thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-09-20 00:30:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' into zvonand-implicit-tz

Browse Source
This commit is contained in:
Andrey Zvonov 2023-04-20 23:13:17 +02:00 committed by GitHub
commit 984e8fde41
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
731 changed files with 17564 additions and 23904 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.github/PULL_REQUEST_TEMPLATE.md vendored
View File

@ -7,11 +7,11 @@ tests/ci/run_check.py
### Changelog category (leave one):
- New Feature
- Improvement
- Bug Fix (user-visible misbehavior in an official stable release)
- Performance Improvement
- Backward Incompatible Change
- Build/Testing/Packaging Improvement
- Documentation (changelog entry is not required)
- Bug Fix (user-visible misbehavior in an official stable release)
- Not for changelog (changelog entry is not required)

38
.github/workflows/pull_request.yml vendored
View File

@ -4846,3 +4846,41 @@ jobs:
cd "$GITHUB_WORKSPACE/tests/ci"
python3 finish_check.py
python3 merge_pr.py --check-approved
##############################################################################################
########################### SQLLOGIC TEST ###################################################
##############################################################################################
SQLLogicTestRelease:
needs: [BuilderDebRelease]
runs-on: [self-hosted, func-tester]
steps:
- name: Set envs
run: |
cat >> "$GITHUB_ENV" << 'EOF'
TEMP_PATH=${{runner.temp}}/sqllogic_debug
REPORTS_PATH=${{runner.temp}}/reports_dir
CHECK_NAME=Sqllogic test (release)
REPO_COPY=${{runner.temp}}/sqllogic_debug/ClickHouse
KILL_TIMEOUT=10800
EOF
- name: Download json reports
uses: actions/download-artifact@v2
with:
path: ${{ env.REPORTS_PATH }}
- name: Clear repository
run: |
sudo rm -fr "$GITHUB_WORKSPACE" && mkdir "$GITHUB_WORKSPACE"
- name: Check out repository code
uses: actions/checkout@v2
- name: Sqllogic test
run: |
sudo rm -fr "$TEMP_PATH"
mkdir -p "$TEMP_PATH"
cp -r "$GITHUB_WORKSPACE" "$TEMP_PATH"
cd "$REPO_COPY/tests/ci"
python3 sqllogic_test.py "$CHECK_NAME" "$KILL_TIMEOUT"
- name: Cleanup
if: always()
run: |
docker ps --quiet | xargs --no-run-if-empty docker kill ||:
docker ps --all --quiet | xargs --no-run-if-empty docker rm -f ||:
sudo rm -fr "$TEMP_PATH"

6
CMakeLists.txt
View File

@ -393,7 +393,11 @@ else()
endif ()
option (ENABLE_GWP_ASAN "Enable Gwp-Asan" ON)
if (NOT OS_LINUX AND NOT OS_ANDROID)
# We use mmap for allocations more heavily in debug builds,
# but GWP-ASan also wants to use mmap frequently,
# and due to a large number of memory mappings,
# it does not work together well.
if ((NOT OS_LINUX AND NOT OS_ANDROID) OR (CMAKE_BUILD_TYPE_UC STREQUAL "DEBUG"))
set(ENABLE_GWP_ASAN OFF)
endif ()

8
README.md
View File

@ -21,10 +21,10 @@ curl https://clickhouse.com/ | sh
* [Contacts](https://clickhouse.com/company/contact) can help to get your questions answered if there are any.
## Upcoming Events
* [**ClickHouse Meetup in Austin**](https://www.meetup.com/clickhouse-austin-user-group/events/291486654/) - Mar 30 - The first ClickHouse Meetup in Austin is happening soon! Interested in speaking, let us know!
* [**v23.3 Release Webinar**](https://clickhouse.com/company/events/v23-3-release-webinar?utm_source=github&utm_medium=social&utm_campaign=release-webinar-2023-02) - Mar 30 - 23.3 is rapidly approaching. Original creator, co-founder, and CTO of ClickHouse Alexey Milovidov will walk us through the highlights of the release.
* [**ClickHouse Spring Meetup in Manhattan**](https://www.meetup.com/clickhouse-new-york-user-group/events/292517734) - April 26 - It's spring, and it's time to meet again in the city! Talks include: "Building a domain specific query language on top of Clickhouse", "A Galaxy of Information", "Our Journey to ClickHouse Cloud from Redshift", and a ClickHouse update!
* [**v23.4 Release Webinar**](https://clickhouse.com/company/events/v23-4-release-webinar?utm_source=github&utm_medium=social&utm_campaign=release-webinar-2023-04) - April 26 - 23.4 is rapidly approaching. Original creator, co-founder, and CTO of ClickHouse Alexey Milovidov will walk us through the highlights of the release.
## Recent Recordings
* **FOSDEM 2023**: In the "Fast and Streaming Data" room Alexey gave a talk entitled "Building Analytical Apps With ClickHouse" that looks at the landscape of data tools, an interesting data set, and how you can interact with data quickly. Check out the recording on **[YouTube](https://www.youtube.com/watch?v=JlcI2Vfz_uk)**.
* **Recording available**: [**v23.2 Release Webinar**](https://www.youtube.com/watch?v=2o0vRMMIrkY) NTILE Window Function support, Partition Key for GROUP By, io_uring, Apache Iceberg support, Dynamic Disks, integrations updates! Watch it now!
* **Recent Meetup Videos**: [Meetup Playlist](https://www.youtube.com/playlist?list=PL0Z2YDlm0b3iNDUzpY1S3L_iV4nARda_U) Whenever possible recordings of the ClickHouse Community Meetups are edited and presented as individual talks. Current featuring "Modern SQL in 2023", "Fast, Concurrent, and Consistent Asynchronous INSERTS in ClickHouse", and "Full-Text Indices: Design and Experiments"
* **Recording available**: [**v23.3 Release Webinar**](https://www.youtube.com/watch?v=ISaGUjvBNao) UNDROP TABLE, server settings introspection, nested dynamic disks, MySQL compatibility, parseDate Time, Lightweight Deletes, Parallel Replicas, integrations updates, and so much more! Watch it now!
* **All release webinar recordings**: [YouTube playlist](https://www.youtube.com/playlist?list=PL0Z2YDlm0b3jAlSy1JxyP8zluvXaN3nxU)

18
base/base/Decimal.h
View File

@ -1,5 +1,6 @@
#pragma once
#include <base/extended_types.h>
#include <base/Decimal_fwd.h>
#if !defined(NO_SANITIZE_UNDEFINED)
#if defined(__clang__)
@ -19,23 +20,6 @@ using Decimal64 = Decimal<Int64>;
using Decimal128 = Decimal<Int128>;
using Decimal256 = Decimal<Int256>;
template <class T>
concept is_decimal =
std::is_same_v<T, Decimal32>
|| std::is_same_v<T, Decimal64>
|| std::is_same_v<T, Decimal128>
|| std::is_same_v<T, Decimal256>
|| std::is_same_v<T, DateTime64>;
template <class T>
concept is_over_big_int =
std::is_same_v<T, Int128>
|| std::is_same_v<T, UInt128>
|| std::is_same_v<T, Int256>
|| std::is_same_v<T, UInt256>
|| std::is_same_v<T, Decimal128>
|| std::is_same_v<T, Decimal256>;
template <class T> struct NativeTypeT { using Type = T; };
template <is_decimal T> struct NativeTypeT<T> { using Type = typename T::NativeType; };
template <class T> using NativeType = typename NativeTypeT<T>::Type;

46
base/base/Decimal_fwd.h Normal file
View File

@ -0,0 +1,46 @@
#pragma once
#include <base/types.h>
namespace wide
{
template <size_t Bits, typename Signed>
class integer;
}
using Int128 = wide::integer<128, signed>;
using UInt128 = wide::integer<128, unsigned>;
using Int256 = wide::integer<256, signed>;
using UInt256 = wide::integer<256, unsigned>;
namespace DB
{
template <class> struct Decimal;
using Decimal32 = Decimal<Int32>;
using Decimal64 = Decimal<Int64>;
using Decimal128 = Decimal<Int128>;
using Decimal256 = Decimal<Int256>;
class DateTime64;
template <class T>
concept is_decimal =
std::is_same_v<T, Decimal32>
|| std::is_same_v<T, Decimal64>
|| std::is_same_v<T, Decimal128>
|| std::is_same_v<T, Decimal256>
|| std::is_same_v<T, DateTime64>;
template <class T>
concept is_over_big_int =
std::is_same_v<T, Int128>
|| std::is_same_v<T, UInt128>
|| std::is_same_v<T, Int256>
|| std::is_same_v<T, UInt256>
|| std::is_same_v<T, Decimal128>
|| std::is_same_v<T, Decimal256>;
}

12
base/base/IPv4andIPv6.h
View File

@ -51,3 +51,15 @@ namespace DB
};
}
namespace std
{
template <>
struct hash<DB::IPv6>
{
size_t operator()(const DB::IPv6 & x) const
{
return std::hash<DB::IPv6::UnderlyingType>()(x.toUnderType());
}
};
}

3
base/poco/Foundation/CMakeLists.txt
View File

@ -233,7 +233,8 @@ target_link_libraries (_poco_foundation
PRIVATE
Poco::Foundation::PCRE
ch_contrib::zlib
ch_contrib::lz4)
ch_contrib::lz4
ch_contrib::double_conversion)
if(OS_DARWIN AND ARCH_AARCH64)
target_compile_definitions (_poco_foundation

20
base/poco/Foundation/src/NumericString.cpp
View File

@ -14,23 +14,9 @@
#include "Poco/Bugcheck.h"
// +++ double conversion +++
#define double_conversion poco_double_conversion // don't collide with standalone double_conversion library
#define UNREACHABLE poco_bugcheck
#define UNIMPLEMENTED poco_bugcheck
#include "diy-fp.cc"
#include "cached-powers.cc"
#include "bignum-dtoa.cc"
#include "bignum.cc"
#include "fast-dtoa.cc"
#include "fixed-dtoa.cc"
#include "strtod.cc"
#include "double-conversion.cc"
// --- double conversion ---
#include <double-conversion/double-conversion.h>
#include "Poco/NumericString.h"
poco_static_assert(POCO_MAX_FLT_STRING_LEN == double_conversion::kMaxSignificantDecimalDigits);
#include "Poco/String.h"
#include <memory>
#include <cctype>
@ -263,7 +249,7 @@ float strToFloat(const char* str)
int processed;
int flags = StringToDoubleConverter::ALLOW_LEADING_SPACES |
StringToDoubleConverter::ALLOW_TRAILING_SPACES;
StringToDoubleConverter converter(flags, 0.0, Single::NaN(), POCO_FLT_INF, POCO_FLT_NAN);
StringToDoubleConverter converter(flags, 0.0, std::numeric_limits<float>::quiet_NaN(), POCO_FLT_INF, POCO_FLT_NAN);
float result = converter.StringToFloat(str, static_cast<int>(strlen(str)), &processed);
return result;
}
@ -275,7 +261,7 @@ double strToDouble(const char* str)
int processed;
int flags = StringToDoubleConverter::ALLOW_LEADING_SPACES |
StringToDoubleConverter::ALLOW_TRAILING_SPACES;
StringToDoubleConverter converter(flags, 0.0, Double::NaN(), POCO_FLT_INF, POCO_FLT_NAN);
StringToDoubleConverter converter(flags, 0.0, std::numeric_limits<double>::quiet_NaN(), POCO_FLT_INF, POCO_FLT_NAN);
double result = converter.StringToDouble(str, static_cast<int>(strlen(str)), &processed);
return result;
}

188
base/poco/Foundation/src/adler32.c
View File

@ -1,188 +0,0 @@
/* adler32.c -- compute the Adler-32 checksum of a data stream
* Copyright (C) 1995-2011, 2016 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* @(#) $Id$ */
#include "zutil.h"
#define local static
local uLong adler32_combine_ OF((uLong adler1, uLong adler2, z_off64_t len2));
#define BASE 65521U /* largest prime smaller than 65536 */
#define NMAX 5552
/* NMAX is the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1 */
#define DO1(buf,i) {adler += (buf)[i]; sum2 += adler;}
#define DO2(buf,i) DO1(buf,i); DO1(buf,i+1);
#define DO4(buf,i) DO2(buf,i); DO2(buf,i+2);
#define DO8(buf,i) DO4(buf,i); DO4(buf,i+4);
#define DO16(buf) DO8(buf,0); DO8(buf,8);
/* use NO_DIVIDE if your processor does not do division in hardware --
try it both ways to see which is faster */
#ifdef NO_DIVIDE
/* note that this assumes BASE is 65521, where 65536 % 65521 == 15
(thank you to John Reiser for pointing this out) */
# define CHOP(a) \
do { \
unsigned long tmp = a >> 16; \
a &= 0xffffUL; \
a += (tmp << 4) - tmp; \
} while (0)
# define MOD28(a) \
do { \
CHOP(a); \
if (a >= BASE) a -= BASE; \
} while (0)
# define MOD(a) \
do { \
CHOP(a); \
MOD28(a); \
} while (0)
# define MOD63(a) \
do { /* this assumes a is not negative */ \
z_off64_t tmp = a >> 32; \
a &= 0xffffffffL; \
a += (tmp << 8) - (tmp << 5) + tmp; \
tmp = a >> 16; \
a &= 0xffffL; \
a += (tmp << 4) - tmp; \
tmp = a >> 16; \
a &= 0xffffL; \
a += (tmp << 4) - tmp; \
if (a >= BASE) a -= BASE; \
} while (0)
#else
# define MOD(a) a %= BASE
# define MOD28(a) a %= BASE
# define MOD63(a) a %= BASE
#endif
/* ========================================================================= */
uLong ZEXPORT adler32_z(adler, buf, len)
uLong adler;
const Bytef *buf;
z_size_t len;
{
unsigned long sum2;
unsigned n;
/* split Adler-32 into component sums */
sum2 = (adler >> 16) & 0xffff;
adler &= 0xffff;
/* in case user likes doing a byte at a time, keep it fast */
if (len == 1) {
adler += buf[0];
if (adler >= BASE)
adler -= BASE;
sum2 += adler;
if (sum2 >= BASE)
sum2 -= BASE;
return adler | (sum2 << 16);
}
/* initial Adler-32 value (deferred check for len == 1 speed) */
if (buf == Z_NULL)
return 1L;
/* in case short lengths are provided, keep it somewhat fast */
if (len < 16) {
while (len--) {
adler += *buf++;
sum2 += adler;
}
if (adler >= BASE)
adler -= BASE;
MOD28(sum2); /* only added so many BASE's */
return adler | (sum2 << 16);
}
/* do length NMAX blocks -- requires just one modulo operation */
while (len >= NMAX) {
len -= NMAX;
n = NMAX / 16; /* NMAX is divisible by 16 */
do {
DO16(buf); /* 16 sums unrolled */
buf += 16;
} while (--n);
MOD(adler);
MOD(sum2);
}
/* do remaining bytes (less than NMAX, still just one modulo) */
if (len) { /* avoid modulos if none remaining */
while (len >= 16) {
len -= 16;
DO16(buf);
buf += 16;
}
while (len--) {
adler += *buf++;
sum2 += adler;
}
MOD(adler);
MOD(sum2);
}
/* return recombined sums */
return adler | (sum2 << 16);
}
/* ========================================================================= */
uLong ZEXPORT adler32(adler, buf, len)
uLong adler;
const Bytef *buf;
uInt len;
{
return adler32_z(adler, buf, len);
}
/* ========================================================================= */
local uLong adler32_combine_(adler1, adler2, len2)
uLong adler1;
uLong adler2;
z_off64_t len2;
{
unsigned long sum1;
unsigned long sum2;
unsigned rem;
/* for negative len, return invalid adler32 as a clue for debugging */
if (len2 < 0)
return 0xffffffffUL;
/* the derivation of this formula is left as an exercise for the reader */
MOD63(len2); /* assumes len2 >= 0 */
rem = (unsigned)len2;
sum1 = adler1 & 0xffff;
sum2 = rem * sum1;
MOD(sum2);
sum1 += (adler2 & 0xffff) + BASE - 1;
sum2 += ((adler1 >> 16) & 0xffff) + ((adler2 >> 16) & 0xffff) + BASE - rem;
if (sum1 >= BASE) sum1 -= BASE;
if (sum1 >= BASE) sum1 -= BASE;
if (sum2 >= ((unsigned long)BASE << 1)) sum2 -= ((unsigned long)BASE << 1);
if (sum2 >= BASE) sum2 -= BASE;
return sum1 | (sum2 << 16);
}
/* ========================================================================= */
uLong ZEXPORT adler32_combine(adler1, adler2, len2)
uLong adler1;
uLong adler2;
z_off_t len2;
{
return adler32_combine_(adler1, adler2, len2);
}
uLong ZEXPORT adler32_combine64(adler1, adler2, len2)
uLong adler1;
uLong adler2;
z_off64_t len2;
{
return adler32_combine_(adler1, adler2, len2);
}

641
base/poco/Foundation/src/bignum-dtoa.cc
View File

@ -1,641 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <math.h>
#include "bignum-dtoa.h"
#include "bignum.h"
#include "ieee.h"
namespace double_conversion {
static int NormalizedExponent(uint64_t significand, int exponent) {
ASSERT(significand != 0);
while ((significand & Double::kHiddenBit) == 0) {
significand = significand << 1;
exponent = exponent - 1;
}
return exponent;
}
// Forward declarations:
// Returns an estimation of k such that 10^(k-1) <= v < 10^k.
static int EstimatePower(int exponent);
// Computes v / 10^estimated_power exactly, as a ratio of two bignums, numerator
// and denominator.
static void InitialScaledStartValues(uint64_t significand,
int exponent,
bool lower_boundary_is_closer,
int estimated_power,
bool need_boundary_deltas,
Bignum* numerator,
Bignum* denominator,
Bignum* delta_minus,
Bignum* delta_plus);
// Multiplies numerator/denominator so that its values lies in the range 1-10.
// Returns decimal_point s.t.
// v = numerator'/denominator' * 10^(decimal_point-1)
// where numerator' and denominator' are the values of numerator and
// denominator after the call to this function.
static void FixupMultiply10(int estimated_power, bool is_even,
int* decimal_point,
Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus);
// Generates digits from the left to the right and stops when the generated
// digits yield the shortest decimal representation of v.
static void GenerateShortestDigits(Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus,
bool is_even,
Vector<char> buffer, int* length);
// Generates 'requested_digits' after the decimal point.
static void BignumToFixed(int requested_digits, int* decimal_point,
Bignum* numerator, Bignum* denominator,
Vector<char>(buffer), int* length);
// Generates 'count' digits of numerator/denominator.
// Once 'count' digits have been produced rounds the result depending on the
// remainder (remainders of exactly .5 round upwards). Might update the
// decimal_point when rounding up (for example for 0.9999).
static void GenerateCountedDigits(int count, int* decimal_point,
Bignum* numerator, Bignum* denominator,
Vector<char>(buffer), int* length);
void BignumDtoa(double v, BignumDtoaMode mode, int requested_digits,
Vector<char> buffer, int* length, int* decimal_point) {
ASSERT(v > 0);
ASSERT(!Double(v).IsSpecial());
uint64_t significand;
int exponent;
bool lower_boundary_is_closer;
if (mode == BIGNUM_DTOA_SHORTEST_SINGLE) {
float f = static_cast<float>(v);
ASSERT(f == v);
significand = Single(f).Significand();
exponent = Single(f).Exponent();
lower_boundary_is_closer = Single(f).LowerBoundaryIsCloser();
} else {
significand = Double(v).Significand();
exponent = Double(v).Exponent();
lower_boundary_is_closer = Double(v).LowerBoundaryIsCloser();
}
bool need_boundary_deltas =
(mode == BIGNUM_DTOA_SHORTEST || mode == BIGNUM_DTOA_SHORTEST_SINGLE);
bool is_even = (significand & 1) == 0;
int normalized_exponent = NormalizedExponent(significand, exponent);
// estimated_power might be too low by 1.
int estimated_power = EstimatePower(normalized_exponent);
// Shortcut for Fixed.
// The requested digits correspond to the digits after the point. If the
// number is much too small, then there is no need in trying to get any
// digits.
if (mode == BIGNUM_DTOA_FIXED && -estimated_power - 1 > requested_digits) {
buffer[0] = '\0';
*length = 0;
// Set decimal-point to -requested_digits. This is what Gay does.
// Note that it should not have any effect anyways since the string is
// empty.
*decimal_point = -requested_digits;
return;
}
Bignum numerator;
Bignum denominator;
Bignum delta_minus;
Bignum delta_plus;
// Make sure the bignum can grow large enough. The smallest double equals
// 4e-324. In this case the denominator needs fewer than 324*4 binary digits.
// The maximum double is 1.7976931348623157e308 which needs fewer than
// 308*4 binary digits.
ASSERT(Bignum::kMaxSignificantBits >= 324*4);
InitialScaledStartValues(significand, exponent, lower_boundary_is_closer,
estimated_power, need_boundary_deltas,
&numerator, &denominator,
&delta_minus, &delta_plus);
// We now have v = (numerator / denominator) * 10^estimated_power.
FixupMultiply10(estimated_power, is_even, decimal_point,
&numerator, &denominator,
&delta_minus, &delta_plus);
// We now have v = (numerator / denominator) * 10^(decimal_point-1), and
// 1 <= (numerator + delta_plus) / denominator < 10
switch (mode) {
case BIGNUM_DTOA_SHORTEST:
case BIGNUM_DTOA_SHORTEST_SINGLE:
GenerateShortestDigits(&numerator, &denominator,
&delta_minus, &delta_plus,
is_even, buffer, length);
break;
case BIGNUM_DTOA_FIXED:
BignumToFixed(requested_digits, decimal_point,
&numerator, &denominator,
buffer, length);
break;
case BIGNUM_DTOA_PRECISION:
GenerateCountedDigits(requested_digits, decimal_point,
&numerator, &denominator,
buffer, length);
break;
default:
UNREACHABLE();
}
buffer[*length] = '\0';
}
// The procedure starts generating digits from the left to the right and stops
// when the generated digits yield the shortest decimal representation of v. A
// decimal representation of v is a number lying closer to v than to any other
// double, so it converts to v when read.
//
// This is true if d, the decimal representation, is between m- and m+, the
// upper and lower boundaries. d must be strictly between them if !is_even.
// m- := (numerator - delta_minus) / denominator
// m+ := (numerator + delta_plus) / denominator
//
// Precondition: 0 <= (numerator+delta_plus) / denominator < 10.
// If 1 <= (numerator+delta_plus) / denominator < 10 then no leading 0 digit
// will be produced. This should be the standard precondition.
static void GenerateShortestDigits(Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus,
bool is_even,
Vector<char> buffer, int* length) {
// Small optimization: if delta_minus and delta_plus are the same just reuse
// one of the two bignums.
if (Bignum::Equal(*delta_minus, *delta_plus)) {
delta_plus = delta_minus;
}
*length = 0;
for (;;) {
uint16_t digit;
digit = numerator->DivideModuloIntBignum(*denominator);
ASSERT(digit <= 9); // digit is a uint16_t and therefore always positive.
// digit = numerator / denominator (integer division).
// numerator = numerator % denominator.
buffer[(*length)++] = static_cast<char>(digit + '0');
// Can we stop already?
// If the remainder of the division is less than the distance to the lower
// boundary we can stop. In this case we simply round down (discarding the
// remainder).
// Similarly we test if we can round up (using the upper boundary).
bool in_delta_room_minus;
bool in_delta_room_plus;
if (is_even) {
in_delta_room_minus = Bignum::LessEqual(*numerator, *delta_minus);
} else {
in_delta_room_minus = Bignum::Less(*numerator, *delta_minus);
}
if (is_even) {
in_delta_room_plus =
Bignum::PlusCompare(*numerator, *delta_plus, *denominator) >= 0;
} else {
in_delta_room_plus =
Bignum::PlusCompare(*numerator, *delta_plus, *denominator) > 0;
}
if (!in_delta_room_minus && !in_delta_room_plus) {
// Prepare for next iteration.
numerator->Times10();
delta_minus->Times10();
// We optimized delta_plus to be equal to delta_minus (if they share the
// same value). So don't multiply delta_plus if they point to the same
// object.
if (delta_minus != delta_plus) {
delta_plus->Times10();
}
} else if (in_delta_room_minus && in_delta_room_plus) {
// Let's see if 2*numerator < denominator.
// If yes, then the next digit would be < 5 and we can round down.
int compare = Bignum::PlusCompare(*numerator, *numerator, *denominator);
if (compare < 0) {
// Remaining digits are less than .5. -> Round down (== do nothing).
} else if (compare > 0) {
// Remaining digits are more than .5 of denominator. -> Round up.
// Note that the last digit could not be a '9' as otherwise the whole
// loop would have stopped earlier.
// We still have an assert here in case the preconditions were not
// satisfied.
ASSERT(buffer[(*length) - 1] != '9');
buffer[(*length) - 1]++;
} else {
// Halfway case.
// TODO(floitsch): need a way to solve half-way cases.
// For now let's round towards even (since this is what Gay seems to
// do).
if ((buffer[(*length) - 1] - '0') % 2 == 0) {
// Round down => Do nothing.
} else {
ASSERT(buffer[(*length) - 1] != '9');
buffer[(*length) - 1]++;
}
}
return;
} else if (in_delta_room_minus) {
// Round down (== do nothing).
return;
} else { // in_delta_room_plus
// Round up.
// Note again that the last digit could not be '9' since this would have
// stopped the loop earlier.
// We still have an ASSERT here, in case the preconditions were not
// satisfied.
ASSERT(buffer[(*length) -1] != '9');
buffer[(*length) - 1]++;
return;
}
}
}
// Let v = numerator / denominator < 10.
// Then we generate 'count' digits of d = x.xxxxx... (without the decimal point)
// from left to right. Once 'count' digits have been produced we decide whether
// to round up or down. Remainders of exactly .5 round upwards. Numbers such
// as 9.999999 propagate a carry all the way, and change the
// exponent (decimal_point), when rounding upwards.
static void GenerateCountedDigits(int count, int* decimal_point,
Bignum* numerator, Bignum* denominator,
Vector<char> buffer, int* length) {
ASSERT(count >= 0);
for (int i = 0; i < count - 1; ++i) {
uint16_t digit;
digit = numerator->DivideModuloIntBignum(*denominator);
ASSERT(digit <= 9); // digit is a uint16_t and therefore always positive.
// digit = numerator / denominator (integer division).
// numerator = numerator % denominator.
buffer[i] = static_cast<char>(digit + '0');
// Prepare for next iteration.
numerator->Times10();
}
// Generate the last digit.
uint16_t digit;
digit = numerator->DivideModuloIntBignum(*denominator);
if (Bignum::PlusCompare(*numerator, *numerator, *denominator) >= 0) {
digit++;
}
ASSERT(digit <= 10);
buffer[count - 1] = static_cast<char>(digit + '0');
// Correct bad digits (in case we had a sequence of '9's). Propagate the
// carry until we hat a non-'9' or til we reach the first digit.
for (int i = count - 1; i > 0; --i) {
if (buffer[i] != '0' + 10) break;
buffer[i] = '0';
buffer[i - 1]++;
}
if (buffer[0] == '0' + 10) {
// Propagate a carry past the top place.
buffer[0] = '1';
(*decimal_point)++;
}
*length = count;
}
// Generates 'requested_digits' after the decimal point. It might omit
// trailing '0's. If the input number is too small then no digits at all are
// generated (ex.: 2 fixed digits for 0.00001).
//
// Input verifies: 1 <= (numerator + delta) / denominator < 10.
static void BignumToFixed(int requested_digits, int* decimal_point,
Bignum* numerator, Bignum* denominator,
Vector<char>(buffer), int* length) {
// Note that we have to look at more than just the requested_digits, since
// a number could be rounded up. Example: v=0.5 with requested_digits=0.
// Even though the power of v equals 0 we can't just stop here.
if (-(*decimal_point) > requested_digits) {
// The number is definitively too small.
// Ex: 0.001 with requested_digits == 1.
// Set decimal-point to -requested_digits. This is what Gay does.
// Note that it should not have any effect anyways since the string is
// empty.
*decimal_point = -requested_digits;
*length = 0;
return;
} else if (-(*decimal_point) == requested_digits) {
// We only need to verify if the number rounds down or up.
// Ex: 0.04 and 0.06 with requested_digits == 1.
ASSERT(*decimal_point == -requested_digits);
// Initially the fraction lies in range (1, 10]. Multiply the denominator
// by 10 so that we can compare more easily.
denominator->Times10();
if (Bignum::PlusCompare(*numerator, *numerator, *denominator) >= 0) {
// If the fraction is >= 0.5 then we have to include the rounded
// digit.
buffer[0] = '1';
*length = 1;
(*decimal_point)++;
} else {
// Note that we caught most of similar cases earlier.
*length = 0;
}
return;
} else {
// The requested digits correspond to the digits after the point.
// The variable 'needed_digits' includes the digits before the point.
int needed_digits = (*decimal_point) + requested_digits;
GenerateCountedDigits(needed_digits, decimal_point,
numerator, denominator,
buffer, length);
}
}
// Returns an estimation of k such that 10^(k-1) <= v < 10^k where
// v = f * 2^exponent and 2^52 <= f < 2^53.
// v is hence a normalized double with the given exponent. The output is an
// approximation for the exponent of the decimal approimation .digits * 10^k.
//
// The result might undershoot by 1 in which case 10^k <= v < 10^k+1.
// Note: this property holds for v's upper boundary m+ too.
// 10^k <= m+ < 10^k+1.
// (see explanation below).
//
// Examples:
// EstimatePower(0) => 16
// EstimatePower(-52) => 0
//
// Note: e >= 0 => EstimatedPower(e) > 0. No similar claim can be made for e<0.
static int EstimatePower(int exponent) {
// This function estimates log10 of v where v = f*2^e (with e == exponent).
// Note that 10^floor(log10(v)) <= v, but v <= 10^ceil(log10(v)).
// Note that f is bounded by its container size. Let p = 53 (the double's
// significand size). Then 2^(p-1) <= f < 2^p.
//
// Given that log10(v) == log2(v)/log2(10) and e+(len(f)-1) is quite close
// to log2(v) the function is simplified to (e+(len(f)-1)/log2(10)).
// The computed number undershoots by less than 0.631 (when we compute log3
// and not log10).
//
// Optimization: since we only need an approximated result this computation
// can be performed on 64 bit integers. On x86/x64 architecture the speedup is
// not really measurable, though.
//
// Since we want to avoid overshooting we decrement by 1e10 so that
// floating-point imprecisions don't affect us.
//
// Explanation for v's boundary m+: the computation takes advantage of
// the fact that 2^(p-1) <= f < 2^p. Boundaries still satisfy this requirement
// (even for denormals where the delta can be much more important).
const double k1Log10 = 0.30102999566398114; // 1/lg(10)
// For doubles len(f) == 53 (don't forget the hidden bit).
const int kSignificandSize = Double::kSignificandSize;
double estimate = ceil((exponent + kSignificandSize - 1) * k1Log10 - 1e-10);
return static_cast<int>(estimate);
}
// See comments for InitialScaledStartValues.
static void InitialScaledStartValuesPositiveExponent(
uint64_t significand, int exponent,
int estimated_power, bool need_boundary_deltas,
Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus) {
// A positive exponent implies a positive power.
ASSERT(estimated_power >= 0);
// Since the estimated_power is positive we simply multiply the denominator
// by 10^estimated_power.
// numerator = v.
numerator->AssignUInt64(significand);
numerator->ShiftLeft(exponent);
// denominator = 10^estimated_power.
denominator->AssignPowerUInt16(10, estimated_power);
if (need_boundary_deltas) {
// Introduce a common denominator so that the deltas to the boundaries are
// integers.
denominator->ShiftLeft(1);
numerator->ShiftLeft(1);
// Let v = f * 2^e, then m+ - v = 1/2 * 2^e; With the common
// denominator (of 2) delta_plus equals 2^e.
delta_plus->AssignUInt16(1);
delta_plus->ShiftLeft(exponent);
// Same for delta_minus. The adjustments if f == 2^p-1 are done later.
delta_minus->AssignUInt16(1);
delta_minus->ShiftLeft(exponent);
}
}
// See comments for InitialScaledStartValues
static void InitialScaledStartValuesNegativeExponentPositivePower(
uint64_t significand, int exponent,
int estimated_power, bool need_boundary_deltas,
Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus) {
// v = f * 2^e with e < 0, and with estimated_power >= 0.
// This means that e is close to 0 (have a look at how estimated_power is
// computed).
// numerator = significand
// since v = significand * 2^exponent this is equivalent to
// numerator = v * / 2^-exponent
numerator->AssignUInt64(significand);
// denominator = 10^estimated_power * 2^-exponent (with exponent < 0)
denominator->AssignPowerUInt16(10, estimated_power);
denominator->ShiftLeft(-exponent);
if (need_boundary_deltas) {
// Introduce a common denominator so that the deltas to the boundaries are
// integers.
denominator->ShiftLeft(1);
numerator->ShiftLeft(1);
// Let v = f * 2^e, then m+ - v = 1/2 * 2^e; With the common
// denominator (of 2) delta_plus equals 2^e.
// Given that the denominator already includes v's exponent the distance
// to the boundaries is simply 1.
delta_plus->AssignUInt16(1);
// Same for delta_minus. The adjustments if f == 2^p-1 are done later.
delta_minus->AssignUInt16(1);
}
}
// See comments for InitialScaledStartValues
static void InitialScaledStartValuesNegativeExponentNegativePower(
uint64_t significand, int exponent,
int estimated_power, bool need_boundary_deltas,
Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus) {
// Instead of multiplying the denominator with 10^estimated_power we
// multiply all values (numerator and deltas) by 10^-estimated_power.
// Use numerator as temporary container for power_ten.
Bignum* power_ten = numerator;
power_ten->AssignPowerUInt16(10, -estimated_power);
if (need_boundary_deltas) {
// Since power_ten == numerator we must make a copy of 10^estimated_power
// before we complete the computation of the numerator.
// delta_plus = delta_minus = 10^estimated_power
delta_plus->AssignBignum(*power_ten);
delta_minus->AssignBignum(*power_ten);
}
// numerator = significand * 2 * 10^-estimated_power
// since v = significand * 2^exponent this is equivalent to
// numerator = v * 10^-estimated_power * 2 * 2^-exponent.
// Remember: numerator has been abused as power_ten. So no need to assign it
// to itself.
ASSERT(numerator == power_ten);
numerator->MultiplyByUInt64(significand);
// denominator = 2 * 2^-exponent with exponent < 0.
denominator->AssignUInt16(1);
denominator->ShiftLeft(-exponent);
if (need_boundary_deltas) {
// Introduce a common denominator so that the deltas to the boundaries are
// integers.
numerator->ShiftLeft(1);
denominator->ShiftLeft(1);
// With this shift the boundaries have their correct value, since
// delta_plus = 10^-estimated_power, and
// delta_minus = 10^-estimated_power.
// These assignments have been done earlier.
// The adjustments if f == 2^p-1 (lower boundary is closer) are done later.
}
}
// Let v = significand * 2^exponent.
// Computes v / 10^estimated_power exactly, as a ratio of two bignums, numerator
// and denominator. The functions GenerateShortestDigits and
// GenerateCountedDigits will then convert this ratio to its decimal
// representation d, with the required accuracy.
// Then d * 10^estimated_power is the representation of v.
// (Note: the fraction and the estimated_power might get adjusted before
// generating the decimal representation.)
//
// The initial start values consist of:
// - a scaled numerator: s.t. numerator/denominator == v / 10^estimated_power.
// - a scaled (common) denominator.
// optionally (used by GenerateShortestDigits to decide if it has the shortest
// decimal converting back to v):
// - v - m-: the distance to the lower boundary.
// - m+ - v: the distance to the upper boundary.
//
// v, m+, m-, and therefore v - m- and m+ - v all share the same denominator.
//
// Let ep == estimated_power, then the returned values will satisfy:
// v / 10^ep = numerator / denominator.
// v's boundaries m- and m+:
// m- / 10^ep == v / 10^ep - delta_minus / denominator
// m+ / 10^ep == v / 10^ep + delta_plus / denominator
// Or in other words:
// m- == v - delta_minus * 10^ep / denominator;
// m+ == v + delta_plus * 10^ep / denominator;
//
// Since 10^(k-1) <= v < 10^k (with k == estimated_power)
// or 10^k <= v < 10^(k+1)
// we then have 0.1 <= numerator/denominator < 1
// or 1 <= numerator/denominator < 10
//
// It is then easy to kickstart the digit-generation routine.
//
// The boundary-deltas are only filled if the mode equals BIGNUM_DTOA_SHORTEST
// or BIGNUM_DTOA_SHORTEST_SINGLE.
static void InitialScaledStartValues(uint64_t significand,
int exponent,
bool lower_boundary_is_closer,
int estimated_power,
bool need_boundary_deltas,
Bignum* numerator,
Bignum* denominator,
Bignum* delta_minus,
Bignum* delta_plus) {
if (exponent >= 0) {
InitialScaledStartValuesPositiveExponent(
significand, exponent, estimated_power, need_boundary_deltas,
numerator, denominator, delta_minus, delta_plus);
} else if (estimated_power >= 0) {
InitialScaledStartValuesNegativeExponentPositivePower(
significand, exponent, estimated_power, need_boundary_deltas,
numerator, denominator, delta_minus, delta_plus);
} else {
InitialScaledStartValuesNegativeExponentNegativePower(
significand, exponent, estimated_power, need_boundary_deltas,
numerator, denominator, delta_minus, delta_plus);
}
if (need_boundary_deltas && lower_boundary_is_closer) {
// The lower boundary is closer at half the distance of "normal" numbers.
// Increase the common denominator and adapt all but the delta_minus.
denominator->ShiftLeft(1); // *2
numerator->ShiftLeft(1); // *2
delta_plus->ShiftLeft(1); // *2
}
}
// This routine multiplies numerator/denominator so that its values lies in the
// range 1-10. That is after a call to this function we have:
// 1 <= (numerator + delta_plus) /denominator < 10.
// Let numerator the input before modification and numerator' the argument
// after modification, then the output-parameter decimal_point is such that
// numerator / denominator * 10^estimated_power ==
// numerator' / denominator' * 10^(decimal_point - 1)
// In some cases estimated_power was too low, and this is already the case. We
// then simply adjust the power so that 10^(k-1) <= v < 10^k (with k ==
// estimated_power) but do not touch the numerator or denominator.
// Otherwise the routine multiplies the numerator and the deltas by 10.
static void FixupMultiply10(int estimated_power, bool is_even,
int* decimal_point,
Bignum* numerator, Bignum* denominator,
Bignum* delta_minus, Bignum* delta_plus) {
bool in_range;
if (is_even) {
// For IEEE doubles half-way cases (in decimal system numbers ending with 5)
// are rounded to the closest floating-point number with even significand.
in_range = Bignum::PlusCompare(*numerator, *delta_plus, *denominator) >= 0;
} else {
in_range = Bignum::PlusCompare(*numerator, *delta_plus, *denominator) > 0;
}
if (in_range) {
// Since numerator + delta_plus >= denominator we already have
// 1 <= numerator/denominator < 10. Simply update the estimated_power.
*decimal_point = estimated_power + 1;
} else {
*decimal_point = estimated_power;
numerator->Times10();
if (Bignum::Equal(*delta_minus, *delta_plus)) {
delta_minus->Times10();
delta_plus->AssignBignum(*delta_minus);
} else {
delta_minus->Times10();
delta_plus->Times10();
}
}
}
} // namespace double_conversion

85
base/poco/Foundation/src/bignum-dtoa.h
View File

@ -1,85 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_BIGNUM_DTOA_H_
#define DOUBLE_CONVERSION_BIGNUM_DTOA_H_
#include "utils.h"
namespace double_conversion
{
enum BignumDtoaMode
{
// Return the shortest correct representation.
// For example the output of 0.299999999999999988897 is (the less accurate but
// correct) 0.3.
BIGNUM_DTOA_SHORTEST,
// Same as BIGNUM_DTOA_SHORTEST but for single-precision floats.
BIGNUM_DTOA_SHORTEST_SINGLE,
// Return a fixed number of digits after the decimal point.
// For instance fixed(0.1, 4) becomes 0.1000
// If the input number is big, the output will be big.
BIGNUM_DTOA_FIXED,
// Return a fixed number of digits, no matter what the exponent is.
BIGNUM_DTOA_PRECISION
};
// Converts the given double 'v' to ascii.
// The result should be interpreted as buffer * 10^(point-length).
// The buffer will be null-terminated.
//
// The input v must be > 0 and different from NaN, and Infinity.
//
// The output depends on the given mode:
// - SHORTEST: produce the least amount of digits for which the internal
// identity requirement is still satisfied. If the digits are printed
// (together with the correct exponent) then reading this number will give
// 'v' again. The buffer will choose the representation that is closest to
// 'v'. If there are two at the same distance, than the number is round up.
// In this mode the 'requested_digits' parameter is ignored.
// - FIXED: produces digits necessary to print a given number with
// 'requested_digits' digits after the decimal point. The produced digits
// might be too short in which case the caller has to fill the gaps with '0's.
// Example: toFixed(0.001, 5) is allowed to return buffer="1", point=-2.
// Halfway cases are rounded up. The call toFixed(0.15, 2) thus returns
// buffer="2", point=0.
// Note: the length of the returned buffer has no meaning wrt the significance
// of its digits. That is, just because it contains '0's does not mean that
// any other digit would not satisfy the internal identity requirement.
// - PRECISION: produces 'requested_digits' where the first digit is not '0'.
// Even though the length of produced digits usually equals
// 'requested_digits', the function is allowed to return fewer digits, in
// which case the caller has to fill the missing digits with '0's.
// Halfway cases are again rounded up.
// 'BignumDtoa' expects the given buffer to be big enough to hold all digits
// and a terminating null-character.
void BignumDtoa(double v, BignumDtoaMode mode, int requested_digits, Vector<char> buffer, int * length, int * point);
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_BIGNUM_DTOA_H_

766
base/poco/Foundation/src/bignum.cc
View File

@ -1,766 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include "bignum.h"
#include "utils.h"
namespace double_conversion {
Bignum::Bignum()
: bigits_(bigits_buffer_, kBigitCapacity), used_digits_(0), exponent_(0) {
for (int i = 0; i < kBigitCapacity; ++i) {
bigits_[i] = 0;
}
}
template<typename S>
static int BitSize(S value) {
(void) value; // Mark variable as used.
return 8 * sizeof(value);
}
// Guaranteed to lie in one Bigit.
void Bignum::AssignUInt16(uint16_t value) {
ASSERT(kBigitSize >= BitSize(value));
Zero();
if (value == 0) return;
EnsureCapacity(1);
bigits_[0] = value;
used_digits_ = 1;
}
void Bignum::AssignUInt64(uint64_t value) {
const int kUInt64Size = 64;
Zero();
if (value == 0) return;
int needed_bigits = kUInt64Size / kBigitSize + 1;
EnsureCapacity(needed_bigits);
for (int i = 0; i < needed_bigits; ++i) {
bigits_[i] = value & kBigitMask;
value = value >> kBigitSize;
}
used_digits_ = needed_bigits;
Clamp();
}
void Bignum::AssignBignum(const Bignum& other) {
exponent_ = other.exponent_;
for (int i = 0; i < other.used_digits_; ++i) {
bigits_[i] = other.bigits_[i];
}
// Clear the excess digits (if there were any).
for (int i = other.used_digits_; i < used_digits_; ++i) {
bigits_[i] = 0;
}
used_digits_ = other.used_digits_;
}
static uint64_t ReadUInt64(Vector<const char> buffer,
int from,
int digits_to_read) {
uint64_t result = 0;
for (int i = from; i < from + digits_to_read; ++i) {
int digit = buffer[i] - '0';
ASSERT(0 <= digit && digit <= 9);
result = result * 10 + digit;
}
return result;
}
void Bignum::AssignDecimalString(Vector<const char> value) {
// 2^64 = 18446744073709551616 > 10^19
const int kMaxUint64DecimalDigits = 19;
Zero();
int length = value.length();
int pos = 0;
// Let's just say that each digit needs 4 bits.
while (length >= kMaxUint64DecimalDigits) {
uint64_t digits = ReadUInt64(value, pos, kMaxUint64DecimalDigits);
pos += kMaxUint64DecimalDigits;
length -= kMaxUint64DecimalDigits;
MultiplyByPowerOfTen(kMaxUint64DecimalDigits);
AddUInt64(digits);
}
uint64_t digits = ReadUInt64(value, pos, length);
MultiplyByPowerOfTen(length);
AddUInt64(digits);
Clamp();
}
static int HexCharValue(char c) {
if ('0' <= c && c <= '9') return c - '0';
if ('a' <= c && c <= 'f') return 10 + c - 'a';
ASSERT('A' <= c && c <= 'F');
return 10 + c - 'A';
}
void Bignum::AssignHexString(Vector<const char> value) {
Zero();
int length = value.length();
int needed_bigits = length * 4 / kBigitSize + 1;
EnsureCapacity(needed_bigits);
int string_index = length - 1;
for (int i = 0; i < needed_bigits - 1; ++i) {
// These bigits are guaranteed to be "full".
Chunk current_bigit = 0;
for (int j = 0; j < kBigitSize / 4; j++) {
current_bigit += HexCharValue(value[string_index--]) << (j * 4);
}
bigits_[i] = current_bigit;
}
used_digits_ = needed_bigits - 1;
Chunk most_significant_bigit = 0; // Could be = 0;
for (int j = 0; j <= string_index; ++j) {
most_significant_bigit <<= 4;
most_significant_bigit += HexCharValue(value[j]);
}
if (most_significant_bigit != 0) {
bigits_[used_digits_] = most_significant_bigit;
used_digits_++;
}
Clamp();
}
void Bignum::AddUInt64(uint64_t operand) {
if (operand == 0) return;
Bignum other;
other.AssignUInt64(operand);
AddBignum(other);
}
void Bignum::AddBignum(const Bignum& other) {
ASSERT(IsClamped());
ASSERT(other.IsClamped());
// If this has a greater exponent than other append zero-bigits to this.
// After this call exponent_ <= other.exponent_.
Align(other);
// There are two possibilities:
// aaaaaaaaaaa 0000 (where the 0s represent a's exponent)
// bbbbb 00000000
// ----------------
// ccccccccccc 0000
// or
// aaaaaaaaaa 0000
// bbbbbbbbb 0000000
// -----------------
// cccccccccccc 0000
// In both cases we might need a carry bigit.
EnsureCapacity(1 + Max(BigitLength(), other.BigitLength()) - exponent_);
Chunk carry = 0;
int bigit_pos = other.exponent_ - exponent_;
ASSERT(bigit_pos >= 0);
for (int i = 0; i < other.used_digits_; ++i) {
Chunk sum = bigits_[bigit_pos] + other.bigits_[i] + carry;
bigits_[bigit_pos] = sum & kBigitMask;
carry = sum >> kBigitSize;
bigit_pos++;
}
while (carry != 0) {
Chunk sum = bigits_[bigit_pos] + carry;
bigits_[bigit_pos] = sum & kBigitMask;
carry = sum >> kBigitSize;
bigit_pos++;
}
used_digits_ = Max(bigit_pos, used_digits_);
ASSERT(IsClamped());
}
void Bignum::SubtractBignum(const Bignum& other) {
ASSERT(IsClamped());
ASSERT(other.IsClamped());
// We require this to be bigger than other.
ASSERT(LessEqual(other, *this));
Align(other);
int offset = other.exponent_ - exponent_;
Chunk borrow = 0;
int i;
for (i = 0; i < other.used_digits_; ++i) {
ASSERT((borrow == 0) || (borrow == 1));
Chunk difference = bigits_[i + offset] - other.bigits_[i] - borrow;
bigits_[i + offset] = difference & kBigitMask;
borrow = difference >> (kChunkSize - 1);
}
while (borrow != 0) {
Chunk difference = bigits_[i + offset] - borrow;
bigits_[i + offset] = difference & kBigitMask;
borrow = difference >> (kChunkSize - 1);
++i;
}
Clamp();
}
void Bignum::ShiftLeft(int shift_amount) {
if (used_digits_ == 0) return;
exponent_ += shift_amount / kBigitSize;
int local_shift = shift_amount % kBigitSize;
EnsureCapacity(used_digits_ + 1);
BigitsShiftLeft(local_shift);
}
void Bignum::MultiplyByUInt32(uint32_t factor) {
if (factor == 1) return;
if (factor == 0) {
Zero();
return;
}
if (used_digits_ == 0) return;
// The product of a bigit with the factor is of size kBigitSize + 32.
// Assert that this number + 1 (for the carry) fits into double chunk.
ASSERT(kDoubleChunkSize >= kBigitSize + 32 + 1);
DoubleChunk carry = 0;
for (int i = 0; i < used_digits_; ++i) {
DoubleChunk product = static_cast<DoubleChunk>(factor) * bigits_[i] + carry;
bigits_[i] = static_cast<Chunk>(product & kBigitMask);
carry = (product >> kBigitSize);
}
while (carry != 0) {
EnsureCapacity(used_digits_ + 1);
bigits_[used_digits_] = carry & kBigitMask;
used_digits_++;
carry >>= kBigitSize;
}
}
void Bignum::MultiplyByUInt64(uint64_t factor) {
if (factor == 1) return;
if (factor == 0) {
Zero();
return;
}
ASSERT(kBigitSize < 32);
uint64_t carry = 0;
uint64_t low = factor & 0xFFFFFFFF;
uint64_t high = factor >> 32;
for (int i = 0; i < used_digits_; ++i) {
uint64_t product_low = low * bigits_[i];
uint64_t product_high = high * bigits_[i];
uint64_t tmp = (carry & kBigitMask) + product_low;
bigits_[i] = tmp & kBigitMask;
carry = (carry >> kBigitSize) + (tmp >> kBigitSize) +
(product_high << (32 - kBigitSize));
}
while (carry != 0) {
EnsureCapacity(used_digits_ + 1);
bigits_[used_digits_] = carry & kBigitMask;
used_digits_++;
carry >>= kBigitSize;
}
}
void Bignum::MultiplyByPowerOfTen(int exponent) {
const uint64_t kFive27 = UINT64_2PART_C(0x6765c793, fa10079d);
const uint16_t kFive1 = 5;
const uint16_t kFive2 = kFive1 * 5;
const uint16_t kFive3 = kFive2 * 5;
const uint16_t kFive4 = kFive3 * 5;
const uint16_t kFive5 = kFive4 * 5;
const uint16_t kFive6 = kFive5 * 5;
const uint32_t kFive7 = kFive6 * 5;
const uint32_t kFive8 = kFive7 * 5;
const uint32_t kFive9 = kFive8 * 5;
const uint32_t kFive10 = kFive9 * 5;
const uint32_t kFive11 = kFive10 * 5;
const uint32_t kFive12 = kFive11 * 5;
const uint32_t kFive13 = kFive12 * 5;
const uint32_t kFive1_to_12[] =
{ kFive1, kFive2, kFive3, kFive4, kFive5, kFive6,
kFive7, kFive8, kFive9, kFive10, kFive11, kFive12 };
ASSERT(exponent >= 0);
if (exponent == 0) return;
if (used_digits_ == 0) return;
// We shift by exponent at the end just before returning.
int remaining_exponent = exponent;
while (remaining_exponent >= 27) {
MultiplyByUInt64(kFive27);
remaining_exponent -= 27;
}
while (remaining_exponent >= 13) {
MultiplyByUInt32(kFive13);
remaining_exponent -= 13;
}
if (remaining_exponent > 0) {
MultiplyByUInt32(kFive1_to_12[remaining_exponent - 1]);
}
ShiftLeft(exponent);
}
void Bignum::Square() {
ASSERT(IsClamped());
int product_length = 2 * used_digits_;
EnsureCapacity(product_length);
// Comba multiplication: compute each column separately.
// Example: r = a2a1a0 * b2b1b0.
// r = 1 * a0b0 +
// 10 * (a1b0 + a0b1) +
// 100 * (a2b0 + a1b1 + a0b2) +
// 1000 * (a2b1 + a1b2) +
// 10000 * a2b2
//
// In the worst case we have to accumulate nb-digits products of digit*digit.
//
// Assert that the additional number of bits in a DoubleChunk are enough to
// sum up used_digits of Bigit*Bigit.
if ((1 << (2 * (kChunkSize - kBigitSize))) <= used_digits_) {
UNIMPLEMENTED();
}
DoubleChunk accumulator = 0;
// First shift the digits so we don't overwrite them.
int copy_offset = used_digits_;
for (int i = 0; i < used_digits_; ++i) {
bigits_[copy_offset + i] = bigits_[i];
}
// We have two loops to avoid some 'if's in the loop.
for (int i = 0; i < used_digits_; ++i) {
// Process temporary digit i with power i.
// The sum of the two indices must be equal to i.
int bigit_index1 = i;
int bigit_index2 = 0;
// Sum all of the sub-products.
while (bigit_index1 >= 0) {
Chunk chunk1 = bigits_[copy_offset + bigit_index1];
Chunk chunk2 = bigits_[copy_offset + bigit_index2];
accumulator += static_cast<DoubleChunk>(chunk1) * chunk2;
bigit_index1--;
bigit_index2++;
}
bigits_[i] = static_cast<Chunk>(accumulator) & kBigitMask;
accumulator >>= kBigitSize;
}
for (int i = used_digits_; i < product_length; ++i) {
int bigit_index1 = used_digits_ - 1;
int bigit_index2 = i - bigit_index1;
// Invariant: sum of both indices is again equal to i.
// Inner loop runs 0 times on last iteration, emptying accumulator.
while (bigit_index2 < used_digits_) {
Chunk chunk1 = bigits_[copy_offset + bigit_index1];
Chunk chunk2 = bigits_[copy_offset + bigit_index2];
accumulator += static_cast<DoubleChunk>(chunk1) * chunk2;
bigit_index1--;
bigit_index2++;
}
// The overwritten bigits_[i] will never be read in further loop iterations,
// because bigit_index1 and bigit_index2 are always greater
// than i - used_digits_.
bigits_[i] = static_cast<Chunk>(accumulator) & kBigitMask;
accumulator >>= kBigitSize;
}
// Since the result was guaranteed to lie inside the number the
// accumulator must be 0 now.
ASSERT(accumulator == 0);
// Don't forget to update the used_digits and the exponent.
used_digits_ = product_length;
exponent_ *= 2;
Clamp();
}
void Bignum::AssignPowerUInt16(uint16_t base, int power_exponent) {
ASSERT(base != 0);
ASSERT(power_exponent >= 0);
if (power_exponent == 0) {
AssignUInt16(1);
return;
}
Zero();
int shifts = 0;
// We expect base to be in range 2-32, and most often to be 10.
// It does not make much sense to implement different algorithms for counting
// the bits.
while ((base & 1) == 0) {
base >>= 1;
shifts++;
}
int bit_size = 0;
int tmp_base = base;
while (tmp_base != 0) {
tmp_base >>= 1;
bit_size++;
}
int final_size = bit_size * power_exponent;
// 1 extra bigit for the shifting, and one for rounded final_size.
EnsureCapacity(final_size / kBigitSize + 2);
// Left to Right exponentiation.
int mask = 1;
while (power_exponent >= mask) mask <<= 1;
// The mask is now pointing to the bit above the most significant 1-bit of
// power_exponent.
// Get rid of first 1-bit;
mask >>= 2;
uint64_t this_value = base;
bool delayed_multipliciation = false;
const uint64_t max_32bits = 0xFFFFFFFF;
while (mask != 0 && this_value <= max_32bits) {
this_value = this_value * this_value;
// Verify that there is enough space in this_value to perform the
// multiplication. The first bit_size bits must be 0.
if ((power_exponent & mask) != 0) {
uint64_t base_bits_mask =
~((static_cast<uint64_t>(1) << (64 - bit_size)) - 1);
bool high_bits_zero = (this_value & base_bits_mask) == 0;
if (high_bits_zero) {
this_value *= base;
} else {
delayed_multipliciation = true;
}
}
mask >>= 1;
}
AssignUInt64(this_value);
if (delayed_multipliciation) {
MultiplyByUInt32(base);
}
// Now do the same thing as a bignum.
while (mask != 0) {
Square();
if ((power_exponent & mask) != 0) {
MultiplyByUInt32(base);
}
mask >>= 1;
}
// And finally add the saved shifts.
ShiftLeft(shifts * power_exponent);
}
// Precondition: this/other < 16bit.
uint16_t Bignum::DivideModuloIntBignum(const Bignum& other) {
ASSERT(IsClamped());
ASSERT(other.IsClamped());
ASSERT(other.used_digits_ > 0);
// Easy case: if we have less digits than the divisor than the result is 0.
// Note: this handles the case where this == 0, too.
if (BigitLength() < other.BigitLength()) {
return 0;
}
Align(other);
uint16_t result = 0;
// Start by removing multiples of 'other' until both numbers have the same
// number of digits.
while (BigitLength() > other.BigitLength()) {
// This naive approach is extremely inefficient if `this` divided by other
// is big. This function is implemented for doubleToString where
// the result should be small (less than 10).
ASSERT(other.bigits_[other.used_digits_ - 1] >= ((1 << kBigitSize) / 16));
ASSERT(bigits_[used_digits_ - 1] < 0x10000);
// Remove the multiples of the first digit.
// Example this = 23 and other equals 9. -> Remove 2 multiples.
result += static_cast<uint16_t>(bigits_[used_digits_ - 1]);
SubtractTimes(other, bigits_[used_digits_ - 1]);
}
ASSERT(BigitLength() == other.BigitLength());
// Both bignums are at the same length now.
// Since other has more than 0 digits we know that the access to
// bigits_[used_digits_ - 1] is safe.
Chunk this_bigit = bigits_[used_digits_ - 1];
Chunk other_bigit = other.bigits_[other.used_digits_ - 1];
if (other.used_digits_ == 1) {
// Shortcut for easy (and common) case.
int quotient = this_bigit / other_bigit;
bigits_[used_digits_ - 1] = this_bigit - other_bigit * quotient;
ASSERT(quotient < 0x10000);
result += static_cast<uint16_t>(quotient);
Clamp();
return result;
}
int division_estimate = this_bigit / (other_bigit + 1);
ASSERT(division_estimate < 0x10000);
result += static_cast<uint16_t>(division_estimate);
SubtractTimes(other, division_estimate);
if (other_bigit * (division_estimate + 1) > this_bigit) {
// No need to even try to subtract. Even if other's remaining digits were 0
// another subtraction would be too much.
return result;
}
while (LessEqual(other, *this)) {
SubtractBignum(other);
result++;
}
return result;
}
template<typename S>
static int SizeInHexChars(S number) {
ASSERT(number > 0);
int result = 0;
while (number != 0) {
number >>= 4;
result++;
}
return result;
}
static char HexCharOfValue(int value) {
ASSERT(0 <= value && value <= 16);
if (value < 10) return static_cast<char>(value + '0');
return static_cast<char>(value - 10 + 'A');
}
bool Bignum::ToHexString(char* buffer, int buffer_size) const {
ASSERT(IsClamped());
// Each bigit must be printable as separate hex-character.
ASSERT(kBigitSize % 4 == 0);
const int kHexCharsPerBigit = kBigitSize / 4;
if (used_digits_ == 0) {
if (buffer_size < 2) return false;
buffer[0] = '0';
buffer[1] = '\0';
return true;
}
// We add 1 for the terminating '\0' character.
int needed_chars = (BigitLength() - 1) * kHexCharsPerBigit +
SizeInHexChars(bigits_[used_digits_ - 1]) + 1;
if (needed_chars > buffer_size) return false;
int string_index = needed_chars - 1;
buffer[string_index--] = '\0';
for (int i = 0; i < exponent_; ++i) {
for (int j = 0; j < kHexCharsPerBigit; ++j) {
buffer[string_index--] = '0';
}
}
for (int i = 0; i < used_digits_ - 1; ++i) {
Chunk current_bigit = bigits_[i];
for (int j = 0; j < kHexCharsPerBigit; ++j) {
buffer[string_index--] = HexCharOfValue(current_bigit & 0xF);
current_bigit >>= 4;
}
}
// And finally the last bigit.
Chunk most_significant_bigit = bigits_[used_digits_ - 1];
while (most_significant_bigit != 0) {
buffer[string_index--] = HexCharOfValue(most_significant_bigit & 0xF);
most_significant_bigit >>= 4;
}
return true;
}
Bignum::Chunk Bignum::BigitAt(int index) const {
if (index >= BigitLength()) return 0;
if (index < exponent_) return 0;
return bigits_[index - exponent_];
}
int Bignum::Compare(const Bignum& a, const Bignum& b) {
ASSERT(a.IsClamped());
ASSERT(b.IsClamped());
int bigit_length_a = a.BigitLength();
int bigit_length_b = b.BigitLength();
if (bigit_length_a < bigit_length_b) return -1;
if (bigit_length_a > bigit_length_b) return +1;
for (int i = bigit_length_a - 1; i >= Min(a.exponent_, b.exponent_); --i) {
Chunk bigit_a = a.BigitAt(i);
Chunk bigit_b = b.BigitAt(i);
if (bigit_a < bigit_b) return -1;
if (bigit_a > bigit_b) return +1;
// Otherwise they are equal up to this digit. Try the next digit.
}
return 0;
}
int Bignum::PlusCompare(const Bignum& a, const Bignum& b, const Bignum& c) {
ASSERT(a.IsClamped());
ASSERT(b.IsClamped());
ASSERT(c.IsClamped());
if (a.BigitLength() < b.BigitLength()) {
return PlusCompare(b, a, c);
}
if (a.BigitLength() + 1 < c.BigitLength()) return -1;
if (a.BigitLength() > c.BigitLength()) return +1;
// The exponent encodes 0-bigits. So if there are more 0-digits in 'a' than
// 'b' has digits, then the bigit-length of 'a'+'b' must be equal to the one
// of 'a'.
if (a.exponent_ >= b.BigitLength() && a.BigitLength() < c.BigitLength()) {
return -1;
}
Chunk borrow = 0;
// Starting at min_exponent all digits are == 0. So no need to compare them.
int min_exponent = Min(Min(a.exponent_, b.exponent_), c.exponent_);
for (int i = c.BigitLength() - 1; i >= min_exponent; --i) {
Chunk chunk_a = a.BigitAt(i);
Chunk chunk_b = b.BigitAt(i);
Chunk chunk_c = c.BigitAt(i);
Chunk sum = chunk_a + chunk_b;
if (sum > chunk_c + borrow) {
return +1;
} else {
borrow = chunk_c + borrow - sum;
if (borrow > 1) return -1;
borrow <<= kBigitSize;
}
}
if (borrow == 0) return 0;
return -1;
}
void Bignum::Clamp() {
while (used_digits_ > 0 && bigits_[used_digits_ - 1] == 0) {
used_digits_--;
}
if (used_digits_ == 0) {
// Zero.
exponent_ = 0;
}
}
bool Bignum::IsClamped() const {
return used_digits_ == 0 || bigits_[used_digits_ - 1] != 0;
}
void Bignum::Zero() {
for (int i = 0; i < used_digits_; ++i) {
bigits_[i] = 0;
}
used_digits_ = 0;
exponent_ = 0;
}
void Bignum::Align(const Bignum& other) {
if (exponent_ > other.exponent_) {
// If "X" represents a "hidden" digit (by the exponent) then we are in the
// following case (a == this, b == other):
// a: aaaaaaXXXX or a: aaaaaXXX
// b: bbbbbbX b: bbbbbbbbXX
// We replace some of the hidden digits (X) of a with 0 digits.
// a: aaaaaa000X or a: aaaaa0XX
int zero_digits = exponent_ - other.exponent_;
EnsureCapacity(used_digits_ + zero_digits);
for (int i = used_digits_ - 1; i >= 0; --i) {
bigits_[i + zero_digits] = bigits_[i];
}
for (int i = 0; i < zero_digits; ++i) {
bigits_[i] = 0;
}
used_digits_ += zero_digits;
exponent_ -= zero_digits;
ASSERT(used_digits_ >= 0);
ASSERT(exponent_ >= 0);
}
}
void Bignum::BigitsShiftLeft(int shift_amount) {
ASSERT(shift_amount < kBigitSize);
ASSERT(shift_amount >= 0);
Chunk carry = 0;
for (int i = 0; i < used_digits_; ++i) {
Chunk new_carry = bigits_[i] >> (kBigitSize - shift_amount);
bigits_[i] = ((bigits_[i] << shift_amount) + carry) & kBigitMask;
carry = new_carry;
}
if (carry != 0) {
bigits_[used_digits_] = carry;
used_digits_++;
}
}
void Bignum::SubtractTimes(const Bignum& other, int factor) {
ASSERT(exponent_ <= other.exponent_);
if (factor < 3) {
for (int i = 0; i < factor; ++i) {
SubtractBignum(other);
}
return;
}
Chunk borrow = 0;
int exponent_diff = other.exponent_ - exponent_;
for (int i = 0; i < other.used_digits_; ++i) {
DoubleChunk product = static_cast<DoubleChunk>(factor) * other.bigits_[i];
DoubleChunk remove = borrow + product;
Chunk difference = bigits_[i + exponent_diff] - (remove & kBigitMask);
bigits_[i + exponent_diff] = difference & kBigitMask;
borrow = static_cast<Chunk>((difference >> (kChunkSize - 1)) +
(remove >> kBigitSize));
}
for (int i = other.used_digits_ + exponent_diff; i < used_digits_; ++i) {
if (borrow == 0) return;
Chunk difference = bigits_[i] - borrow;
bigits_[i] = difference & kBigitMask;
borrow = difference >> (kChunkSize - 1);
}
Clamp();
}
} // namespace double_conversion

138
base/poco/Foundation/src/bignum.h
View File

@ -1,138 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_BIGNUM_H_
#define DOUBLE_CONVERSION_BIGNUM_H_
#include "utils.h"
namespace double_conversion
{
class Bignum
{
public:
// 3584 = 128 * 28. We can represent 2^3584 > 10^1000 accurately.
// This bignum can encode much bigger numbers, since it contains an
// exponent.
static const int kMaxSignificantBits = 3584;
Bignum();
void AssignUInt16(uint16_t value);
void AssignUInt64(uint64_t value);
void AssignBignum(const Bignum & other);
void AssignDecimalString(Vector<const char> value);
void AssignHexString(Vector<const char> value);
void AssignPowerUInt16(uint16_t base, int exponent);
void AddUInt16(uint16_t operand);
void AddUInt64(uint64_t operand);
void AddBignum(const Bignum & other);
// Precondition: this >= other.
void SubtractBignum(const Bignum & other);
void Square();
void ShiftLeft(int shift_amount);
void MultiplyByUInt32(uint32_t factor);
void MultiplyByUInt64(uint64_t factor);
void MultiplyByPowerOfTen(int exponent);
void Times10() { return MultiplyByUInt32(10); }
// Pseudocode:
// int result = this / other;
// this = this % other;
// In the worst case this function is in O(this/other).
uint16_t DivideModuloIntBignum(const Bignum & other);
bool ToHexString(char * buffer, int buffer_size) const;
// Returns
// -1 if a < b,
// 0 if a == b, and
// +1 if a > b.
static int Compare(const Bignum & a, const Bignum & b);
static bool Equal(const Bignum & a, const Bignum & b) { return Compare(a, b) == 0; }
static bool LessEqual(const Bignum & a, const Bignum & b) { return Compare(a, b) <= 0; }
static bool Less(const Bignum & a, const Bignum & b) { return Compare(a, b) < 0; }
// Returns Compare(a + b, c);
static int PlusCompare(const Bignum & a, const Bignum & b, const Bignum & c);
// Returns a + b == c
static bool PlusEqual(const Bignum & a, const Bignum & b, const Bignum & c) { return PlusCompare(a, b, c) == 0; }
// Returns a + b <= c
static bool PlusLessEqual(const Bignum & a, const Bignum & b, const Bignum & c) { return PlusCompare(a, b, c) <= 0; }
// Returns a + b < c
static bool PlusLess(const Bignum & a, const Bignum & b, const Bignum & c) { return PlusCompare(a, b, c) < 0; }
private:
typedef uint32_t Chunk;
typedef uint64_t DoubleChunk;
static const int kChunkSize = sizeof(Chunk) * 8;
static const int kDoubleChunkSize = sizeof(DoubleChunk) * 8;
// With bigit size of 28 we loose some bits, but a double still fits easily
// into two chunks, and more importantly we can use the Comba multiplication.
static const int kBigitSize = 28;
static const Chunk kBigitMask = (1 << kBigitSize) - 1;
// Every instance allocates kBigitLength chunks on the stack. Bignums cannot
// grow. There are no checks if the stack-allocated space is sufficient.
static const int kBigitCapacity = kMaxSignificantBits / kBigitSize;
void EnsureCapacity(int size)
{
if (size > kBigitCapacity)
{
UNREACHABLE();
}
}
void Align(const Bignum & other);
void Clamp();
bool IsClamped() const;
void Zero();
// Requires this to have enough capacity (no tests done).
// Updates used_digits_ if necessary.
// shift_amount must be < kBigitSize.
void BigitsShiftLeft(int shift_amount);
// BigitLength includes the "hidden" digits encoded in the exponent.
int BigitLength() const { return used_digits_ + exponent_; }
Chunk BigitAt(int index) const;
void SubtractTimes(const Bignum & other, int factor);
Chunk bigits_buffer_[kBigitCapacity];
// A vector backed by bigits_buffer_. This way accesses to the array are
// checked for out-of-bounds errors.
Vector<Chunk> bigits_;
int used_digits_;
// The Bignum's value equals value(bigits_) * 2^(exponent_ * kBigitSize).
int exponent_;
DISALLOW_COPY_AND_ASSIGN(Bignum);
};
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_BIGNUM_H_

176
base/poco/Foundation/src/cached-powers.cc
View File

@ -1,176 +0,0 @@
// Copyright 2006-2008 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <stdarg.h>
#include <limits.h>
#include <math.h>
#include "utils.h"
#include "cached-powers.h"
namespace double_conversion {
struct CachedPower {
uint64_t significand;
int16_t binary_exponent;
int16_t decimal_exponent;
};
static const CachedPower kCachedPowers[] = {
{UINT64_2PART_C(0xfa8fd5a0, 081c0288), -1220, -348},
{UINT64_2PART_C(0xbaaee17f, a23ebf76), -1193, -340},
{UINT64_2PART_C(0x8b16fb20, 3055ac76), -1166, -332},
{UINT64_2PART_C(0xcf42894a, 5dce35ea), -1140, -324},
{UINT64_2PART_C(0x9a6bb0aa, 55653b2d), -1113, -316},
{UINT64_2PART_C(0xe61acf03, 3d1a45df), -1087, -308},
{UINT64_2PART_C(0xab70fe17, c79ac6ca), -1060, -300},
{UINT64_2PART_C(0xff77b1fc, bebcdc4f), -1034, -292},
{UINT64_2PART_C(0xbe5691ef, 416bd60c), -1007, -284},
{UINT64_2PART_C(0x8dd01fad, 907ffc3c), -980, -276},
{UINT64_2PART_C(0xd3515c28, 31559a83), -954, -268},
{UINT64_2PART_C(0x9d71ac8f, ada6c9b5), -927, -260},
{UINT64_2PART_C(0xea9c2277, 23ee8bcb), -901, -252},
{UINT64_2PART_C(0xaecc4991, 4078536d), -874, -244},
{UINT64_2PART_C(0x823c1279, 5db6ce57), -847, -236},
{UINT64_2PART_C(0xc2109436, 4dfb5637), -821, -228},
{UINT64_2PART_C(0x9096ea6f, 3848984f), -794, -220},
{UINT64_2PART_C(0xd77485cb, 25823ac7), -768, -212},
{UINT64_2PART_C(0xa086cfcd, 97bf97f4), -741, -204},
{UINT64_2PART_C(0xef340a98, 172aace5), -715, -196},
{UINT64_2PART_C(0xb23867fb, 2a35b28e), -688, -188},
{UINT64_2PART_C(0x84c8d4df, d2c63f3b), -661, -180},
{UINT64_2PART_C(0xc5dd4427, 1ad3cdba), -635, -172},
{UINT64_2PART_C(0x936b9fce, bb25c996), -608, -164},
{UINT64_2PART_C(0xdbac6c24, 7d62a584), -582, -156},
{UINT64_2PART_C(0xa3ab6658, 0d5fdaf6), -555, -148},
{UINT64_2PART_C(0xf3e2f893, dec3f126), -529, -140},
{UINT64_2PART_C(0xb5b5ada8, aaff80b8), -502, -132},
{UINT64_2PART_C(0x87625f05, 6c7c4a8b), -475, -124},
{UINT64_2PART_C(0xc9bcff60, 34c13053), -449, -116},
{UINT64_2PART_C(0x964e858c, 91ba2655), -422, -108},
{UINT64_2PART_C(0xdff97724, 70297ebd), -396, -100},
{UINT64_2PART_C(0xa6dfbd9f, b8e5b88f), -369, -92},
{UINT64_2PART_C(0xf8a95fcf, 88747d94), -343, -84},
{UINT64_2PART_C(0xb9447093, 8fa89bcf), -316, -76},
{UINT64_2PART_C(0x8a08f0f8, bf0f156b), -289, -68},
{UINT64_2PART_C(0xcdb02555, 653131b6), -263, -60},
{UINT64_2PART_C(0x993fe2c6, d07b7fac), -236, -52},
{UINT64_2PART_C(0xe45c10c4, 2a2b3b06), -210, -44},
{UINT64_2PART_C(0xaa242499, 697392d3), -183, -36},
{UINT64_2PART_C(0xfd87b5f2, 8300ca0e), -157, -28},
{UINT64_2PART_C(0xbce50864, 92111aeb), -130, -20},
{UINT64_2PART_C(0x8cbccc09, 6f5088cc), -103, -12},
{UINT64_2PART_C(0xd1b71758, e219652c), -77, -4},
{UINT64_2PART_C(0x9c400000, 00000000), -50, 4},
{UINT64_2PART_C(0xe8d4a510, 00000000), -24, 12},
{UINT64_2PART_C(0xad78ebc5, ac620000), 3, 20},
{UINT64_2PART_C(0x813f3978, f8940984), 30, 28},
{UINT64_2PART_C(0xc097ce7b, c90715b3), 56, 36},
{UINT64_2PART_C(0x8f7e32ce, 7bea5c70), 83, 44},
{UINT64_2PART_C(0xd5d238a4, abe98068), 109, 52},
{UINT64_2PART_C(0x9f4f2726, 179a2245), 136, 60},
{UINT64_2PART_C(0xed63a231, d4c4fb27), 162, 68},
{UINT64_2PART_C(0xb0de6538, 8cc8ada8), 189, 76},
{UINT64_2PART_C(0x83c7088e, 1aab65db), 216, 84},
{UINT64_2PART_C(0xc45d1df9, 42711d9a), 242, 92},
{UINT64_2PART_C(0x924d692c, a61be758), 269, 100},
{UINT64_2PART_C(0xda01ee64, 1a708dea), 295, 108},
{UINT64_2PART_C(0xa26da399, 9aef774a), 322, 116},
{UINT64_2PART_C(0xf209787b, b47d6b85), 348, 124},
{UINT64_2PART_C(0xb454e4a1, 79dd1877), 375, 132},
{UINT64_2PART_C(0x865b8692, 5b9bc5c2), 402, 140},
{UINT64_2PART_C(0xc83553c5, c8965d3d), 428, 148},
{UINT64_2PART_C(0x952ab45c, fa97a0b3), 455, 156},
{UINT64_2PART_C(0xde469fbd, 99a05fe3), 481, 164},
{UINT64_2PART_C(0xa59bc234, db398c25), 508, 172},
{UINT64_2PART_C(0xf6c69a72, a3989f5c), 534, 180},
{UINT64_2PART_C(0xb7dcbf53, 54e9bece), 561, 188},
{UINT64_2PART_C(0x88fcf317, f22241e2), 588, 196},
{UINT64_2PART_C(0xcc20ce9b, d35c78a5), 614, 204},
{UINT64_2PART_C(0x98165af3, 7b2153df), 641, 212},
{UINT64_2PART_C(0xe2a0b5dc, 971f303a), 667, 220},
{UINT64_2PART_C(0xa8d9d153, 5ce3b396), 694, 228},
{UINT64_2PART_C(0xfb9b7cd9, a4a7443c), 720, 236},
{UINT64_2PART_C(0xbb764c4c, a7a44410), 747, 244},
{UINT64_2PART_C(0x8bab8eef, b6409c1a), 774, 252},
{UINT64_2PART_C(0xd01fef10, a657842c), 800, 260},
{UINT64_2PART_C(0x9b10a4e5, e9913129), 827, 268},
{UINT64_2PART_C(0xe7109bfb, a19c0c9d), 853, 276},
{UINT64_2PART_C(0xac2820d9, 623bf429), 880, 284},
{UINT64_2PART_C(0x80444b5e, 7aa7cf85), 907, 292},
{UINT64_2PART_C(0xbf21e440, 03acdd2d), 933, 300},
{UINT64_2PART_C(0x8e679c2f, 5e44ff8f), 960, 308},
{UINT64_2PART_C(0xd433179d, 9c8cb841), 986, 316},
{UINT64_2PART_C(0x9e19db92, b4e31ba9), 1013, 324},
{UINT64_2PART_C(0xeb96bf6e, badf77d9), 1039, 332},
{UINT64_2PART_C(0xaf87023b, 9bf0ee6b), 1066, 340},
};
static const int kCachedPowersLength = ARRAY_SIZE(kCachedPowers);
static const int kCachedPowersOffset = 348; // -1 * the first decimal_exponent.
static const double kD_1_LOG2_10 = 0.30102999566398114; // 1 / lg(10)
// Difference between the decimal exponents in the table above.
const int PowersOfTenCache::kDecimalExponentDistance = 8;
const int PowersOfTenCache::kMinDecimalExponent = -348;
const int PowersOfTenCache::kMaxDecimalExponent = 340;
void PowersOfTenCache::GetCachedPowerForBinaryExponentRange(
int min_exponent,
int max_exponent,
DiyFp* power,
int* decimal_exponent) {
int kQ = DiyFp::kSignificandSize;
double k = ceil((min_exponent + kQ - 1) * kD_1_LOG2_10);
int foo = kCachedPowersOffset;
int index =
(foo + static_cast<int>(k) - 1) / kDecimalExponentDistance + 1;
ASSERT(0 <= index && index < kCachedPowersLength);
CachedPower cached_power = kCachedPowers[index];
ASSERT(min_exponent <= cached_power.binary_exponent);
(void) max_exponent; // Mark variable as used.
ASSERT(cached_power.binary_exponent <= max_exponent);
*decimal_exponent = cached_power.decimal_exponent;
*power = DiyFp(cached_power.significand, cached_power.binary_exponent);
}
void PowersOfTenCache::GetCachedPowerForDecimalExponent(int requested_exponent,
DiyFp* power,
int* found_exponent) {
ASSERT(kMinDecimalExponent <= requested_exponent);
ASSERT(requested_exponent < kMaxDecimalExponent + kDecimalExponentDistance);
int index =
(requested_exponent + kCachedPowersOffset) / kDecimalExponentDistance;
CachedPower cached_power = kCachedPowers[index];
*power = DiyFp(cached_power.significand, cached_power.binary_exponent);
*found_exponent = cached_power.decimal_exponent;
ASSERT(*found_exponent <= requested_exponent);
ASSERT(requested_exponent < *found_exponent + kDecimalExponentDistance);
}
} // namespace double_conversion

60
base/poco/Foundation/src/cached-powers.h
View File

@ -1,60 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_CACHED_POWERS_H_
#define DOUBLE_CONVERSION_CACHED_POWERS_H_
#include "diy-fp.h"
namespace double_conversion
{
class PowersOfTenCache
{
public:
// Not all powers of ten are cached. The decimal exponent of two neighboring
// cached numbers will differ by kDecimalExponentDistance.
static const int kDecimalExponentDistance;
static const int kMinDecimalExponent;
static const int kMaxDecimalExponent;
// Returns a cached power-of-ten with a binary exponent in the range
// [min_exponent; max_exponent] (boundaries included).
static void GetCachedPowerForBinaryExponentRange(int min_exponent, int max_exponent, DiyFp * power, int * decimal_exponent);
// Returns a cached power of ten x ~= 10^k such that
// k <= decimal_exponent < k + kCachedPowersDecimalDistance.
// The given decimal_exponent must satisfy
// kMinDecimalExponent <= requested_exponent, and
// requested_exponent < kMaxDecimalExponent + kDecimalExponentDistance.
static void GetCachedPowerForDecimalExponent(int requested_exponent, DiyFp * power, int * found_exponent);
};
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_CACHED_POWERS_H_

86
base/poco/Foundation/src/compress.c
View File

@ -1,86 +0,0 @@
/* compress.c -- compress a memory buffer
* Copyright (C) 1995-2005, 2014, 2016 Jean-loup Gailly, Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* @(#) $Id$ */
#define ZLIB_INTERNAL
#include "zlib.h"
/* ===========================================================================
Compresses the source buffer into the destination buffer. The level
parameter has the same meaning as in deflateInit. sourceLen is the byte
length of the source buffer. Upon entry, destLen is the total size of the
destination buffer, which must be at least 0.1% larger than sourceLen plus
12 bytes. Upon exit, destLen is the actual size of the compressed buffer.
compress2 returns Z_OK if success, Z_MEM_ERROR if there was not enough
memory, Z_BUF_ERROR if there was not enough room in the output buffer,
Z_STREAM_ERROR if the level parameter is invalid.
*/
int ZEXPORT compress2 (dest, destLen, source, sourceLen, level)
Bytef *dest;
uLongf *destLen;
const Bytef *source;
uLong sourceLen;
int level;
{
z_stream stream;
int err;
const uInt max = (uInt)-1;
uLong left;
left = *destLen;
*destLen = 0;
stream.zalloc = (alloc_func)0;
stream.zfree = (free_func)0;
stream.opaque = (voidpf)0;
err = deflateInit(&stream, level);
if (err != Z_OK) return err;
stream.next_out = dest;
stream.avail_out = 0;
stream.next_in = (z_const Bytef *)source;
stream.avail_in = 0;
do {
if (stream.avail_out == 0) {
stream.avail_out = left > (uLong)max ? max : (uInt)left;
left -= stream.avail_out;
}
if (stream.avail_in == 0) {
stream.avail_in = sourceLen > (uLong)max ? max : (uInt)sourceLen;
sourceLen -= stream.avail_in;
}
err = deflate(&stream, sourceLen ? Z_NO_FLUSH : Z_FINISH);
} while (err == Z_OK);
*destLen = stream.total_out;
deflateEnd(&stream);
return err == Z_STREAM_END ? Z_OK : err;
}
/* ===========================================================================
*/
int ZEXPORT compress (dest, destLen, source, sourceLen)
Bytef *dest;
uLongf *destLen;
const Bytef *source;
uLong sourceLen;
{
return compress2(dest, destLen, source, sourceLen, Z_DEFAULT_COMPRESSION);
}
/* ===========================================================================
If the default memLevel or windowBits for deflateInit() is changed, then
this function needs to be updated.
*/
uLong ZEXPORT compressBound (sourceLen)
uLong sourceLen;
{
return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
(sourceLen >> 25) + 13;
}

444
base/poco/Foundation/src/crc32.c
View File

@ -1,444 +0,0 @@
/* crc32.c -- compute the CRC-32 of a data stream
* Copyright (C) 1995-2006, 2010, 2011, 2012 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*
* Thanks to Rodney Brown <rbrown64@csc.com.au> for his contribution of faster
* CRC methods: exclusive-oring 32 bits of data at a time, and pre-computing
* tables for updating the shift register in one step with three exclusive-ors
* instead of four steps with four exclusive-ors. This results in about a
* factor of two increase in speed on a Power PC G4 (PPC7455) using gcc -O3.
*/
/* @(#) $Id$ */
/*
Note on the use of DYNAMIC_CRC_TABLE: there is no mutex or semaphore
protection on the static variables used to control the first-use generation
of the crc tables. Therefore, if you #define DYNAMIC_CRC_TABLE, you should
first call get_crc_table() to initialize the tables before allowing more than
one thread to use crc32().
DYNAMIC_CRC_TABLE and MAKECRCH can be #defined to write out crc32.h.
*/
#ifdef MAKECRCH
# include <stdio.h>
# ifndef DYNAMIC_CRC_TABLE
# define DYNAMIC_CRC_TABLE
# endif /* !DYNAMIC_CRC_TABLE */
#endif /* MAKECRCH */
#include "zutil.h" /* for STDC and FAR definitions */
#define local static
/* Definitions for doing the crc four data bytes at a time. */
#if !defined(NOBYFOUR) && defined(Z_U4)
# define BYFOUR
#endif
#ifdef BYFOUR
local unsigned long crc32_little OF((unsigned long,
const unsigned char FAR *, z_size_t));
local unsigned long crc32_big OF((unsigned long,
const unsigned char FAR *, z_size_t));
# define TBLS 8
#else
# define TBLS 1
#endif /* BYFOUR */
/* Local functions for crc concatenation */
local unsigned long gf2_matrix_times OF((unsigned long *mat,
unsigned long vec));
local void gf2_matrix_square OF((unsigned long *square, unsigned long *mat));
local uLong crc32_combine_ OF((uLong crc1, uLong crc2, z_off64_t len2));
#ifdef DYNAMIC_CRC_TABLE
local volatile int crc_table_empty = 1;
local z_crc_t FAR crc_table[TBLS][256];
local void make_crc_table OF((void));
#ifdef MAKECRCH
local void write_table OF((FILE *, const z_crc_t FAR *));
#endif /* MAKECRCH */
/*
Generate tables for a byte-wise 32-bit CRC calculation on the polynomial:
x^32+x^26+x^23+x^22+x^16+x^12+x^11+x^10+x^8+x^7+x^5+x^4+x^2+x+1.
Polynomials over GF(2) are represented in binary, one bit per coefficient,
with the lowest powers in the most significant bit. Then adding polynomials
is just exclusive-or, and multiplying a polynomial by x is a right shift by
one. If we call the above polynomial p, and represent a byte as the
polynomial q, also with the lowest power in the most significant bit (so the
byte 0xb1 is the polynomial x^7+x^3+x+1), then the CRC is (q*x^32) mod p,
where a mod b means the remainder after dividing a by b.
This calculation is done using the shift-register method of multiplying and
taking the remainder. The register is initialized to zero, and for each
incoming bit, x^32 is added mod p to the register if the bit is a one (where
x^32 mod p is p+x^32 = x^26+...+1), and the register is multiplied mod p by
x (which is shifting right by one and adding x^32 mod p if the bit shifted
out is a one). We start with the highest power (least significant bit) of
q and repeat for all eight bits of q.
The first table is simply the CRC of all possible eight bit values. This is
all the information needed to generate CRCs on data a byte at a time for all
combinations of CRC register values and incoming bytes. The remaining tables
allow for word-at-a-time CRC calculation for both big-endian and little-
endian machines, where a word is four bytes.
*/
local void make_crc_table()
{
z_crc_t c;
int n, k;
z_crc_t poly; /* polynomial exclusive-or pattern */
/* terms of polynomial defining this crc (except x^32): */
static volatile int first = 1; /* flag to limit concurrent making */
static const unsigned char p[] = {0,1,2,4,5,7,8,10,11,12,16,22,23,26};
/* See if another task is already doing this (not thread-safe, but better
than nothing -- significantly reduces duration of vulnerability in
case the advice about DYNAMIC_CRC_TABLE is ignored) */
if (first) {
first = 0;
/* make exclusive-or pattern from polynomial (0xedb88320UL) */
poly = 0;
for (n = 0; n < (int)(sizeof(p)/sizeof(unsigned char)); n++)
poly |= (z_crc_t)1 << (31 - p[n]);
/* generate a crc for every 8-bit value */
for (n = 0; n < 256; n++) {
c = (z_crc_t)n;
for (k = 0; k < 8; k++)
c = c & 1 ? poly ^ (c >> 1) : c >> 1;
crc_table[0][n] = c;
}
#ifdef BYFOUR
/* generate crc for each value followed by one, two, and three zeros,
and then the byte reversal of those as well as the first table */
for (n = 0; n < 256; n++) {
c = crc_table[0][n];
crc_table[4][n] = ZSWAP32(c);
for (k = 1; k < 4; k++) {
c = crc_table[0][c & 0xff] ^ (c >> 8);
crc_table[k][n] = c;
crc_table[k + 4][n] = ZSWAP32(c);
}
}
#endif /* BYFOUR */
crc_table_empty = 0;
}
else { /* not first */
/* wait for the other guy to finish (not efficient, but rare) */
while (crc_table_empty)
;
}
#ifdef MAKECRCH
/* write out CRC tables to crc32.h */
{
FILE *out;
out = fopen("crc32.h", "w");
if (out == NULL) return;
fprintf(out, "/* crc32.h -- tables for rapid CRC calculation\n");
fprintf(out, " * Generated automatically by crc32.c\n */\n\n");
fprintf(out, "local const z_crc_t FAR ");
fprintf(out, "crc_table[TBLS][256] =\n{\n {\n");
write_table(out, crc_table[0]);
# ifdef BYFOUR
fprintf(out, "#ifdef BYFOUR\n");
for (k = 1; k < 8; k++) {
fprintf(out, " },\n {\n");
write_table(out, crc_table[k]);
}
fprintf(out, "#endif\n");
# endif /* BYFOUR */
fprintf(out, " }\n};\n");
fclose(out);
}
#endif /* MAKECRCH */
}
#ifdef MAKECRCH
local void write_table(out, table)
FILE *out;
const z_crc_t FAR *table;
{
int n;
for (n = 0; n < 256; n++)
fprintf(out, "%s0x%08lxUL%s", n % 5 ? "" : " ",
(unsigned long)(table[n]),
n == 255 ? "\n" : (n % 5 == 4 ? ",\n" : ", "));
}
#endif /* MAKECRCH */
#else /* !DYNAMIC_CRC_TABLE */
/* ========================================================================
* Tables of CRC-32s of all single-byte values, made by make_crc_table().
*/
#include "crc32.h"
#endif /* DYNAMIC_CRC_TABLE */
/* =========================================================================
* This function can be used by asm versions of crc32()
*/
const z_crc_t FAR * ZEXPORT get_crc_table()
{
#ifdef DYNAMIC_CRC_TABLE
if (crc_table_empty)
make_crc_table();
#endif /* DYNAMIC_CRC_TABLE */
return (const z_crc_t FAR *)crc_table;
}
/* ========================================================================= */
#define DO1 crc = crc_table[0][((int)crc ^ (*buf++)) & 0xff] ^ (crc >> 8)
#define DO8 DO1; DO1; DO1; DO1; DO1; DO1; DO1; DO1
/* ========================================================================= */
unsigned long ZEXPORT crc32_z(crc, buf, len)
unsigned long crc;
const unsigned char FAR *buf;
z_size_t len;
{
if (buf == Z_NULL) return 0UL;
#ifdef DYNAMIC_CRC_TABLE
if (crc_table_empty)
make_crc_table();
#endif /* DYNAMIC_CRC_TABLE */
#ifdef BYFOUR
if (sizeof(void *) == sizeof(ptrdiff_t)) {
z_crc_t endian;
endian = 1;
if (*((unsigned char *)(&endian)))
return crc32_little(crc, buf, len);
else
return crc32_big(crc, buf, len);
}
#endif /* BYFOUR */
crc = crc ^ 0xffffffffUL;
while (len >= 8) {
DO8;
len -= 8;
}
if (len) do {
DO1;
} while (--len);
return crc ^ 0xffffffffUL;
}
/* ========================================================================= */
unsigned long ZEXPORT crc32(crc, buf, len)
unsigned long crc;
const unsigned char FAR *buf;
uInt len;
{
return crc32_z(crc, buf, len);
}
#ifdef BYFOUR
/*
This BYFOUR code accesses the passed unsigned char * buffer with a 32-bit
integer pointer type. This violates the strict aliasing rule, where a
compiler can assume, for optimization purposes, that two pointers to
fundamentally different types won't ever point to the same memory. This can
manifest as a problem only if one of the pointers is written to. This code
only reads from those pointers. So long as this code remains isolated in
this compilation unit, there won't be a problem. For this reason, this code
should not be copied and pasted into a compilation unit in which other code
writes to the buffer that is passed to these routines.
*/
/* ========================================================================= */
#define DOLIT4 c ^= *buf4++; \
c = crc_table[3][c & 0xff] ^ crc_table[2][(c >> 8) & 0xff] ^ \
crc_table[1][(c >> 16) & 0xff] ^ crc_table[0][c >> 24]
#define DOLIT32 DOLIT4; DOLIT4; DOLIT4; DOLIT4; DOLIT4; DOLIT4; DOLIT4; DOLIT4
/* ========================================================================= */
local unsigned long crc32_little(crc, buf, len)
unsigned long crc;
const unsigned char FAR *buf;
z_size_t len;
{
register z_crc_t c;
register const z_crc_t FAR *buf4;
c = (z_crc_t)crc;
c = ~c;
while (len && ((ptrdiff_t)buf & 3)) {
c = crc_table[0][(c ^ *buf++) & 0xff] ^ (c >> 8);
len--;
}
buf4 = (const z_crc_t FAR *)(const void FAR *)buf;
while (len >= 32) {
DOLIT32;
len -= 32;
}
while (len >= 4) {
DOLIT4;
len -= 4;
}
buf = (const unsigned char FAR *)buf4;
if (len) do {
c = crc_table[0][(c ^ *buf++) & 0xff] ^ (c >> 8);
} while (--len);
c = ~c;
return (unsigned long)c;
}
/* ========================================================================= */
#define DOBIG4 c ^= *buf4++; \
c = crc_table[4][c & 0xff] ^ crc_table[5][(c >> 8) & 0xff] ^ \
crc_table[6][(c >> 16) & 0xff] ^ crc_table[7][c >> 24]
#define DOBIG32 DOBIG4; DOBIG4; DOBIG4; DOBIG4; DOBIG4; DOBIG4; DOBIG4; DOBIG4
/* ========================================================================= */
local unsigned long crc32_big(crc, buf, len)
unsigned long crc;
const unsigned char FAR *buf;
z_size_t len;
{
register z_crc_t c;
register const z_crc_t FAR *buf4;
c = ZSWAP32((z_crc_t)crc);
c = ~c;
while (len && ((ptrdiff_t)buf & 3)) {
c = crc_table[4][(c >> 24) ^ *buf++] ^ (c << 8);
len--;
}
buf4 = (const z_crc_t FAR *)(const void FAR *)buf;
while (len >= 32) {
DOBIG32;
len -= 32;
}
while (len >= 4) {
DOBIG4;
len -= 4;
}
buf = (const unsigned char FAR *)buf4;
if (len) do {
c = crc_table[4][(c >> 24) ^ *buf++] ^ (c << 8);
} while (--len);
c = ~c;
return (unsigned long)(ZSWAP32(c));
}
#endif /* BYFOUR */
#define GF2_DIM 32 /* dimension of GF(2) vectors (length of CRC) */
/* ========================================================================= */
local unsigned long gf2_matrix_times(mat, vec)
unsigned long *mat;
unsigned long vec;
{
unsigned long sum;
sum = 0;
while (vec) {
if (vec & 1)
sum ^= *mat;
vec >>= 1;
mat++;
}
return sum;
}
/* ========================================================================= */
local void gf2_matrix_square(square, mat)
unsigned long *square;
unsigned long *mat;
{
int n;
for (n = 0; n < GF2_DIM; n++)
square[n] = gf2_matrix_times(mat, mat[n]);
}
/* ========================================================================= */
local uLong crc32_combine_(crc1, crc2, len2)
uLong crc1;
uLong crc2;
z_off64_t len2;
{
int n;
unsigned long row;
unsigned long even[GF2_DIM]; /* even-power-of-two zeros operator */
unsigned long odd[GF2_DIM]; /* odd-power-of-two zeros operator */
/* degenerate case (also disallow negative lengths) */
if (len2 <= 0)
return crc1;
/* put operator for one zero bit in odd */
odd[0] = 0xedb88320UL; /* CRC-32 polynomial */
row = 1;
for (n = 1; n < GF2_DIM; n++) {
odd[n] = row;
row <<= 1;
}
/* put operator for two zero bits in even */
gf2_matrix_square(even, odd);
/* put operator for four zero bits in odd */
gf2_matrix_square(odd, even);
/* apply len2 zeros to crc1 (first square will put the operator for one
zero byte, eight zero bits, in even) */
do {
/* apply zeros operator for this bit of len2 */
gf2_matrix_square(even, odd);
if (len2 & 1)
crc1 = gf2_matrix_times(even, crc1);
len2 >>= 1;
/* if no more bits set, then done */
if (len2 == 0)
break;
/* another iteration of the loop with odd and even swapped */
gf2_matrix_square(odd, even);
if (len2 & 1)
crc1 = gf2_matrix_times(odd, crc1);
len2 >>= 1;
/* if no more bits set, then done */
} while (len2 != 0);
/* return combined crc */
crc1 ^= crc2;
return crc1;
}
/* ========================================================================= */
uLong ZEXPORT crc32_combine(crc1, crc2, len2)
uLong crc1;
uLong crc2;
z_off_t len2;
{
return crc32_combine_(crc1, crc2, len2);
}
uLong ZEXPORT crc32_combine64(crc1, crc2, len2)
uLong crc1;
uLong crc2;
z_off64_t len2;
{
return crc32_combine_(crc1, crc2, len2);
}

241
base/poco/Foundation/src/crc32.h
View File

@ -1,241 +0,0 @@
/* crc32.h -- tables for rapid CRC calculation
* Generated automatically by crc32.c
*/
local const z_crc_t FAR crc_table[TBLS][256]
= {{0x00000000UL, 0x77073096UL, 0xee0e612cUL, 0x990951baUL, 0x076dc419UL, 0x706af48fUL, 0xe963a535UL, 0x9e6495a3UL, 0x0edb8832UL,
0x79dcb8a4UL, 0xe0d5e91eUL, 0x97d2d988UL, 0x09b64c2bUL, 0x7eb17cbdUL, 0xe7b82d07UL, 0x90bf1d91UL, 0x1db71064UL, 0x6ab020f2UL,
0xf3b97148UL, 0x84be41deUL, 0x1adad47dUL, 0x6ddde4ebUL, 0xf4d4b551UL, 0x83d385c7UL, 0x136c9856UL, 0x646ba8c0UL, 0xfd62f97aUL,
0x8a65c9ecUL, 0x14015c4fUL, 0x63066cd9UL, 0xfa0f3d63UL, 0x8d080df5UL, 0x3b6e20c8UL, 0x4c69105eUL, 0xd56041e4UL, 0xa2677172UL,
0x3c03e4d1UL, 0x4b04d447UL, 0xd20d85fdUL, 0xa50ab56bUL, 0x35b5a8faUL, 0x42b2986cUL, 0xdbbbc9d6UL, 0xacbcf940UL, 0x32d86ce3UL,
0x45df5c75UL, 0xdcd60dcfUL, 0xabd13d59UL, 0x26d930acUL, 0x51de003aUL, 0xc8d75180UL, 0xbfd06116UL, 0x21b4f4b5UL, 0x56b3c423UL,
0xcfba9599UL, 0xb8bda50fUL, 0x2802b89eUL, 0x5f058808UL, 0xc60cd9b2UL, 0xb10be924UL, 0x2f6f7c87UL, 0x58684c11UL, 0xc1611dabUL,
0xb6662d3dUL, 0x76dc4190UL, 0x01db7106UL, 0x98d220bcUL, 0xefd5102aUL, 0x71b18589UL, 0x06b6b51fUL, 0x9fbfe4a5UL, 0xe8b8d433UL,
0x7807c9a2UL, 0x0f00f934UL, 0x9609a88eUL, 0xe10e9818UL, 0x7f6a0dbbUL, 0x086d3d2dUL, 0x91646c97UL, 0xe6635c01UL, 0x6b6b51f4UL,
0x1c6c6162UL, 0x856530d8UL, 0xf262004eUL, 0x6c0695edUL, 0x1b01a57bUL, 0x8208f4c1UL, 0xf50fc457UL, 0x65b0d9c6UL, 0x12b7e950UL,
0x8bbeb8eaUL, 0xfcb9887cUL, 0x62dd1ddfUL, 0x15da2d49UL, 0x8cd37cf3UL, 0xfbd44c65UL, 0x4db26158UL, 0x3ab551ceUL, 0xa3bc0074UL,
0xd4bb30e2UL, 0x4adfa541UL, 0x3dd895d7UL, 0xa4d1c46dUL, 0xd3d6f4fbUL, 0x4369e96aUL, 0x346ed9fcUL, 0xad678846UL, 0xda60b8d0UL,
0x44042d73UL, 0x33031de5UL, 0xaa0a4c5fUL, 0xdd0d7cc9UL, 0x5005713cUL, 0x270241aaUL, 0xbe0b1010UL, 0xc90c2086UL, 0x5768b525UL,
0x206f85b3UL, 0xb966d409UL, 0xce61e49fUL, 0x5edef90eUL, 0x29d9c998UL, 0xb0d09822UL, 0xc7d7a8b4UL, 0x59b33d17UL, 0x2eb40d81UL,
0xb7bd5c3bUL, 0xc0ba6cadUL, 0xedb88320UL, 0x9abfb3b6UL, 0x03b6e20cUL, 0x74b1d29aUL, 0xead54739UL, 0x9dd277afUL, 0x04db2615UL,
0x73dc1683UL, 0xe3630b12UL, 0x94643b84UL, 0x0d6d6a3eUL, 0x7a6a5aa8UL, 0xe40ecf0bUL, 0x9309ff9dUL, 0x0a00ae27UL, 0x7d079eb1UL,
0xf00f9344UL, 0x8708a3d2UL, 0x1e01f268UL, 0x6906c2feUL, 0xf762575dUL, 0x806567cbUL, 0x196c3671UL, 0x6e6b06e7UL, 0xfed41b76UL,
0x89d32be0UL, 0x10da7a5aUL, 0x67dd4accUL, 0xf9b9df6fUL, 0x8ebeeff9UL, 0x17b7be43UL, 0x60b08ed5UL, 0xd6d6a3e8UL, 0xa1d1937eUL,
0x38d8c2c4UL, 0x4fdff252UL, 0xd1bb67f1UL, 0xa6bc5767UL, 0x3fb506ddUL, 0x48b2364bUL, 0xd80d2bdaUL, 0xaf0a1b4cUL, 0x36034af6UL,
0x41047a60UL, 0xdf60efc3UL, 0xa867df55UL, 0x316e8eefUL, 0x4669be79UL, 0xcb61b38cUL, 0xbc66831aUL, 0x256fd2a0UL, 0x5268e236UL,
0xcc0c7795UL, 0xbb0b4703UL, 0x220216b9UL, 0x5505262fUL, 0xc5ba3bbeUL, 0xb2bd0b28UL, 0x2bb45a92UL, 0x5cb36a04UL, 0xc2d7ffa7UL,
0xb5d0cf31UL, 0x2cd99e8bUL, 0x5bdeae1dUL, 0x9b64c2b0UL, 0xec63f226UL, 0x756aa39cUL, 0x026d930aUL, 0x9c0906a9UL, 0xeb0e363fUL,
0x72076785UL, 0x05005713UL, 0x95bf4a82UL, 0xe2b87a14UL, 0x7bb12baeUL, 0x0cb61b38UL, 0x92d28e9bUL, 0xe5d5be0dUL, 0x7cdcefb7UL,
0x0bdbdf21UL, 0x86d3d2d4UL, 0xf1d4e242UL, 0x68ddb3f8UL, 0x1fda836eUL, 0x81be16cdUL, 0xf6b9265bUL, 0x6fb077e1UL, 0x18b74777UL,
0x88085ae6UL, 0xff0f6a70UL, 0x66063bcaUL, 0x11010b5cUL, 0x8f659effUL, 0xf862ae69UL, 0x616bffd3UL, 0x166ccf45UL, 0xa00ae278UL,
0xd70dd2eeUL, 0x4e048354UL, 0x3903b3c2UL, 0xa7672661UL, 0xd06016f7UL, 0x4969474dUL, 0x3e6e77dbUL, 0xaed16a4aUL, 0xd9d65adcUL,
0x40df0b66UL, 0x37d83bf0UL, 0xa9bcae53UL, 0xdebb9ec5UL, 0x47b2cf7fUL, 0x30b5ffe9UL, 0xbdbdf21cUL, 0xcabac28aUL, 0x53b39330UL,
0x24b4a3a6UL, 0xbad03605UL, 0xcdd70693UL, 0x54de5729UL, 0x23d967bfUL, 0xb3667a2eUL, 0xc4614ab8UL, 0x5d681b02UL, 0x2a6f2b94UL,
0xb40bbe37UL, 0xc30c8ea1UL, 0x5a05df1bUL, 0x2d02ef8dUL
#ifdef BYFOUR
},
{0x00000000UL, 0x191b3141UL, 0x32366282UL, 0x2b2d53c3UL, 0x646cc504UL, 0x7d77f445UL, 0x565aa786UL, 0x4f4196c7UL, 0xc8d98a08UL,
0xd1c2bb49UL, 0xfaefe88aUL, 0xe3f4d9cbUL, 0xacb54f0cUL, 0xb5ae7e4dUL, 0x9e832d8eUL, 0x87981ccfUL, 0x4ac21251UL, 0x53d92310UL,
0x78f470d3UL, 0x61ef4192UL, 0x2eaed755UL, 0x37b5e614UL, 0x1c98b5d7UL, 0x05838496UL, 0x821b9859UL, 0x9b00a918UL, 0xb02dfadbUL,
0xa936cb9aUL, 0xe6775d5dUL, 0xff6c6c1cUL, 0xd4413fdfUL, 0xcd5a0e9eUL, 0x958424a2UL, 0x8c9f15e3UL, 0xa7b24620UL, 0xbea97761UL,
0xf1e8e1a6UL, 0xe8f3d0e7UL, 0xc3de8324UL, 0xdac5b265UL, 0x5d5daeaaUL, 0x44469febUL, 0x6f6bcc28UL, 0x7670fd69UL, 0x39316baeUL,
0x202a5aefUL, 0x0b07092cUL, 0x121c386dUL, 0xdf4636f3UL, 0xc65d07b2UL, 0xed705471UL, 0xf46b6530UL, 0xbb2af3f7UL, 0xa231c2b6UL,
0x891c9175UL, 0x9007a034UL, 0x179fbcfbUL, 0x0e848dbaUL, 0x25a9de79UL, 0x3cb2ef38UL, 0x73f379ffUL, 0x6ae848beUL, 0x41c51b7dUL,
0x58de2a3cUL, 0xf0794f05UL, 0xe9627e44UL, 0xc24f2d87UL, 0xdb541cc6UL, 0x94158a01UL, 0x8d0ebb40UL, 0xa623e883UL, 0xbf38d9c2UL,
0x38a0c50dUL, 0x21bbf44cUL, 0x0a96a78fUL, 0x138d96ceUL, 0x5ccc0009UL, 0x45d73148UL, 0x6efa628bUL, 0x77e153caUL, 0xbabb5d54UL,
0xa3a06c15UL, 0x888d3fd6UL, 0x91960e97UL, 0xded79850UL, 0xc7cca911UL, 0xece1fad2UL, 0xf5facb93UL, 0x7262d75cUL, 0x6b79e61dUL,
0x4054b5deUL, 0x594f849fUL, 0x160e1258UL, 0x0f152319UL, 0x243870daUL, 0x3d23419bUL, 0x65fd6ba7UL, 0x7ce65ae6UL, 0x57cb0925UL,
0x4ed03864UL, 0x0191aea3UL, 0x188a9fe2UL, 0x33a7cc21UL, 0x2abcfd60UL, 0xad24e1afUL, 0xb43fd0eeUL, 0x9f12832dUL, 0x8609b26cUL,
0xc94824abUL, 0xd05315eaUL, 0xfb7e4629UL, 0xe2657768UL, 0x2f3f79f6UL, 0x362448b7UL, 0x1d091b74UL, 0x04122a35UL, 0x4b53bcf2UL,
0x52488db3UL, 0x7965de70UL, 0x607eef31UL, 0xe7e6f3feUL, 0xfefdc2bfUL, 0xd5d0917cUL, 0xcccba03dUL, 0x838a36faUL, 0x9a9107bbUL,
0xb1bc5478UL, 0xa8a76539UL, 0x3b83984bUL, 0x2298a90aUL, 0x09b5fac9UL, 0x10aecb88UL, 0x5fef5d4fUL, 0x46f46c0eUL, 0x6dd93fcdUL,
0x74c20e8cUL, 0xf35a1243UL, 0xea412302UL, 0xc16c70c1UL, 0xd8774180UL, 0x9736d747UL, 0x8e2de606UL, 0xa500b5c5UL, 0xbc1b8484UL,
0x71418a1aUL, 0x685abb5bUL, 0x4377e898UL, 0x5a6cd9d9UL, 0x152d4f1eUL, 0x0c367e5fUL, 0x271b2d9cUL, 0x3e001cddUL, 0xb9980012UL,
0xa0833153UL, 0x8bae6290UL, 0x92b553d1UL, 0xddf4c516UL, 0xc4eff457UL, 0xefc2a794UL, 0xf6d996d5UL, 0xae07bce9UL, 0xb71c8da8UL,
0x9c31de6bUL, 0x852aef2aUL, 0xca6b79edUL, 0xd37048acUL, 0xf85d1b6fUL, 0xe1462a2eUL, 0x66de36e1UL, 0x7fc507a0UL, 0x54e85463UL,
0x4df36522UL, 0x02b2f3e5UL, 0x1ba9c2a4UL, 0x30849167UL, 0x299fa026UL, 0xe4c5aeb8UL, 0xfdde9ff9UL, 0xd6f3cc3aUL, 0xcfe8fd7bUL,
0x80a96bbcUL, 0x99b25afdUL, 0xb29f093eUL, 0xab84387fUL, 0x2c1c24b0UL, 0x350715f1UL, 0x1e2a4632UL, 0x07317773UL, 0x4870e1b4UL,
0x516bd0f5UL, 0x7a468336UL, 0x635db277UL, 0xcbfad74eUL, 0xd2e1e60fUL, 0xf9ccb5ccUL, 0xe0d7848dUL, 0xaf96124aUL, 0xb68d230bUL,
0x9da070c8UL, 0x84bb4189UL, 0x03235d46UL, 0x1a386c07UL, 0x31153fc4UL, 0x280e0e85UL, 0x674f9842UL, 0x7e54a903UL, 0x5579fac0UL,
0x4c62cb81UL, 0x8138c51fUL, 0x9823f45eUL, 0xb30ea79dUL, 0xaa1596dcUL, 0xe554001bUL, 0xfc4f315aUL, 0xd7626299UL, 0xce7953d8UL,
0x49e14f17UL, 0x50fa7e56UL, 0x7bd72d95UL, 0x62cc1cd4UL, 0x2d8d8a13UL, 0x3496bb52UL, 0x1fbbe891UL, 0x06a0d9d0UL, 0x5e7ef3ecUL,
0x4765c2adUL, 0x6c48916eUL, 0x7553a02fUL, 0x3a1236e8UL, 0x230907a9UL, 0x0824546aUL, 0x113f652bUL, 0x96a779e4UL, 0x8fbc48a5UL,
0xa4911b66UL, 0xbd8a2a27UL, 0xf2cbbce0UL, 0xebd08da1UL, 0xc0fdde62UL, 0xd9e6ef23UL, 0x14bce1bdUL, 0x0da7d0fcUL, 0x268a833fUL,
0x3f91b27eUL, 0x70d024b9UL, 0x69cb15f8UL, 0x42e6463bUL, 0x5bfd777aUL, 0xdc656bb5UL, 0xc57e5af4UL, 0xee530937UL, 0xf7483876UL,
0xb809aeb1UL, 0xa1129ff0UL, 0x8a3fcc33UL, 0x9324fd72UL},
{0x00000000UL, 0x01c26a37UL, 0x0384d46eUL, 0x0246be59UL, 0x0709a8dcUL, 0x06cbc2ebUL, 0x048d7cb2UL, 0x054f1685UL, 0x0e1351b8UL,
0x0fd13b8fUL, 0x0d9785d6UL, 0x0c55efe1UL, 0x091af964UL, 0x08d89353UL, 0x0a9e2d0aUL, 0x0b5c473dUL, 0x1c26a370UL, 0x1de4c947UL,
0x1fa2771eUL, 0x1e601d29UL, 0x1b2f0bacUL, 0x1aed619bUL, 0x18abdfc2UL, 0x1969b5f5UL, 0x1235f2c8UL, 0x13f798ffUL, 0x11b126a6UL,
0x10734c91UL, 0x153c5a14UL, 0x14fe3023UL, 0x16b88e7aUL, 0x177ae44dUL, 0x384d46e0UL, 0x398f2cd7UL, 0x3bc9928eUL, 0x3a0bf8b9UL,
0x3f44ee3cUL, 0x3e86840bUL, 0x3cc03a52UL, 0x3d025065UL, 0x365e1758UL, 0x379c7d6fUL, 0x35dac336UL, 0x3418a901UL, 0x3157bf84UL,
0x3095d5b3UL, 0x32d36beaUL, 0x331101ddUL, 0x246be590UL, 0x25a98fa7UL, 0x27ef31feUL, 0x262d5bc9UL, 0x23624d4cUL, 0x22a0277bUL,
0x20e69922UL, 0x2124f315UL, 0x2a78b428UL, 0x2bbade1fUL, 0x29fc6046UL, 0x283e0a71UL, 0x2d711cf4UL, 0x2cb376c3UL, 0x2ef5c89aUL,
0x2f37a2adUL, 0x709a8dc0UL, 0x7158e7f7UL, 0x731e59aeUL, 0x72dc3399UL, 0x7793251cUL, 0x76514f2bUL, 0x7417f172UL, 0x75d59b45UL,
0x7e89dc78UL, 0x7f4bb64fUL, 0x7d0d0816UL, 0x7ccf6221UL, 0x798074a4UL, 0x78421e93UL, 0x7a04a0caUL, 0x7bc6cafdUL, 0x6cbc2eb0UL,
0x6d7e4487UL, 0x6f38fadeUL, 0x6efa90e9UL, 0x6bb5866cUL, 0x6a77ec5bUL, 0x68315202UL, 0x69f33835UL, 0x62af7f08UL, 0x636d153fUL,
0x612bab66UL, 0x60e9c151UL, 0x65a6d7d4UL, 0x6464bde3UL, 0x662203baUL, 0x67e0698dUL, 0x48d7cb20UL, 0x4915a117UL, 0x4b531f4eUL,
0x4a917579UL, 0x4fde63fcUL, 0x4e1c09cbUL, 0x4c5ab792UL, 0x4d98dda5UL, 0x46c49a98UL, 0x4706f0afUL, 0x45404ef6UL, 0x448224c1UL,
0x41cd3244UL, 0x400f5873UL, 0x4249e62aUL, 0x438b8c1dUL, 0x54f16850UL, 0x55330267UL, 0x5775bc3eUL, 0x56b7d609UL, 0x53f8c08cUL,
0x523aaabbUL, 0x507c14e2UL, 0x51be7ed5UL, 0x5ae239e8UL, 0x5b2053dfUL, 0x5966ed86UL, 0x58a487b1UL, 0x5deb9134UL, 0x5c29fb03UL,
0x5e6f455aUL, 0x5fad2f6dUL, 0xe1351b80UL, 0xe0f771b7UL, 0xe2b1cfeeUL, 0xe373a5d9UL, 0xe63cb35cUL, 0xe7fed96bUL, 0xe5b86732UL,
0xe47a0d05UL, 0xef264a38UL, 0xeee4200fUL, 0xeca29e56UL, 0xed60f461UL, 0xe82fe2e4UL, 0xe9ed88d3UL, 0xebab368aUL, 0xea695cbdUL,
0xfd13b8f0UL, 0xfcd1d2c7UL, 0xfe976c9eUL, 0xff5506a9UL, 0xfa1a102cUL, 0xfbd87a1bUL, 0xf99ec442UL, 0xf85cae75UL, 0xf300e948UL,
0xf2c2837fUL, 0xf0843d26UL, 0xf1465711UL, 0xf4094194UL, 0xf5cb2ba3UL, 0xf78d95faUL, 0xf64fffcdUL, 0xd9785d60UL, 0xd8ba3757UL,
0xdafc890eUL, 0xdb3ee339UL, 0xde71f5bcUL, 0xdfb39f8bUL, 0xddf521d2UL, 0xdc374be5UL, 0xd76b0cd8UL, 0xd6a966efUL, 0xd4efd8b6UL,
0xd52db281UL, 0xd062a404UL, 0xd1a0ce33UL, 0xd3e6706aUL, 0xd2241a5dUL, 0xc55efe10UL, 0xc49c9427UL, 0xc6da2a7eUL, 0xc7184049UL,
0xc25756ccUL, 0xc3953cfbUL, 0xc1d382a2UL, 0xc011e895UL, 0xcb4dafa8UL, 0xca8fc59fUL, 0xc8c97bc6UL, 0xc90b11f1UL, 0xcc440774UL,
0xcd866d43UL, 0xcfc0d31aUL, 0xce02b92dUL, 0x91af9640UL, 0x906dfc77UL, 0x922b422eUL, 0x93e92819UL, 0x96a63e9cUL, 0x976454abUL,
0x9522eaf2UL, 0x94e080c5UL, 0x9fbcc7f8UL, 0x9e7eadcfUL, 0x9c381396UL, 0x9dfa79a1UL, 0x98b56f24UL, 0x99770513UL, 0x9b31bb4aUL,
0x9af3d17dUL, 0x8d893530UL, 0x8c4b5f07UL, 0x8e0de15eUL, 0x8fcf8b69UL, 0x8a809decUL, 0x8b42f7dbUL, 0x89044982UL, 0x88c623b5UL,
0x839a6488UL, 0x82580ebfUL, 0x801eb0e6UL, 0x81dcdad1UL, 0x8493cc54UL, 0x8551a663UL, 0x8717183aUL, 0x86d5720dUL, 0xa9e2d0a0UL,
0xa820ba97UL, 0xaa6604ceUL, 0xaba46ef9UL, 0xaeeb787cUL, 0xaf29124bUL, 0xad6fac12UL, 0xacadc625UL, 0xa7f18118UL, 0xa633eb2fUL,
0xa4755576UL, 0xa5b73f41UL, 0xa0f829c4UL, 0xa13a43f3UL, 0xa37cfdaaUL, 0xa2be979dUL, 0xb5c473d0UL, 0xb40619e7UL, 0xb640a7beUL,
0xb782cd89UL, 0xb2cddb0cUL, 0xb30fb13bUL, 0xb1490f62UL, 0xb08b6555UL, 0xbbd72268UL, 0xba15485fUL, 0xb853f606UL, 0xb9919c31UL,
0xbcde8ab4UL, 0xbd1ce083UL, 0xbf5a5edaUL, 0xbe9834edUL},
{0x00000000UL, 0xb8bc6765UL, 0xaa09c88bUL, 0x12b5afeeUL, 0x8f629757UL, 0x37def032UL, 0x256b5fdcUL, 0x9dd738b9UL, 0xc5b428efUL,
0x7d084f8aUL, 0x6fbde064UL, 0xd7018701UL, 0x4ad6bfb8UL, 0xf26ad8ddUL, 0xe0df7733UL, 0x58631056UL, 0x5019579fUL, 0xe8a530faUL,
0xfa109f14UL, 0x42acf871UL, 0xdf7bc0c8UL, 0x67c7a7adUL, 0x75720843UL, 0xcdce6f26UL, 0x95ad7f70UL, 0x2d111815UL, 0x3fa4b7fbUL,
0x8718d09eUL, 0x1acfe827UL, 0xa2738f42UL, 0xb0c620acUL, 0x087a47c9UL, 0xa032af3eUL, 0x188ec85bUL, 0x0a3b67b5UL, 0xb28700d0UL,
0x2f503869UL, 0x97ec5f0cUL, 0x8559f0e2UL, 0x3de59787UL, 0x658687d1UL, 0xdd3ae0b4UL, 0xcf8f4f5aUL, 0x7733283fUL, 0xeae41086UL,
0x525877e3UL, 0x40edd80dUL, 0xf851bf68UL, 0xf02bf8a1UL, 0x48979fc4UL, 0x5a22302aUL, 0xe29e574fUL, 0x7f496ff6UL, 0xc7f50893UL,
0xd540a77dUL, 0x6dfcc018UL, 0x359fd04eUL, 0x8d23b72bUL, 0x9f9618c5UL, 0x272a7fa0UL, 0xbafd4719UL, 0x0241207cUL, 0x10f48f92UL,
0xa848e8f7UL, 0x9b14583dUL, 0x23a83f58UL, 0x311d90b6UL, 0x89a1f7d3UL, 0x1476cf6aUL, 0xaccaa80fUL, 0xbe7f07e1UL, 0x06c36084UL,
0x5ea070d2UL, 0xe61c17b7UL, 0xf4a9b859UL, 0x4c15df3cUL, 0xd1c2e785UL, 0x697e80e0UL, 0x7bcb2f0eUL, 0xc377486bUL, 0xcb0d0fa2UL,
0x73b168c7UL, 0x6104c729UL, 0xd9b8a04cUL, 0x446f98f5UL, 0xfcd3ff90UL, 0xee66507eUL, 0x56da371bUL, 0x0eb9274dUL, 0xb6054028UL,
0xa4b0efc6UL, 0x1c0c88a3UL, 0x81dbb01aUL, 0x3967d77fUL, 0x2bd27891UL, 0x936e1ff4UL, 0x3b26f703UL, 0x839a9066UL, 0x912f3f88UL,
0x299358edUL, 0xb4446054UL, 0x0cf80731UL, 0x1e4da8dfUL, 0xa6f1cfbaUL, 0xfe92dfecUL, 0x462eb889UL, 0x549b1767UL, 0xec277002UL,
0x71f048bbUL, 0xc94c2fdeUL, 0xdbf98030UL, 0x6345e755UL, 0x6b3fa09cUL, 0xd383c7f9UL, 0xc1366817UL, 0x798a0f72UL, 0xe45d37cbUL,
0x5ce150aeUL, 0x4e54ff40UL, 0xf6e89825UL, 0xae8b8873UL, 0x1637ef16UL, 0x048240f8UL, 0xbc3e279dUL, 0x21e91f24UL, 0x99557841UL,
0x8be0d7afUL, 0x335cb0caUL, 0xed59b63bUL, 0x55e5d15eUL, 0x47507eb0UL, 0xffec19d5UL, 0x623b216cUL, 0xda874609UL, 0xc832e9e7UL,
0x708e8e82UL, 0x28ed9ed4UL, 0x9051f9b1UL, 0x82e4565fUL, 0x3a58313aUL, 0xa78f0983UL, 0x1f336ee6UL, 0x0d86c108UL, 0xb53aa66dUL,
0xbd40e1a4UL, 0x05fc86c1UL, 0x1749292fUL, 0xaff54e4aUL, 0x322276f3UL, 0x8a9e1196UL, 0x982bbe78UL, 0x2097d91dUL, 0x78f4c94bUL,
0xc048ae2eUL, 0xd2fd01c0UL, 0x6a4166a5UL, 0xf7965e1cUL, 0x4f2a3979UL, 0x5d9f9697UL, 0xe523f1f2UL, 0x4d6b1905UL, 0xf5d77e60UL,
0xe762d18eUL, 0x5fdeb6ebUL, 0xc2098e52UL, 0x7ab5e937UL, 0x680046d9UL, 0xd0bc21bcUL, 0x88df31eaUL, 0x3063568fUL, 0x22d6f961UL,
0x9a6a9e04UL, 0x07bda6bdUL, 0xbf01c1d8UL, 0xadb46e36UL, 0x15080953UL, 0x1d724e9aUL, 0xa5ce29ffUL, 0xb77b8611UL, 0x0fc7e174UL,
0x9210d9cdUL, 0x2aacbea8UL, 0x38191146UL, 0x80a57623UL, 0xd8c66675UL, 0x607a0110UL, 0x72cfaefeUL, 0xca73c99bUL, 0x57a4f122UL,
0xef189647UL, 0xfdad39a9UL, 0x45115eccUL, 0x764dee06UL, 0xcef18963UL, 0xdc44268dUL, 0x64f841e8UL, 0xf92f7951UL, 0x41931e34UL,
0x5326b1daUL, 0xeb9ad6bfUL, 0xb3f9c6e9UL, 0x0b45a18cUL, 0x19f00e62UL, 0xa14c6907UL, 0x3c9b51beUL, 0x842736dbUL, 0x96929935UL,
0x2e2efe50UL, 0x2654b999UL, 0x9ee8defcUL, 0x8c5d7112UL, 0x34e11677UL, 0xa9362eceUL, 0x118a49abUL, 0x033fe645UL, 0xbb838120UL,
0xe3e09176UL, 0x5b5cf613UL, 0x49e959fdUL, 0xf1553e98UL, 0x6c820621UL, 0xd43e6144UL, 0xc68bceaaUL, 0x7e37a9cfUL, 0xd67f4138UL,
0x6ec3265dUL, 0x7c7689b3UL, 0xc4caeed6UL, 0x591dd66fUL, 0xe1a1b10aUL, 0xf3141ee4UL, 0x4ba87981UL, 0x13cb69d7UL, 0xab770eb2UL,
0xb9c2a15cUL, 0x017ec639UL, 0x9ca9fe80UL, 0x241599e5UL, 0x36a0360bUL, 0x8e1c516eUL, 0x866616a7UL, 0x3eda71c2UL, 0x2c6fde2cUL,
0x94d3b949UL, 0x090481f0UL, 0xb1b8e695UL, 0xa30d497bUL, 0x1bb12e1eUL, 0x43d23e48UL, 0xfb6e592dUL, 0xe9dbf6c3UL, 0x516791a6UL,
0xccb0a91fUL, 0x740cce7aUL, 0x66b96194UL, 0xde0506f1UL},
{0x00000000UL, 0x96300777UL, 0x2c610eeeUL, 0xba510999UL, 0x19c46d07UL, 0x8ff46a70UL, 0x35a563e9UL, 0xa395649eUL, 0x3288db0eUL,
0xa4b8dc79UL, 0x1ee9d5e0UL, 0x88d9d297UL, 0x2b4cb609UL, 0xbd7cb17eUL, 0x072db8e7UL, 0x911dbf90UL, 0x6410b71dUL, 0xf220b06aUL,
0x4871b9f3UL, 0xde41be84UL, 0x7dd4da1aUL, 0xebe4dd6dUL, 0x51b5d4f4UL, 0xc785d383UL, 0x56986c13UL, 0xc0a86b64UL, 0x7af962fdUL,
0xecc9658aUL, 0x4f5c0114UL, 0xd96c0663UL, 0x633d0ffaUL, 0xf50d088dUL, 0xc8206e3bUL, 0x5e10694cUL, 0xe44160d5UL, 0x727167a2UL,
0xd1e4033cUL, 0x47d4044bUL, 0xfd850dd2UL, 0x6bb50aa5UL, 0xfaa8b535UL, 0x6c98b242UL, 0xd6c9bbdbUL, 0x40f9bcacUL, 0xe36cd832UL,
0x755cdf45UL, 0xcf0dd6dcUL, 0x593dd1abUL, 0xac30d926UL, 0x3a00de51UL, 0x8051d7c8UL, 0x1661d0bfUL, 0xb5f4b421UL, 0x23c4b356UL,
0x9995bacfUL, 0x0fa5bdb8UL, 0x9eb80228UL, 0x0888055fUL, 0xb2d90cc6UL, 0x24e90bb1UL, 0x877c6f2fUL, 0x114c6858UL, 0xab1d61c1UL,
0x3d2d66b6UL, 0x9041dc76UL, 0x0671db01UL, 0xbc20d298UL, 0x2a10d5efUL, 0x8985b171UL, 0x1fb5b606UL, 0xa5e4bf9fUL, 0x33d4b8e8UL,
0xa2c90778UL, 0x34f9000fUL, 0x8ea80996UL, 0x18980ee1UL, 0xbb0d6a7fUL, 0x2d3d6d08UL, 0x976c6491UL, 0x015c63e6UL, 0xf4516b6bUL,
0x62616c1cUL, 0xd8306585UL, 0x4e0062f2UL, 0xed95066cUL, 0x7ba5011bUL, 0xc1f40882UL, 0x57c40ff5UL, 0xc6d9b065UL, 0x50e9b712UL,
0xeab8be8bUL, 0x7c88b9fcUL, 0xdf1ddd62UL, 0x492dda15UL, 0xf37cd38cUL, 0x654cd4fbUL, 0x5861b24dUL, 0xce51b53aUL, 0x7400bca3UL,
0xe230bbd4UL, 0x41a5df4aUL, 0xd795d83dUL, 0x6dc4d1a4UL, 0xfbf4d6d3UL, 0x6ae96943UL, 0xfcd96e34UL, 0x468867adUL, 0xd0b860daUL,
0x732d0444UL, 0xe51d0333UL, 0x5f4c0aaaUL, 0xc97c0dddUL, 0x3c710550UL, 0xaa410227UL, 0x10100bbeUL, 0x86200cc9UL, 0x25b56857UL,
0xb3856f20UL, 0x09d466b9UL, 0x9fe461ceUL, 0x0ef9de5eUL, 0x98c9d929UL, 0x2298d0b0UL, 0xb4a8d7c7UL, 0x173db359UL, 0x810db42eUL,
0x3b5cbdb7UL, 0xad6cbac0UL, 0x2083b8edUL, 0xb6b3bf9aUL, 0x0ce2b603UL, 0x9ad2b174UL, 0x3947d5eaUL, 0xaf77d29dUL, 0x1526db04UL,
0x8316dc73UL, 0x120b63e3UL, 0x843b6494UL, 0x3e6a6d0dUL, 0xa85a6a7aUL, 0x0bcf0ee4UL, 0x9dff0993UL, 0x27ae000aUL, 0xb19e077dUL,
0x44930ff0UL, 0xd2a30887UL, 0x68f2011eUL, 0xfec20669UL, 0x5d5762f7UL, 0xcb676580UL, 0x71366c19UL, 0xe7066b6eUL, 0x761bd4feUL,
0xe02bd389UL, 0x5a7ada10UL, 0xcc4add67UL, 0x6fdfb9f9UL, 0xf9efbe8eUL, 0x43beb717UL, 0xd58eb060UL, 0xe8a3d6d6UL, 0x7e93d1a1UL,
0xc4c2d838UL, 0x52f2df4fUL, 0xf167bbd1UL, 0x6757bca6UL, 0xdd06b53fUL, 0x4b36b248UL, 0xda2b0dd8UL, 0x4c1b0aafUL, 0xf64a0336UL,
0x607a0441UL, 0xc3ef60dfUL, 0x55df67a8UL, 0xef8e6e31UL, 0x79be6946UL, 0x8cb361cbUL, 0x1a8366bcUL, 0xa0d26f25UL, 0x36e26852UL,
0x95770cccUL, 0x03470bbbUL, 0xb9160222UL, 0x2f260555UL, 0xbe3bbac5UL, 0x280bbdb2UL, 0x925ab42bUL, 0x046ab35cUL, 0xa7ffd7c2UL,
0x31cfd0b5UL, 0x8b9ed92cUL, 0x1daede5bUL, 0xb0c2649bUL, 0x26f263ecUL, 0x9ca36a75UL, 0x0a936d02UL, 0xa906099cUL, 0x3f360eebUL,
0x85670772UL, 0x13570005UL, 0x824abf95UL, 0x147ab8e2UL, 0xae2bb17bUL, 0x381bb60cUL, 0x9b8ed292UL, 0x0dbed5e5UL, 0xb7efdc7cUL,
0x21dfdb0bUL, 0xd4d2d386UL, 0x42e2d4f1UL, 0xf8b3dd68UL, 0x6e83da1fUL, 0xcd16be81UL, 0x5b26b9f6UL, 0xe177b06fUL, 0x7747b718UL,
0xe65a0888UL, 0x706a0fffUL, 0xca3b0666UL, 0x5c0b0111UL, 0xff9e658fUL, 0x69ae62f8UL, 0xd3ff6b61UL, 0x45cf6c16UL, 0x78e20aa0UL,
0xeed20dd7UL, 0x5483044eUL, 0xc2b30339UL, 0x612667a7UL, 0xf71660d0UL, 0x4d476949UL, 0xdb776e3eUL, 0x4a6ad1aeUL, 0xdc5ad6d9UL,
0x660bdf40UL, 0xf03bd837UL, 0x53aebca9UL, 0xc59ebbdeUL, 0x7fcfb247UL, 0xe9ffb530UL, 0x1cf2bdbdUL, 0x8ac2bacaUL, 0x3093b353UL,
0xa6a3b424UL, 0x0536d0baUL, 0x9306d7cdUL, 0x2957de54UL, 0xbf67d923UL, 0x2e7a66b3UL, 0xb84a61c4UL, 0x021b685dUL, 0x942b6f2aUL,
0x37be0bb4UL, 0xa18e0cc3UL, 0x1bdf055aUL, 0x8def022dUL},
{0x00000000UL, 0x41311b19UL, 0x82623632UL, 0xc3532d2bUL, 0x04c56c64UL, 0x45f4777dUL, 0x86a75a56UL, 0xc796414fUL, 0x088ad9c8UL,
0x49bbc2d1UL, 0x8ae8effaUL, 0xcbd9f4e3UL, 0x0c4fb5acUL, 0x4d7eaeb5UL, 0x8e2d839eUL, 0xcf1c9887UL, 0x5112c24aUL, 0x1023d953UL,
0xd370f478UL, 0x9241ef61UL, 0x55d7ae2eUL, 0x14e6b537UL, 0xd7b5981cUL, 0x96848305UL, 0x59981b82UL, 0x18a9009bUL, 0xdbfa2db0UL,
0x9acb36a9UL, 0x5d5d77e6UL, 0x1c6c6cffUL, 0xdf3f41d4UL, 0x9e0e5acdUL, 0xa2248495UL, 0xe3159f8cUL, 0x2046b2a7UL, 0x6177a9beUL,
0xa6e1e8f1UL, 0xe7d0f3e8UL, 0x2483dec3UL, 0x65b2c5daUL, 0xaaae5d5dUL, 0xeb9f4644UL, 0x28cc6b6fUL, 0x69fd7076UL, 0xae6b3139UL,
0xef5a2a20UL, 0x2c09070bUL, 0x6d381c12UL, 0xf33646dfUL, 0xb2075dc6UL, 0x715470edUL, 0x30656bf4UL, 0xf7f32abbUL, 0xb6c231a2UL,
0x75911c89UL, 0x34a00790UL, 0xfbbc9f17UL, 0xba8d840eUL, 0x79dea925UL, 0x38efb23cUL, 0xff79f373UL, 0xbe48e86aUL, 0x7d1bc541UL,
0x3c2ade58UL, 0x054f79f0UL, 0x447e62e9UL, 0x872d4fc2UL, 0xc61c54dbUL, 0x018a1594UL, 0x40bb0e8dUL, 0x83e823a6UL, 0xc2d938bfUL,
0x0dc5a038UL, 0x4cf4bb21UL, 0x8fa7960aUL, 0xce968d13UL, 0x0900cc5cUL, 0x4831d745UL, 0x8b62fa6eUL, 0xca53e177UL, 0x545dbbbaUL,
0x156ca0a3UL, 0xd63f8d88UL, 0x970e9691UL, 0x5098d7deUL, 0x11a9ccc7UL, 0xd2fae1ecUL, 0x93cbfaf5UL, 0x5cd76272UL, 0x1de6796bUL,
0xdeb55440UL, 0x9f844f59UL, 0x58120e16UL, 0x1923150fUL, 0xda703824UL, 0x9b41233dUL, 0xa76bfd65UL, 0xe65ae67cUL, 0x2509cb57UL,
0x6438d04eUL, 0xa3ae9101UL, 0xe29f8a18UL, 0x21cca733UL, 0x60fdbc2aUL, 0xafe124adUL, 0xeed03fb4UL, 0x2d83129fUL, 0x6cb20986UL,
0xab2448c9UL, 0xea1553d0UL, 0x29467efbUL, 0x687765e2UL, 0xf6793f2fUL, 0xb7482436UL, 0x741b091dUL, 0x352a1204UL, 0xf2bc534bUL,
0xb38d4852UL, 0x70de6579UL, 0x31ef7e60UL, 0xfef3e6e7UL, 0xbfc2fdfeUL, 0x7c91d0d5UL, 0x3da0cbccUL, 0xfa368a83UL, 0xbb07919aUL,
0x7854bcb1UL, 0x3965a7a8UL, 0x4b98833bUL, 0x0aa99822UL, 0xc9fab509UL, 0x88cbae10UL, 0x4f5def5fUL, 0x0e6cf446UL, 0xcd3fd96dUL,
0x8c0ec274UL, 0x43125af3UL, 0x022341eaUL, 0xc1706cc1UL, 0x804177d8UL, 0x47d73697UL, 0x06e62d8eUL, 0xc5b500a5UL, 0x84841bbcUL,
0x1a8a4171UL, 0x5bbb5a68UL, 0x98e87743UL, 0xd9d96c5aUL, 0x1e4f2d15UL, 0x5f7e360cUL, 0x9c2d1b27UL, 0xdd1c003eUL, 0x120098b9UL,
0x533183a0UL, 0x9062ae8bUL, 0xd153b592UL, 0x16c5f4ddUL, 0x57f4efc4UL, 0x94a7c2efUL, 0xd596d9f6UL, 0xe9bc07aeUL, 0xa88d1cb7UL,
0x6bde319cUL, 0x2aef2a85UL, 0xed796bcaUL, 0xac4870d3UL, 0x6f1b5df8UL, 0x2e2a46e1UL, 0xe136de66UL, 0xa007c57fUL, 0x6354e854UL,
0x2265f34dUL, 0xe5f3b202UL, 0xa4c2a91bUL, 0x67918430UL, 0x26a09f29UL, 0xb8aec5e4UL, 0xf99fdefdUL, 0x3accf3d6UL, 0x7bfde8cfUL,
0xbc6ba980UL, 0xfd5ab299UL, 0x3e099fb2UL, 0x7f3884abUL, 0xb0241c2cUL, 0xf1150735UL, 0x32462a1eUL, 0x73773107UL, 0xb4e17048UL,
0xf5d06b51UL, 0x3683467aUL, 0x77b25d63UL, 0x4ed7facbUL, 0x0fe6e1d2UL, 0xccb5ccf9UL, 0x8d84d7e0UL, 0x4a1296afUL, 0x0b238db6UL,
0xc870a09dUL, 0x8941bb84UL, 0x465d2303UL, 0x076c381aUL, 0xc43f1531UL, 0x850e0e28UL, 0x42984f67UL, 0x03a9547eUL, 0xc0fa7955UL,
0x81cb624cUL, 0x1fc53881UL, 0x5ef42398UL, 0x9da70eb3UL, 0xdc9615aaUL, 0x1b0054e5UL, 0x5a314ffcUL, 0x996262d7UL, 0xd85379ceUL,
0x174fe149UL, 0x567efa50UL, 0x952dd77bUL, 0xd41ccc62UL, 0x138a8d2dUL, 0x52bb9634UL, 0x91e8bb1fUL, 0xd0d9a006UL, 0xecf37e5eUL,
0xadc26547UL, 0x6e91486cUL, 0x2fa05375UL, 0xe836123aUL, 0xa9070923UL, 0x6a542408UL, 0x2b653f11UL, 0xe479a796UL, 0xa548bc8fUL,
0x661b91a4UL, 0x272a8abdUL, 0xe0bccbf2UL, 0xa18dd0ebUL, 0x62defdc0UL, 0x23efe6d9UL, 0xbde1bc14UL, 0xfcd0a70dUL, 0x3f838a26UL,
0x7eb2913fUL, 0xb924d070UL, 0xf815cb69UL, 0x3b46e642UL, 0x7a77fd5bUL, 0xb56b65dcUL, 0xf45a7ec5UL, 0x370953eeUL, 0x763848f7UL,
0xb1ae09b8UL, 0xf09f12a1UL, 0x33cc3f8aUL, 0x72fd2493UL},
{0x00000000UL, 0x376ac201UL, 0x6ed48403UL, 0x59be4602UL, 0xdca80907UL, 0xebc2cb06UL, 0xb27c8d04UL, 0x85164f05UL, 0xb851130eUL,
0x8f3bd10fUL, 0xd685970dUL, 0xe1ef550cUL, 0x64f91a09UL, 0x5393d808UL, 0x0a2d9e0aUL, 0x3d475c0bUL, 0x70a3261cUL, 0x47c9e41dUL,
0x1e77a21fUL, 0x291d601eUL, 0xac0b2f1bUL, 0x9b61ed1aUL, 0xc2dfab18UL, 0xf5b56919UL, 0xc8f23512UL, 0xff98f713UL, 0xa626b111UL,
0x914c7310UL, 0x145a3c15UL, 0x2330fe14UL, 0x7a8eb816UL, 0x4de47a17UL, 0xe0464d38UL, 0xd72c8f39UL, 0x8e92c93bUL, 0xb9f80b3aUL,
0x3cee443fUL, 0x0b84863eUL, 0x523ac03cUL, 0x6550023dUL, 0x58175e36UL, 0x6f7d9c37UL, 0x36c3da35UL, 0x01a91834UL, 0x84bf5731UL,
0xb3d59530UL, 0xea6bd332UL, 0xdd011133UL, 0x90e56b24UL, 0xa78fa925UL, 0xfe31ef27UL, 0xc95b2d26UL, 0x4c4d6223UL, 0x7b27a022UL,
0x2299e620UL, 0x15f32421UL, 0x28b4782aUL, 0x1fdeba2bUL, 0x4660fc29UL, 0x710a3e28UL, 0xf41c712dUL, 0xc376b32cUL, 0x9ac8f52eUL,
0xada2372fUL, 0xc08d9a70UL, 0xf7e75871UL, 0xae591e73UL, 0x9933dc72UL, 0x1c259377UL, 0x2b4f5176UL, 0x72f11774UL, 0x459bd575UL,
0x78dc897eUL, 0x4fb64b7fUL, 0x16080d7dUL, 0x2162cf7cUL, 0xa4748079UL, 0x931e4278UL, 0xcaa0047aUL, 0xfdcac67bUL, 0xb02ebc6cUL,
0x87447e6dUL, 0xdefa386fUL, 0xe990fa6eUL, 0x6c86b56bUL, 0x5bec776aUL, 0x02523168UL, 0x3538f369UL, 0x087faf62UL, 0x3f156d63UL,
0x66ab2b61UL, 0x51c1e960UL, 0xd4d7a665UL, 0xe3bd6464UL, 0xba032266UL, 0x8d69e067UL, 0x20cbd748UL, 0x17a11549UL, 0x4e1f534bUL,
0x7975914aUL, 0xfc63de4fUL, 0xcb091c4eUL, 0x92b75a4cUL, 0xa5dd984dUL, 0x989ac446UL, 0xaff00647UL, 0xf64e4045UL, 0xc1248244UL,
0x4432cd41UL, 0x73580f40UL, 0x2ae64942UL, 0x1d8c8b43UL, 0x5068f154UL, 0x67023355UL, 0x3ebc7557UL, 0x09d6b756UL, 0x8cc0f853UL,
0xbbaa3a52UL, 0xe2147c50UL, 0xd57ebe51UL, 0xe839e25aUL, 0xdf53205bUL, 0x86ed6659UL, 0xb187a458UL, 0x3491eb5dUL, 0x03fb295cUL,
0x5a456f5eUL, 0x6d2fad5fUL, 0x801b35e1UL, 0xb771f7e0UL, 0xeecfb1e2UL, 0xd9a573e3UL, 0x5cb33ce6UL, 0x6bd9fee7UL, 0x3267b8e5UL,
0x050d7ae4UL, 0x384a26efUL, 0x0f20e4eeUL, 0x569ea2ecUL, 0x61f460edUL, 0xe4e22fe8UL, 0xd388ede9UL, 0x8a36abebUL, 0xbd5c69eaUL,
0xf0b813fdUL, 0xc7d2d1fcUL, 0x9e6c97feUL, 0xa90655ffUL, 0x2c101afaUL, 0x1b7ad8fbUL, 0x42c49ef9UL, 0x75ae5cf8UL, 0x48e900f3UL,
0x7f83c2f2UL, 0x263d84f0UL, 0x115746f1UL, 0x944109f4UL, 0xa32bcbf5UL, 0xfa958df7UL, 0xcdff4ff6UL, 0x605d78d9UL, 0x5737bad8UL,
0x0e89fcdaUL, 0x39e33edbUL, 0xbcf571deUL, 0x8b9fb3dfUL, 0xd221f5ddUL, 0xe54b37dcUL, 0xd80c6bd7UL, 0xef66a9d6UL, 0xb6d8efd4UL,
0x81b22dd5UL, 0x04a462d0UL, 0x33cea0d1UL, 0x6a70e6d3UL, 0x5d1a24d2UL, 0x10fe5ec5UL, 0x27949cc4UL, 0x7e2adac6UL, 0x494018c7UL,
0xcc5657c2UL, 0xfb3c95c3UL, 0xa282d3c1UL, 0x95e811c0UL, 0xa8af4dcbUL, 0x9fc58fcaUL, 0xc67bc9c8UL, 0xf1110bc9UL, 0x740744ccUL,
0x436d86cdUL, 0x1ad3c0cfUL, 0x2db902ceUL, 0x4096af91UL, 0x77fc6d90UL, 0x2e422b92UL, 0x1928e993UL, 0x9c3ea696UL, 0xab546497UL,
0xf2ea2295UL, 0xc580e094UL, 0xf8c7bc9fUL, 0xcfad7e9eUL, 0x9613389cUL, 0xa179fa9dUL, 0x246fb598UL, 0x13057799UL, 0x4abb319bUL,
0x7dd1f39aUL, 0x3035898dUL, 0x075f4b8cUL, 0x5ee10d8eUL, 0x698bcf8fUL, 0xec9d808aUL, 0xdbf7428bUL, 0x82490489UL, 0xb523c688UL,
0x88649a83UL, 0xbf0e5882UL, 0xe6b01e80UL, 0xd1dadc81UL, 0x54cc9384UL, 0x63a65185UL, 0x3a181787UL, 0x0d72d586UL, 0xa0d0e2a9UL,
0x97ba20a8UL, 0xce0466aaUL, 0xf96ea4abUL, 0x7c78ebaeUL, 0x4b1229afUL, 0x12ac6fadUL, 0x25c6adacUL, 0x1881f1a7UL, 0x2feb33a6UL,
0x765575a4UL, 0x413fb7a5UL, 0xc429f8a0UL, 0xf3433aa1UL, 0xaafd7ca3UL, 0x9d97bea2UL, 0xd073c4b5UL, 0xe71906b4UL, 0xbea740b6UL,
0x89cd82b7UL, 0x0cdbcdb2UL, 0x3bb10fb3UL, 0x620f49b1UL, 0x55658bb0UL, 0x6822d7bbUL, 0x5f4815baUL, 0x06f653b8UL, 0x319c91b9UL,
0xb48adebcUL, 0x83e01cbdUL, 0xda5e5abfUL, 0xed3498beUL},
{0x00000000UL, 0x6567bcb8UL, 0x8bc809aaUL, 0xeeafb512UL, 0x5797628fUL, 0x32f0de37UL, 0xdc5f6b25UL, 0xb938d79dUL, 0xef28b4c5UL,
0x8a4f087dUL, 0x64e0bd6fUL, 0x018701d7UL, 0xb8bfd64aUL, 0xddd86af2UL, 0x3377dfe0UL, 0x56106358UL, 0x9f571950UL, 0xfa30a5e8UL,
0x149f10faUL, 0x71f8ac42UL, 0xc8c07bdfUL, 0xada7c767UL, 0x43087275UL, 0x266fcecdUL, 0x707fad95UL, 0x1518112dUL, 0xfbb7a43fUL,
0x9ed01887UL, 0x27e8cf1aUL, 0x428f73a2UL, 0xac20c6b0UL, 0xc9477a08UL, 0x3eaf32a0UL, 0x5bc88e18UL, 0xb5673b0aUL, 0xd00087b2UL,
0x6938502fUL, 0x0c5fec97UL, 0xe2f05985UL, 0x8797e53dUL, 0xd1878665UL, 0xb4e03addUL, 0x5a4f8fcfUL, 0x3f283377UL, 0x8610e4eaUL,
0xe3775852UL, 0x0dd8ed40UL, 0x68bf51f8UL, 0xa1f82bf0UL, 0xc49f9748UL, 0x2a30225aUL, 0x4f579ee2UL, 0xf66f497fUL, 0x9308f5c7UL,
0x7da740d5UL, 0x18c0fc6dUL, 0x4ed09f35UL, 0x2bb7238dUL, 0xc518969fUL, 0xa07f2a27UL, 0x1947fdbaUL, 0x7c204102UL, 0x928ff410UL,
0xf7e848a8UL, 0x3d58149bUL, 0x583fa823UL, 0xb6901d31UL, 0xd3f7a189UL, 0x6acf7614UL, 0x0fa8caacUL, 0xe1077fbeUL, 0x8460c306UL,
0xd270a05eUL, 0xb7171ce6UL, 0x59b8a9f4UL, 0x3cdf154cUL, 0x85e7c2d1UL, 0xe0807e69UL, 0x0e2fcb7bUL, 0x6b4877c3UL, 0xa20f0dcbUL,
0xc768b173UL, 0x29c70461UL, 0x4ca0b8d9UL, 0xf5986f44UL, 0x90ffd3fcUL, 0x7e5066eeUL, 0x1b37da56UL, 0x4d27b90eUL, 0x284005b6UL,
0xc6efb0a4UL, 0xa3880c1cUL, 0x1ab0db81UL, 0x7fd76739UL, 0x9178d22bUL, 0xf41f6e93UL, 0x03f7263bUL, 0x66909a83UL, 0x883f2f91UL,
0xed589329UL, 0x546044b4UL, 0x3107f80cUL, 0xdfa84d1eUL, 0xbacff1a6UL, 0xecdf92feUL, 0x89b82e46UL, 0x67179b54UL, 0x027027ecUL,
0xbb48f071UL, 0xde2f4cc9UL, 0x3080f9dbUL, 0x55e74563UL, 0x9ca03f6bUL, 0xf9c783d3UL, 0x176836c1UL, 0x720f8a79UL, 0xcb375de4UL,
0xae50e15cUL, 0x40ff544eUL, 0x2598e8f6UL, 0x73888baeUL, 0x16ef3716UL, 0xf8408204UL, 0x9d273ebcUL, 0x241fe921UL, 0x41785599UL,
0xafd7e08bUL, 0xcab05c33UL, 0x3bb659edUL, 0x5ed1e555UL, 0xb07e5047UL, 0xd519ecffUL, 0x6c213b62UL, 0x094687daUL, 0xe7e932c8UL,
0x828e8e70UL, 0xd49eed28UL, 0xb1f95190UL, 0x5f56e482UL, 0x3a31583aUL, 0x83098fa7UL, 0xe66e331fUL, 0x08c1860dUL, 0x6da63ab5UL,
0xa4e140bdUL, 0xc186fc05UL, 0x2f294917UL, 0x4a4ef5afUL, 0xf3762232UL, 0x96119e8aUL, 0x78be2b98UL, 0x1dd99720UL, 0x4bc9f478UL,
0x2eae48c0UL, 0xc001fdd2UL, 0xa566416aUL, 0x1c5e96f7UL, 0x79392a4fUL, 0x97969f5dUL, 0xf2f123e5UL, 0x05196b4dUL, 0x607ed7f5UL,
0x8ed162e7UL, 0xebb6de5fUL, 0x528e09c2UL, 0x37e9b57aUL, 0xd9460068UL, 0xbc21bcd0UL, 0xea31df88UL, 0x8f566330UL, 0x61f9d622UL,
0x049e6a9aUL, 0xbda6bd07UL, 0xd8c101bfUL, 0x366eb4adUL, 0x53090815UL, 0x9a4e721dUL, 0xff29cea5UL, 0x11867bb7UL, 0x74e1c70fUL,
0xcdd91092UL, 0xa8beac2aUL, 0x46111938UL, 0x2376a580UL, 0x7566c6d8UL, 0x10017a60UL, 0xfeaecf72UL, 0x9bc973caUL, 0x22f1a457UL,
0x479618efUL, 0xa939adfdUL, 0xcc5e1145UL, 0x06ee4d76UL, 0x6389f1ceUL, 0x8d2644dcUL, 0xe841f864UL, 0x51792ff9UL, 0x341e9341UL,
0xdab12653UL, 0xbfd69aebUL, 0xe9c6f9b3UL, 0x8ca1450bUL, 0x620ef019UL, 0x07694ca1UL, 0xbe519b3cUL, 0xdb362784UL, 0x35999296UL,
0x50fe2e2eUL, 0x99b95426UL, 0xfcdee89eUL, 0x12715d8cUL, 0x7716e134UL, 0xce2e36a9UL, 0xab498a11UL, 0x45e63f03UL, 0x208183bbUL,
0x7691e0e3UL, 0x13f65c5bUL, 0xfd59e949UL, 0x983e55f1UL, 0x2106826cUL, 0x44613ed4UL, 0xaace8bc6UL, 0xcfa9377eUL, 0x38417fd6UL,
0x5d26c36eUL, 0xb389767cUL, 0xd6eecac4UL, 0x6fd61d59UL, 0x0ab1a1e1UL, 0xe41e14f3UL, 0x8179a84bUL, 0xd769cb13UL, 0xb20e77abUL,
0x5ca1c2b9UL, 0x39c67e01UL, 0x80fea99cUL, 0xe5991524UL, 0x0b36a036UL, 0x6e511c8eUL, 0xa7166686UL, 0xc271da3eUL, 0x2cde6f2cUL,
0x49b9d394UL, 0xf0810409UL, 0x95e6b8b1UL, 0x7b490da3UL, 0x1e2eb11bUL, 0x483ed243UL, 0x2d596efbUL, 0xc3f6dbe9UL, 0xa6916751UL,
0x1fa9b0ccUL, 0x7ace0c74UL, 0x9461b966UL, 0xf10605deUL
#endif
}};

2163
base/poco/Foundation/src/deflate.c
View File

File diff suppressed because it is too large Load Diff

355
base/poco/Foundation/src/deflate.h
View File

@ -1,355 +0,0 @@
/* deflate.h -- internal compression state
* Copyright (C) 1995-2016 Jean-loup Gailly
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* WARNING: this file should *not* be used by applications. It is
part of the implementation of the compression library and is
subject to change. Applications should only use zlib.h.
*/
/* @(#) $Id$ */
#ifndef DEFLATE_H
#define DEFLATE_H
#include "zutil.h"
/* define NO_GZIP when compiling if you want to disable gzip header and
trailer creation by deflate(). NO_GZIP would be used to avoid linking in
the crc code when it is not needed. For shared libraries, gzip encoding
should be left enabled. */
#ifndef NO_GZIP
# define GZIP
#endif
/* ===========================================================================
* Internal compression state.
*/
#define LENGTH_CODES 29
/* number of length codes, not counting the special END_BLOCK code */
#define LITERALS 256
/* number of literal bytes 0..255 */
#define L_CODES (LITERALS + 1 + LENGTH_CODES)
/* number of Literal or Length codes, including the END_BLOCK code */
#define D_CODES 30
/* number of distance codes */
#define BL_CODES 19
/* number of codes used to transfer the bit lengths */
#define HEAP_SIZE (2 * L_CODES + 1)
/* maximum heap size */
#define MAX_BITS 15
/* All codes must not exceed MAX_BITS bits */
#define Buf_size 16
/* size of bit buffer in bi_buf */
#define INIT_STATE 42 /* zlib header -> BUSY_STATE */
#ifdef GZIP
# define GZIP_STATE 57 /* gzip header -> BUSY_STATE | EXTRA_STATE */
#endif
#define EXTRA_STATE 69 /* gzip extra block -> NAME_STATE */
#define NAME_STATE 73 /* gzip file name -> COMMENT_STATE */
#define COMMENT_STATE 91 /* gzip comment -> HCRC_STATE */
#define HCRC_STATE 103 /* gzip header CRC -> BUSY_STATE */
#define BUSY_STATE 113 /* deflate -> FINISH_STATE */
#define FINISH_STATE 666 /* stream complete */
/* Stream status */
/* Data structure describing a single value and its code string. */
typedef struct ct_data_s
{
union
{
ush freq; /* frequency count */
ush code; /* bit string */
} fc;
union
{
ush dad; /* father node in Huffman tree */
ush len; /* length of bit string */
} dl;
} FAR ct_data;
#define Freq fc.freq
#define Code fc.code
#define Dad dl.dad
#define Len dl.len
typedef struct static_tree_desc_s static_tree_desc;
typedef struct tree_desc_s
{
ct_data * dyn_tree; /* the dynamic tree */
int max_code; /* largest code with non zero frequency */
const static_tree_desc * stat_desc; /* the corresponding static tree */
} FAR tree_desc;
typedef ush Pos;
typedef Pos FAR Posf;
typedef unsigned IPos;
/* A Pos is an index in the character window. We use short instead of int to
* save space in the various tables. IPos is used only for parameter passing.
*/
typedef struct internal_state
{
z_streamp strm; /* pointer back to this zlib stream */
int status; /* as the name implies */
Bytef * pending_buf; /* output still pending */
ulg pending_buf_size; /* size of pending_buf */
Bytef * pending_out; /* next pending byte to output to the stream */
ulg pending; /* nb of bytes in the pending buffer */
int wrap; /* bit 0 true for zlib, bit 1 true for gzip */
gz_headerp gzhead; /* gzip header information to write */
ulg gzindex; /* where in extra, name, or comment */
Byte method; /* can only be DEFLATED */
int last_flush; /* value of flush param for previous deflate call */
/* used by deflate.c: */
uInt w_size; /* LZ77 window size (32K by default) */
uInt w_bits; /* log2(w_size) (8..16) */
uInt w_mask; /* w_size - 1 */
Bytef * window;
/* Sliding window. Input bytes are read into the second half of the window,
* and move to the first half later to keep a dictionary of at least wSize
* bytes. With this organization, matches are limited to a distance of
* wSize-MAX_MATCH bytes, but this ensures that IO is always
* performed with a length multiple of the block size. Also, it limits
* the window size to 64K, which is quite useful on MS-DOS.
* To do: use the user input buffer as sliding window.
*/
ulg window_size;
/* Actual size of window: 2*wSize, except when the user input buffer
* is directly used as sliding window.
*/
Posf * prev;
/* Link to older string with same hash index. To limit the size of this
* array to 64K, this link is maintained only for the last 32K strings.
* An index in this array is thus a window index modulo 32K.
*/
Posf * head; /* Heads of the hash chains or NIL. */
uInt ins_h; /* hash index of string to be inserted */
uInt hash_size; /* number of elements in hash table */
uInt hash_bits; /* log2(hash_size) */
uInt hash_mask; /* hash_size-1 */
uInt hash_shift;
/* Number of bits by which ins_h must be shifted at each input
* step. It must be such that after MIN_MATCH steps, the oldest
* byte no longer takes part in the hash key, that is:
* hash_shift * MIN_MATCH >= hash_bits
*/
long block_start;
/* Window position at the beginning of the current output block. Gets
* negative when the window is moved backwards.
*/
uInt match_length; /* length of best match */
IPos prev_match; /* previous match */
int match_available; /* set if previous match exists */
uInt strstart; /* start of string to insert */
uInt match_start; /* start of matching string */
uInt lookahead; /* number of valid bytes ahead in window */
uInt prev_length;
/* Length of the best match at previous step. Matches not greater than this
* are discarded. This is used in the lazy match evaluation.
*/
uInt max_chain_length;
/* To speed up deflation, hash chains are never searched beyond this
* length. A higher limit improves compression ratio but degrades the
* speed.
*/
uInt max_lazy_match;
/* Attempt to find a better match only when the current match is strictly
* smaller than this value. This mechanism is used only for compression
* levels >= 4.
*/
#define max_insert_length max_lazy_match
/* Insert new strings in the hash table only if the match length is not
* greater than this length. This saves time but degrades compression.
* max_insert_length is used only for compression levels <= 3.
*/
int level; /* compression level (1..9) */
int strategy; /* favor or force Huffman coding*/
uInt good_match;
/* Use a faster search when the previous match is longer than this */
int nice_match; /* Stop searching when current match exceeds this */
/* used by trees.c: */
/* Didn't use ct_data typedef below to suppress compiler warning */
struct ct_data_s dyn_ltree[HEAP_SIZE]; /* literal and length tree */
struct ct_data_s dyn_dtree[2 * D_CODES + 1]; /* distance tree */
struct ct_data_s bl_tree[2 * BL_CODES + 1]; /* Huffman tree for bit lengths */
struct tree_desc_s l_desc; /* desc. for literal tree */
struct tree_desc_s d_desc; /* desc. for distance tree */
struct tree_desc_s bl_desc; /* desc. for bit length tree */
ush bl_count[MAX_BITS + 1];
/* number of codes at each bit length for an optimal tree */
int heap[2 * L_CODES + 1]; /* heap used to build the Huffman trees */
int heap_len; /* number of elements in the heap */
int heap_max; /* element of largest frequency */
/* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
* The same heap array is used to build all trees.
*/
uch depth[2 * L_CODES + 1];
/* Depth of each subtree used as tie breaker for trees of equal frequency
*/
uchf * l_buf; /* buffer for literals or lengths */
uInt lit_bufsize;
/* Size of match buffer for literals/lengths. There are 4 reasons for
* limiting lit_bufsize to 64K:
* - frequencies can be kept in 16 bit counters
* - if compression is not successful for the first block, all input
* data is still in the window so we can still emit a stored block even
* when input comes from standard input. (This can also be done for
* all blocks if lit_bufsize is not greater than 32K.)
* - if compression is not successful for a file smaller than 64K, we can
* even emit a stored file instead of a stored block (saving 5 bytes).
* This is applicable only for zip (not gzip or zlib).
* - creating new Huffman trees less frequently may not provide fast
* adaptation to changes in the input data statistics. (Take for
* example a binary file with poorly compressible code followed by
* a highly compressible string table.) Smaller buffer sizes give
* fast adaptation but have of course the overhead of transmitting
* trees more frequently.
* - I can't count above 4
*/
uInt last_lit; /* running index in l_buf */
ushf * d_buf;
/* Buffer for distances. To simplify the code, d_buf and l_buf have
* the same number of elements. To use different lengths, an extra flag
* array would be necessary.
*/
ulg opt_len; /* bit length of current block with optimal trees */
ulg static_len; /* bit length of current block with static trees */
uInt matches; /* number of string matches in current block */
uInt insert; /* bytes at end of window left to insert */
#ifdef ZLIB_DEBUG
ulg compressed_len; /* total bit length of compressed file mod 2^32 */
ulg bits_sent; /* bit length of compressed data sent mod 2^32 */
#endif
ush bi_buf;
/* Output buffer. bits are inserted starting at the bottom (least
* significant bits).
*/
int bi_valid;
/* Number of valid bits in bi_buf. All bits above the last valid bit
* are always zero.
*/
ulg high_water;
/* High water mark offset in window for initialized bytes -- bytes above
* this are set to zero in order to avoid memory check warnings when
* longest match routines access bytes past the input. This is then
* updated to the new high water mark.
*/
} FAR deflate_state;
/* Output a byte on the stream.
* IN assertion: there is enough room in pending_buf.
*/
#define put_byte(s, c) \
{ \
s->pending_buf[s->pending++] = (Bytef)(c); \
}
#define MIN_LOOKAHEAD (MAX_MATCH + MIN_MATCH + 1)
/* Minimum amount of lookahead, except at the end of the input file.
* See deflate.c for comments about the MIN_MATCH+1.
*/
#define MAX_DIST(s) ((s)->w_size - MIN_LOOKAHEAD)
/* In order to simplify the code, particularly on 16 bit machines, match
* distances are limited to MAX_DIST instead of WSIZE.
*/
#define WIN_INIT MAX_MATCH
/* Number of bytes after end of data in window to initialize in order to avoid
memory checker errors from longest match routines */
/* in trees.c */
void ZLIB_INTERNAL _tr_init OF((deflate_state * s));
int ZLIB_INTERNAL _tr_tally OF((deflate_state * s, unsigned dist, unsigned lc));
void ZLIB_INTERNAL _tr_flush_block OF((deflate_state * s, charf * buf, ulg stored_len, int last));
void ZLIB_INTERNAL _tr_flush_bits OF((deflate_state * s));
void ZLIB_INTERNAL _tr_align OF((deflate_state * s));
void ZLIB_INTERNAL _tr_stored_block OF((deflate_state * s, charf * buf, ulg stored_len, int last));
#define d_code(dist) ((dist) < 256 ? _dist_code[dist] : _dist_code[256 + ((dist) >> 7)])
/* Mapping from a distance to a distance code. dist is the distance - 1 and
* must not have side effects. _dist_code[256] and _dist_code[257] are never
* used.
*/
#ifndef ZLIB_DEBUG
/* Inline versions of _tr_tally for speed: */
# if defined(GEN_TREES_H) || !defined(STDC)
extern uch ZLIB_INTERNAL _length_code[];
extern uch ZLIB_INTERNAL _dist_code[];
# else
extern const uch ZLIB_INTERNAL _length_code[];
extern const uch ZLIB_INTERNAL _dist_code[];
# endif
# define _tr_tally_lit(s, c, flush) \
{ \
uch cc = (c); \
s->d_buf[s->last_lit] = 0; \
s->l_buf[s->last_lit++] = cc; \
s->dyn_ltree[cc].Freq++; \
flush = (s->last_lit == s->lit_bufsize - 1); \
}
# define _tr_tally_dist(s, distance, length, flush) \
{ \
uch len = (uch)(length); \
ush dist = (ush)(distance); \
s->d_buf[s->last_lit] = dist; \
s->l_buf[s->last_lit++] = len; \
dist--; \
s->dyn_ltree[_length_code[len] + LITERALS + 1].Freq++; \
s->dyn_dtree[d_code(dist)].Freq++; \
flush = (s->last_lit == s->lit_bufsize - 1); \
}
#else
# define _tr_tally_lit(s, c, flush) flush = _tr_tally(s, 0, c)
# define _tr_tally_dist(s, distance, length, flush) flush = _tr_tally(s, distance, length)
#endif
#endif /* DEFLATE_H */

57
base/poco/Foundation/src/diy-fp.cc
View File

@ -1,57 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include "diy-fp.h"
#include "utils.h"
namespace double_conversion {
void DiyFp::Multiply(const DiyFp& other) {
// Simply "emulates" a 128 bit multiplication.
// However: the resulting number only contains 64 bits. The least
// significant 64 bits are only used for rounding the most significant 64
// bits.
const uint64_t kM32 = 0xFFFFFFFFU;
uint64_t a = f_ >> 32;
uint64_t b = f_ & kM32;
uint64_t c = other.f_ >> 32;
uint64_t d = other.f_ & kM32;
uint64_t ac = a * c;
uint64_t bc = b * c;
uint64_t ad = a * d;
uint64_t bd = b * d;
uint64_t tmp = (bd >> 32) + (ad & kM32) + (bc & kM32);
// By adding 1U << 31 to tmp we round the final result.
// Halfway cases will be round up.
tmp += 1U << 31;
uint64_t result_f = ac + (ad >> 32) + (bc >> 32) + (tmp >> 32);
e_ += other.e_ + 64;
f_ = result_f;
}
} // namespace double_conversion

127
base/poco/Foundation/src/diy-fp.h
View File

@ -1,127 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_DIY_FP_H_
#define DOUBLE_CONVERSION_DIY_FP_H_
#include "utils.h"
namespace double_conversion
{
// This "Do It Yourself Floating Point" class implements a floating-point number
// with a uint64 significand and an int exponent. Normalized DiyFp numbers will
// have the most significant bit of the significand set.
// Multiplication and Subtraction do not normalize their results.
// DiyFp are not designed to contain special doubles (NaN and Infinity).
class DiyFp
{
public:
static const int kSignificandSize = 64;
DiyFp() : f_(0), e_(0) { }
DiyFp(uint64_t f, int e) : f_(f), e_(e) { }
// this = this - other.
// The exponents of both numbers must be the same and the significand of this
// must be bigger than the significand of other.
// The result will not be normalized.
void Subtract(const DiyFp & other)
{
ASSERT(e_ == other.e_);
ASSERT(f_ >= other.f_);
f_ -= other.f_;
}
// Returns a - b.
// The exponents of both numbers must be the same and this must be bigger
// than other. The result will not be normalized.
static DiyFp Minus(const DiyFp & a, const DiyFp & b)
{
DiyFp result = a;
result.Subtract(b);
return result;
}
// this = this * other.
void Multiply(const DiyFp & other);
// returns a * b;
static DiyFp Times(const DiyFp & a, const DiyFp & b)
{
DiyFp result = a;
result.Multiply(b);
return result;
}
void Normalize()
{
ASSERT(f_ != 0);
uint64_t f = f_;
int e = e_;
// This method is mainly called for normalizing boundaries. In general
// boundaries need to be shifted by 10 bits. We thus optimize for this case.
const uint64_t k10MSBits = UINT64_2PART_C(0xFFC00000, 00000000);
while ((f & k10MSBits) == 0)
{
f <<= 10;
e -= 10;
}
while ((f & kUint64MSB) == 0)
{
f <<= 1;
e--;
}
f_ = f;
e_ = e;
}
static DiyFp Normalize(const DiyFp & a)
{
DiyFp result = a;
result.Normalize();
return result;
}
uint64_t f() const { return f_; }
int e() const { return e_; }
void set_f(uint64_t new_value) { f_ = new_value; }
void set_e(int new_value) { e_ = new_value; }
private:
static const uint64_t kUint64MSB = UINT64_2PART_C(0x80000000, 00000000);
uint64_t f_;
int e_;
};
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_DIY_FP_H_

911
base/poco/Foundation/src/double-conversion.cc
View File

@ -1,911 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <limits.h>
#include <math.h>
#include "double-conversion.h"
#include "bignum-dtoa.h"
#include "fast-dtoa.h"
#include "fixed-dtoa.h"
#include "ieee.h"
#include "strtod.h"
#include "utils.h"
namespace double_conversion {
const DoubleToStringConverter& DoubleToStringConverter::EcmaScriptConverter() {
int flags = UNIQUE_ZERO | EMIT_POSITIVE_EXPONENT_SIGN;
static DoubleToStringConverter converter(flags,
"Infinity",
"NaN",
'e',
-6, 21,
6, 0);
return converter;
}
bool DoubleToStringConverter::HandleSpecialValues(
double value,
StringBuilder* result_builder) const {
Double double_inspect(value);
if (double_inspect.IsInfinite()) {
if (infinity_symbol_ == NULL) return false;
if (value < 0) {
result_builder->AddCharacter('-');
}
result_builder->AddString(infinity_symbol_);
return true;
}
if (double_inspect.IsNan()) {
if (nan_symbol_ == NULL) return false;
result_builder->AddString(nan_symbol_);
return true;
}
return false;
}
void DoubleToStringConverter::CreateExponentialRepresentation(
const char* decimal_digits,
int length,
int exponent,
StringBuilder* result_builder) const {
ASSERT(length != 0);
result_builder->AddCharacter(decimal_digits[0]);
if (length != 1) {
result_builder->AddCharacter('.');
result_builder->AddSubstring(&decimal_digits[1], length-1);
}
result_builder->AddCharacter(exponent_character_);
if (exponent < 0) {
result_builder->AddCharacter('-');
exponent = -exponent;
} else {
if ((flags_ & EMIT_POSITIVE_EXPONENT_SIGN) != 0) {
result_builder->AddCharacter('+');
}
}
if (exponent == 0) {
result_builder->AddCharacter('0');
return;
}
ASSERT(exponent < 1e4);
const int kMaxExponentLength = 5;
char buffer[kMaxExponentLength + 1];
buffer[kMaxExponentLength] = '\0';
int first_char_pos = kMaxExponentLength;
while (exponent > 0) {
buffer[--first_char_pos] = '0' + (exponent % 10);
exponent /= 10;
}
result_builder->AddSubstring(&buffer[first_char_pos],
kMaxExponentLength - first_char_pos);
}
void DoubleToStringConverter::CreateDecimalRepresentation(
const char* decimal_digits,
int length,
int decimal_point,
int digits_after_point,
StringBuilder* result_builder) const {
// Create a representation that is padded with zeros if needed.
if (decimal_point <= 0) {
// "0.00000decimal_rep".
result_builder->AddCharacter('0');
if (digits_after_point > 0) {
result_builder->AddCharacter('.');
result_builder->AddPadding('0', -decimal_point);
ASSERT(length <= digits_after_point - (-decimal_point));
result_builder->AddSubstring(decimal_digits, length);
int remaining_digits = digits_after_point - (-decimal_point) - length;
result_builder->AddPadding('0', remaining_digits);
}
} else if (decimal_point >= length) {
// "decimal_rep0000.00000" or "decimal_rep.0000"
result_builder->AddSubstring(decimal_digits, length);
result_builder->AddPadding('0', decimal_point - length);
if (digits_after_point > 0) {
result_builder->AddCharacter('.');
result_builder->AddPadding('0', digits_after_point);
}
} else {
// "decima.l_rep000"
ASSERT(digits_after_point > 0);
result_builder->AddSubstring(decimal_digits, decimal_point);
result_builder->AddCharacter('.');
ASSERT(length - decimal_point <= digits_after_point);
result_builder->AddSubstring(&decimal_digits[decimal_point],
length - decimal_point);
int remaining_digits = digits_after_point - (length - decimal_point);
result_builder->AddPadding('0', remaining_digits);
}
if (digits_after_point == 0) {
if ((flags_ & EMIT_TRAILING_DECIMAL_POINT) != 0) {
result_builder->AddCharacter('.');
}
if ((flags_ & EMIT_TRAILING_ZERO_AFTER_POINT) != 0) {
result_builder->AddCharacter('0');
}
}
}
bool DoubleToStringConverter::ToShortestIeeeNumber(
double value,
StringBuilder* result_builder,
DoubleToStringConverter::DtoaMode mode) const {
ASSERT(mode == SHORTEST || mode == SHORTEST_SINGLE);
if (Double(value).IsSpecial()) {
return HandleSpecialValues(value, result_builder);
}
int decimal_point;
bool sign;
const int kDecimalRepCapacity = kBase10MaximalLength + 1;
char decimal_rep[kDecimalRepCapacity];
int decimal_rep_length;
DoubleToAscii(value, mode, 0, decimal_rep, kDecimalRepCapacity,
&sign, &decimal_rep_length, &decimal_point);
bool unique_zero = (flags_ & UNIQUE_ZERO) != 0;
if (sign && (value != 0.0 || !unique_zero)) {
result_builder->AddCharacter('-');
}
int exponent = decimal_point - 1;
if ((decimal_in_shortest_low_ <= exponent) &&
(exponent < decimal_in_shortest_high_)) {
CreateDecimalRepresentation(decimal_rep, decimal_rep_length,
decimal_point,
Max(0, decimal_rep_length - decimal_point),
result_builder);
} else {
CreateExponentialRepresentation(decimal_rep, decimal_rep_length, exponent,
result_builder);
}
return true;
}
bool DoubleToStringConverter::ToFixed(double value,
int requested_digits,
StringBuilder* result_builder) const {
ASSERT(kMaxFixedDigitsBeforePoint == 60);
const double kFirstNonFixed = 1e60;
if (Double(value).IsSpecial()) {
return HandleSpecialValues(value, result_builder);
}
if (requested_digits > kMaxFixedDigitsAfterPoint) return false;
if (value >= kFirstNonFixed || value <= -kFirstNonFixed) return false;
// Find a sufficiently precise decimal representation of n.
int decimal_point;
bool sign;
// Add space for the '\0' byte.
const int kDecimalRepCapacity =
kMaxFixedDigitsBeforePoint + kMaxFixedDigitsAfterPoint + 1;
char decimal_rep[kDecimalRepCapacity];
int decimal_rep_length;
DoubleToAscii(value, FIXED, requested_digits,
decimal_rep, kDecimalRepCapacity,
&sign, &decimal_rep_length, &decimal_point);
bool unique_zero = ((flags_ & UNIQUE_ZERO) != 0);
if (sign && (value != 0.0 || !unique_zero)) {
result_builder->AddCharacter('-');
}
CreateDecimalRepresentation(decimal_rep, decimal_rep_length, decimal_point,
requested_digits, result_builder);
return true;
}
bool DoubleToStringConverter::ToExponential(
double value,
int requested_digits,
StringBuilder* result_builder) const {
if (Double(value).IsSpecial()) {
return HandleSpecialValues(value, result_builder);
}
if (requested_digits < -1) return false;
if (requested_digits > kMaxExponentialDigits) return false;
int decimal_point;
bool sign;
// Add space for digit before the decimal point and the '\0' character.
const int kDecimalRepCapacity = kMaxExponentialDigits + 2;
ASSERT(kDecimalRepCapacity > kBase10MaximalLength);
char decimal_rep[kDecimalRepCapacity];
int decimal_rep_length;
if (requested_digits == -1) {
DoubleToAscii(value, SHORTEST, 0,
decimal_rep, kDecimalRepCapacity,
&sign, &decimal_rep_length, &decimal_point);
} else {
DoubleToAscii(value, PRECISION, requested_digits + 1,
decimal_rep, kDecimalRepCapacity,
&sign, &decimal_rep_length, &decimal_point);
ASSERT(decimal_rep_length <= requested_digits + 1);
for (int i = decimal_rep_length; i < requested_digits + 1; ++i) {
decimal_rep[i] = '0';
}
decimal_rep_length = requested_digits + 1;
}
bool unique_zero = ((flags_ & UNIQUE_ZERO) != 0);
if (sign && (value != 0.0 || !unique_zero)) {
result_builder->AddCharacter('-');
}
int exponent = decimal_point - 1;
CreateExponentialRepresentation(decimal_rep,
decimal_rep_length,
exponent,
result_builder);
return true;
}
bool DoubleToStringConverter::ToPrecision(double value,
int precision,
StringBuilder* result_builder) const {
if (Double(value).IsSpecial()) {
return HandleSpecialValues(value, result_builder);
}
if (precision < kMinPrecisionDigits || precision > kMaxPrecisionDigits) {
return false;
}
// Find a sufficiently precise decimal representation of n.
int decimal_point;
bool sign;
// Add one for the terminating null character.
const int kDecimalRepCapacity = kMaxPrecisionDigits + 1;
char decimal_rep[kDecimalRepCapacity];
int decimal_rep_length;
DoubleToAscii(value, PRECISION, precision,
decimal_rep, kDecimalRepCapacity,
&sign, &decimal_rep_length, &decimal_point);
ASSERT(decimal_rep_length <= precision);
bool unique_zero = ((flags_ & UNIQUE_ZERO) != 0);
if (sign && (value != 0.0 || !unique_zero)) {
result_builder->AddCharacter('-');
}
// The exponent if we print the number as x.xxeyyy. That is with the
// decimal point after the first digit.
int exponent = decimal_point - 1;
int extra_zero = ((flags_ & EMIT_TRAILING_ZERO_AFTER_POINT) != 0) ? 1 : 0;
if ((-decimal_point + 1 > max_leading_padding_zeroes_in_precision_mode_) ||
(decimal_point - precision + extra_zero >
max_trailing_padding_zeroes_in_precision_mode_)) {
// Fill buffer to contain 'precision' digits.
// Usually the buffer is already at the correct length, but 'DoubleToAscii'
// is allowed to return less characters.
for (int i = decimal_rep_length; i < precision; ++i) {
decimal_rep[i] = '0';
}
CreateExponentialRepresentation(decimal_rep,
precision,
exponent,
result_builder);
} else {
CreateDecimalRepresentation(decimal_rep, decimal_rep_length, decimal_point,
Max(0, precision - decimal_point),
result_builder);
}
return true;
}
static BignumDtoaMode DtoaToBignumDtoaMode(
DoubleToStringConverter::DtoaMode dtoa_mode) {
switch (dtoa_mode) {
case DoubleToStringConverter::SHORTEST: return BIGNUM_DTOA_SHORTEST;
case DoubleToStringConverter::SHORTEST_SINGLE:
return BIGNUM_DTOA_SHORTEST_SINGLE;
case DoubleToStringConverter::FIXED: return BIGNUM_DTOA_FIXED;
case DoubleToStringConverter::PRECISION: return BIGNUM_DTOA_PRECISION;
default:
UNREACHABLE();
return BIGNUM_DTOA_SHORTEST;
}
}
void DoubleToStringConverter::DoubleToAscii(double v,
DtoaMode mode,
int requested_digits,
char* buffer,
int buffer_length,
bool* sign,
int* length,
int* point) {
Vector<char> vector(buffer, buffer_length);
ASSERT(!Double(v).IsSpecial());
ASSERT(mode == SHORTEST || mode == SHORTEST_SINGLE || requested_digits >= 0);
if (Double(v).Sign() < 0) {
*sign = true;
v = -v;
} else {
*sign = false;
}
if (mode == PRECISION && requested_digits == 0) {
vector[0] = '\0';
*length = 0;
return;
}
if (v == 0) {
vector[0] = '0';
vector[1] = '\0';
*length = 1;
*point = 1;
return;
}
bool fast_worked;
switch (mode) {
case SHORTEST:
fast_worked = FastDtoa(v, FAST_DTOA_SHORTEST, 0, vector, length, point);
break;
case SHORTEST_SINGLE:
fast_worked = FastDtoa(v, FAST_DTOA_SHORTEST_SINGLE, 0,
vector, length, point);
break;
case FIXED:
fast_worked = FastFixedDtoa(v, requested_digits, vector, length, point);
break;
case PRECISION:
fast_worked = FastDtoa(v, FAST_DTOA_PRECISION, requested_digits,
vector, length, point);
break;
default:
fast_worked = false;
UNREACHABLE();
}
if (fast_worked) return;
// If the fast dtoa didn't succeed use the slower bignum version.
BignumDtoaMode bignum_mode = DtoaToBignumDtoaMode(mode);
BignumDtoa(v, bignum_mode, requested_digits, vector, length, point);
vector[*length] = '\0';
}
// Consumes the given substring from the iterator.
// Returns false, if the substring does not match.
static bool ConsumeSubString(const char** current,
const char* end,
const char* substring) {
ASSERT(**current == *substring);
for (substring++; *substring != '\0'; substring++) {
++*current;
if (*current == end || **current != *substring) return false;
}
++*current;
return true;
}
// Maximum number of significant digits in decimal representation.
// The longest possible double in decimal representation is
// (2^53 - 1) * 2 ^ -1074 that is (2 ^ 53 - 1) * 5 ^ 1074 / 10 ^ 1074
// (768 digits). If we parse a number whose first digits are equal to a
// mean of 2 adjacent doubles (that could have up to 769 digits) the result
// must be rounded to the bigger one unless the tail consists of zeros, so
// we don't need to preserve all the digits.
const int kMaxSignificantDigits = 772;
// Returns true if a nonspace found and false if the end has reached.
static inline bool AdvanceToNonspace(const char** current, const char* end) {
while (*current != end) {
if (**current != ' ') return true;
++*current;
}
return false;
}
static bool isDigit(int x, int radix) {
return (x >= '0' && x <= '9' && x < '0' + radix)
|| (radix > 10 && x >= 'a' && x < 'a' + radix - 10)
|| (radix > 10 && x >= 'A' && x < 'A' + radix - 10);
}
static double SignedZero(bool sign) {
return sign ? -0.0 : 0.0;
}
// Returns true if 'c' is a decimal digit that is valid for the given radix.
//
// The function is small and could be inlined, but VS2012 emitted a warning
// because it constant-propagated the radix and concluded that the last
// condition was always true. By moving it into a separate function the
// compiler wouldn't warn anymore.
static bool IsDecimalDigitForRadix(int c, int radix) {
return '0' <= c && c <= '9' && (c - '0') < radix;
}
// Returns true if 'c' is a character digit that is valid for the given radix.
// The 'a_character' should be 'a' or 'A'.
//
// The function is small and could be inlined, but VS2012 emitted a warning
// because it constant-propagated the radix and concluded that the first
// condition was always false. By moving it into a separate function the
// compiler wouldn't warn anymore.
static bool IsCharacterDigitForRadix(int c, int radix, char a_character) {
return radix > 10 && c >= a_character && c < a_character + radix - 10;
}
// Parsing integers with radix 2, 4, 8, 16, 32. Assumes current != end.
template <int radix_log_2>
static double RadixStringToIeee(const char* current,
const char* end,
bool sign,
bool allow_trailing_junk,
double junk_string_value,
bool read_as_double,
const char** trailing_pointer) {
ASSERT(current != end);
const int kDoubleSize = Double::kSignificandSize;
const int kSingleSize = Single::kSignificandSize;
const int kSignificandSize = read_as_double? kDoubleSize: kSingleSize;
// Skip leading 0s.
while (*current == '0') {
++current;
if (current == end) {
*trailing_pointer = end;
return SignedZero(sign);
}
}
int64_t number = 0;
int exponent = 0;
const int radix = (1 << radix_log_2);
do {
int digit;
if (IsDecimalDigitForRadix(*current, radix)) {
digit = static_cast<char>(*current) - '0';
} else if (IsCharacterDigitForRadix(*current, radix, 'a')) {
digit = static_cast<char>(*current) - 'a' + 10;
} else if (IsCharacterDigitForRadix(*current, radix, 'A')) {
digit = static_cast<char>(*current) - 'A' + 10;
} else {
if (allow_trailing_junk || !AdvanceToNonspace(&current, end)) {
break;
} else {
return junk_string_value;
}
}
number = number * radix + digit;
int overflow = static_cast<int>(number >> kSignificandSize);
if (overflow != 0) {
// Overflow occurred. Need to determine which direction to round the
// result.
int overflow_bits_count = 1;
while (overflow > 1) {
overflow_bits_count++;
overflow >>= 1;
}
int dropped_bits_mask = ((1 << overflow_bits_count) - 1);
int dropped_bits = static_cast<int>(number) & dropped_bits_mask;
number >>= overflow_bits_count;
exponent = overflow_bits_count;
bool zero_tail = true;
for (;;) {
++current;
if (current == end || !isDigit(*current, radix)) break;
zero_tail = zero_tail && *current == '0';
exponent += radix_log_2;
}
if (!allow_trailing_junk && AdvanceToNonspace(&current, end)) {
return junk_string_value;
}
int middle_value = (1 << (overflow_bits_count - 1));
if (dropped_bits > middle_value) {
number++; // Rounding up.
} else if (dropped_bits == middle_value) {
// Rounding to even to consistency with decimals: half-way case rounds
// up if significant part is odd and down otherwise.
if ((number & 1) != 0 || !zero_tail) {
number++; // Rounding up.
}
}
// Rounding up may cause overflow.
if ((number & ((int64_t)1 << kSignificandSize)) != 0) {
exponent++;
number >>= 1;
}
break;
}
++current;
} while (current != end);
ASSERT(number < ((int64_t)1 << kSignificandSize));
ASSERT(static_cast<int64_t>(static_cast<double>(number)) == number);
*trailing_pointer = current;
if (exponent == 0) {
if (sign) {
if (number == 0) return -0.0;
number = -number;
}
return static_cast<double>(number);
}
ASSERT(number != 0);
return Double(DiyFp(number, exponent)).value();
}
double StringToDoubleConverter::StringToIeee(
const char* input,
int length,
int* processed_characters_count,
bool read_as_double) const {
const char* current = input;
const char* end = input + length;
*processed_characters_count = 0;
const bool allow_trailing_junk = (flags_ & ALLOW_TRAILING_JUNK) != 0;
const bool allow_leading_spaces = (flags_ & ALLOW_LEADING_SPACES) != 0;
const bool allow_trailing_spaces = (flags_ & ALLOW_TRAILING_SPACES) != 0;
const bool allow_spaces_after_sign = (flags_ & ALLOW_SPACES_AFTER_SIGN) != 0;
// To make sure that iterator dereferencing is valid the following
// convention is used:
// 1. Each '++current' statement is followed by check for equality to 'end'.
// 2. If AdvanceToNonspace returned false then current == end.
// 3. If 'current' becomes equal to 'end' the function returns or goes to
// 'parsing_done'.
// 4. 'current' is not dereferenced after the 'parsing_done' label.
// 5. Code before 'parsing_done' may rely on 'current != end'.
if (current == end) return empty_string_value_;
if (allow_leading_spaces || allow_trailing_spaces) {
if (!AdvanceToNonspace(&current, end)) {
*processed_characters_count = static_cast<int>(current - input);
return empty_string_value_;
}
if (!allow_leading_spaces && (input != current)) {
// No leading spaces allowed, but AdvanceToNonspace moved forward.
return junk_string_value_;
}
}
// The longest form of simplified number is: "-<significant digits>.1eXXX\0".
const int kBufferSize = kMaxSignificantDigits + 10;
char buffer[kBufferSize]; // NOLINT: size is known at compile time.
int buffer_pos = 0;
// Exponent will be adjusted if insignificant digits of the integer part
// or insignificant leading zeros of the fractional part are dropped.
int exponent = 0;
int significant_digits = 0;
int insignificant_digits = 0;
bool nonzero_digit_dropped = false;
bool sign = false;
if (*current == '+' || *current == '-') {
sign = (*current == '-');
++current;
const char* next_non_space = current;
// Skip following spaces (if allowed).
if (!AdvanceToNonspace(&next_non_space, end)) return junk_string_value_;
if (!allow_spaces_after_sign && (current != next_non_space)) {
return junk_string_value_;
}
current = next_non_space;
}
if (infinity_symbol_ != NULL) {
if (*current == infinity_symbol_[0]) {
if (!ConsumeSubString(&current, end, infinity_symbol_)) {
return junk_string_value_;
}
if (!(allow_trailing_spaces || allow_trailing_junk) && (current != end)) {
return junk_string_value_;
}
if (!allow_trailing_junk && AdvanceToNonspace(&current, end)) {
return junk_string_value_;
}
ASSERT(buffer_pos == 0);
*processed_characters_count = static_cast<int>(current - input);
return sign ? -Double::Infinity() : Double::Infinity();
}
}
if (nan_symbol_ != NULL) {
if (*current == nan_symbol_[0]) {
if (!ConsumeSubString(&current, end, nan_symbol_)) {
return junk_string_value_;
}
if (!(allow_trailing_spaces || allow_trailing_junk) && (current != end)) {
return junk_string_value_;
}
if (!allow_trailing_junk && AdvanceToNonspace(&current, end)) {
return junk_string_value_;
}
ASSERT(buffer_pos == 0);
*processed_characters_count = static_cast<int>(current - input);
return sign ? -Double::NaN() : Double::NaN();
}
}
bool leading_zero = false;
if (*current == '0') {
++current;
if (current == end) {
*processed_characters_count = static_cast<int>(current - input);
return SignedZero(sign);
}
leading_zero = true;
// It could be hexadecimal value.
if ((flags_ & ALLOW_HEX) && (*current == 'x' || *current == 'X')) {
++current;
if (current == end || !isDigit(*current, 16)) {
return junk_string_value_; // "0x".
}
const char* tail_pointer = NULL;
double result = RadixStringToIeee<4>(current,
end,
sign,
allow_trailing_junk,
junk_string_value_,
read_as_double,
&tail_pointer);
if (tail_pointer != NULL) {
if (allow_trailing_spaces) AdvanceToNonspace(&tail_pointer, end);
*processed_characters_count = static_cast<int>(tail_pointer - input);
}
return result;
}
// Ignore leading zeros in the integer part.
while (*current == '0') {
++current;
if (current == end) {
*processed_characters_count = static_cast<int>(current - input);
return SignedZero(sign);
}
}
}
bool octal = leading_zero && (flags_ & ALLOW_OCTALS) != 0;
// Copy significant digits of the integer part (if any) to the buffer.
while (*current >= '0' && *current <= '9') {
if (significant_digits < kMaxSignificantDigits) {
ASSERT(buffer_pos < kBufferSize);
buffer[buffer_pos++] = static_cast<char>(*current);
significant_digits++;
// Will later check if it's an octal in the buffer.
} else {
insignificant_digits++; // Move the digit into the exponential part.
nonzero_digit_dropped = nonzero_digit_dropped || *current != '0';
}
octal = octal && *current < '8';
++current;
if (current == end) goto parsing_done;
}
if (significant_digits == 0) {
octal = false;
}
if (*current == '.') {
if (octal && !allow_trailing_junk) return junk_string_value_;
if (octal) goto parsing_done;
++current;
if (current == end) {
if (significant_digits == 0 && !leading_zero) {
return junk_string_value_;
} else {
goto parsing_done;
}
}
if (significant_digits == 0) {
// octal = false;
// Integer part consists of 0 or is absent. Significant digits start after
// leading zeros (if any).
while (*current == '0') {
++current;
if (current == end) {
*processed_characters_count = static_cast<int>(current - input);
return SignedZero(sign);
}
exponent--; // Move this 0 into the exponent.
}
}
// There is a fractional part.
// We don't emit a '.', but adjust the exponent instead.
while (*current >= '0' && *current <= '9') {
if (significant_digits < kMaxSignificantDigits) {
ASSERT(buffer_pos < kBufferSize);
buffer[buffer_pos++] = static_cast<char>(*current);
significant_digits++;
exponent--;
} else {
// Ignore insignificant digits in the fractional part.
nonzero_digit_dropped = nonzero_digit_dropped || *current != '0';
}
++current;
if (current == end) goto parsing_done;
}
}
if (!leading_zero && exponent == 0 && significant_digits == 0) {
// If leading_zeros is true then the string contains zeros.
// If exponent < 0 then string was [+-]\.0*...
// If significant_digits != 0 the string is not equal to 0.
// Otherwise there are no digits in the string.
return junk_string_value_;
}
// Parse exponential part.
if (*current == 'e' || *current == 'E') {
if (octal && !allow_trailing_junk) return junk_string_value_;
if (octal) goto parsing_done;
++current;
if (current == end) {
if (allow_trailing_junk) {
goto parsing_done;
} else {
return junk_string_value_;
}
}
char sign = '+';
if (*current == '+' || *current == '-') {
sign = static_cast<char>(*current);
++current;
if (current == end) {
if (allow_trailing_junk) {
goto parsing_done;
} else {
return junk_string_value_;
}
}
}
if (current == end || *current < '0' || *current > '9') {
if (allow_trailing_junk) {
goto parsing_done;
} else {
return junk_string_value_;
}
}
const int max_exponent = INT_MAX / 2;
ASSERT(-max_exponent / 2 <= exponent && exponent <= max_exponent / 2);
int num = 0;
do {
// Check overflow.
int digit = *current - '0';
if (num >= max_exponent / 10
&& !(num == max_exponent / 10 && digit <= max_exponent % 10)) {
num = max_exponent;
} else {
num = num * 10 + digit;
}
++current;
} while (current != end && *current >= '0' && *current <= '9');
exponent += (sign == '-' ? -num : num);
}
if (!(allow_trailing_spaces || allow_trailing_junk) && (current != end)) {
return junk_string_value_;
}
if (!allow_trailing_junk && AdvanceToNonspace(&current, end)) {
return junk_string_value_;
}
if (allow_trailing_spaces) {
AdvanceToNonspace(&current, end);
}
parsing_done:
exponent += insignificant_digits;
if (octal) {
double result;
const char* tail_pointer = NULL;
result = RadixStringToIeee<3>(buffer,
buffer + buffer_pos,
sign,
allow_trailing_junk,
junk_string_value_,
read_as_double,
&tail_pointer);
ASSERT(tail_pointer != NULL);
*processed_characters_count = static_cast<int>(current - input);
return result;
}
if (nonzero_digit_dropped) {
buffer[buffer_pos++] = '1';
exponent--;
}
ASSERT(buffer_pos < kBufferSize);
buffer[buffer_pos] = '\0';
double converted;
if (read_as_double) {
converted = Strtod(Vector<const char>(buffer, buffer_pos), exponent);
} else {
converted = Strtof(Vector<const char>(buffer, buffer_pos), exponent);
}
*processed_characters_count = static_cast<int>(current - input);
return sign? -converted: converted;
}
} // namespace double_conversion

512
base/poco/Foundation/src/double-conversion.h
View File

@ -1,512 +0,0 @@
// Copyright 2012 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_DOUBLE_CONVERSION_H_
#define DOUBLE_CONVERSION_DOUBLE_CONVERSION_H_
#include "utils.h"
namespace double_conversion
{
class DoubleToStringConverter
{
public:
// When calling ToFixed with a double > 10^kMaxFixedDigitsBeforePoint
// or a requested_digits parameter > kMaxFixedDigitsAfterPoint then the
// function returns false.
static const int kMaxFixedDigitsBeforePoint = 60;
static const int kMaxFixedDigitsAfterPoint = 60;
// When calling ToExponential with a requested_digits
// parameter > kMaxExponentialDigits then the function returns false.
static const int kMaxExponentialDigits = 120;
// When calling ToPrecision with a requested_digits
// parameter < kMinPrecisionDigits or requested_digits > kMaxPrecisionDigits
// then the function returns false.
static const int kMinPrecisionDigits = 1;
static const int kMaxPrecisionDigits = 120;
enum Flags
{
NO_FLAGS = 0,
EMIT_POSITIVE_EXPONENT_SIGN = 1,
EMIT_TRAILING_DECIMAL_POINT = 2,
EMIT_TRAILING_ZERO_AFTER_POINT = 4,
UNIQUE_ZERO = 8
};
// Flags should be a bit-or combination of the possible Flags-enum.
// - NO_FLAGS: no special flags.
// - EMIT_POSITIVE_EXPONENT_SIGN: when the number is converted into exponent
// form, emits a '+' for positive exponents. Example: 1.2e+2.
// - EMIT_TRAILING_DECIMAL_POINT: when the input number is an integer and is
// converted into decimal format then a trailing decimal point is appended.
// Example: 2345.0 is converted to "2345.".
// - EMIT_TRAILING_ZERO_AFTER_POINT: in addition to a trailing decimal point
// emits a trailing '0'-character. This flag requires the
// EXMIT_TRAILING_DECIMAL_POINT flag.
// Example: 2345.0 is converted to "2345.0".
// - UNIQUE_ZERO: "-0.0" is converted to "0.0".
//
// Infinity symbol and nan_symbol provide the string representation for these
// special values. If the string is NULL and the special value is encountered
// then the conversion functions return false.
//
// The exponent_character is used in exponential representations. It is
// usually 'e' or 'E'.
//
// When converting to the shortest representation the converter will
// represent input numbers in decimal format if they are in the interval
// [10^decimal_in_shortest_low; 10^decimal_in_shortest_high[
// (lower boundary included, greater boundary excluded).
// Example: with decimal_in_shortest_low = -6 and
// decimal_in_shortest_high = 21:
// ToShortest(0.000001) -> "0.000001"
// ToShortest(0.0000001) -> "1e-7"
// ToShortest(111111111111111111111.0) -> "111111111111111110000"
// ToShortest(100000000000000000000.0) -> "100000000000000000000"
// ToShortest(1111111111111111111111.0) -> "1.1111111111111111e+21"
//
// When converting to precision mode the converter may add
// max_leading_padding_zeroes before returning the number in exponential
// format.
// Example with max_leading_padding_zeroes_in_precision_mode = 6.
// ToPrecision(0.0000012345, 2) -> "0.0000012"
// ToPrecision(0.00000012345, 2) -> "1.2e-7"
// Similarly the converter may add up to
// max_trailing_padding_zeroes_in_precision_mode in precision mode to avoid
// returning an exponential representation. A zero added by the
// EMIT_TRAILING_ZERO_AFTER_POINT flag is counted for this limit.
// Examples for max_trailing_padding_zeroes_in_precision_mode = 1:
// ToPrecision(230.0, 2) -> "230"
// ToPrecision(230.0, 2) -> "230." with EMIT_TRAILING_DECIMAL_POINT.
// ToPrecision(230.0, 2) -> "2.3e2" with EMIT_TRAILING_ZERO_AFTER_POINT.
DoubleToStringConverter(
int flags,
const char * infinity_symbol,
const char * nan_symbol,
char exponent_character,
int decimal_in_shortest_low,
int decimal_in_shortest_high,
int max_leading_padding_zeroes_in_precision_mode,
int max_trailing_padding_zeroes_in_precision_mode)
: flags_(flags)
, infinity_symbol_(infinity_symbol)
, nan_symbol_(nan_symbol)
, exponent_character_(exponent_character)
, decimal_in_shortest_low_(decimal_in_shortest_low)
, decimal_in_shortest_high_(decimal_in_shortest_high)
, max_leading_padding_zeroes_in_precision_mode_(max_leading_padding_zeroes_in_precision_mode)
, max_trailing_padding_zeroes_in_precision_mode_(max_trailing_padding_zeroes_in_precision_mode)
{
// When 'trailing zero after the point' is set, then 'trailing point'
// must be set too.
ASSERT(((flags & EMIT_TRAILING_DECIMAL_POINT) != 0) || !((flags & EMIT_TRAILING_ZERO_AFTER_POINT) != 0));
}
// Returns a converter following the EcmaScript specification.
static const DoubleToStringConverter & EcmaScriptConverter();
// Computes the shortest string of digits that correctly represent the input
// number. Depending on decimal_in_shortest_low and decimal_in_shortest_high
// (see constructor) it then either returns a decimal representation, or an
// exponential representation.
// Example with decimal_in_shortest_low = -6,
// decimal_in_shortest_high = 21,
// EMIT_POSITIVE_EXPONENT_SIGN activated, and
// EMIT_TRAILING_DECIMAL_POINT deactivated:
// ToShortest(0.000001) -> "0.000001"
// ToShortest(0.0000001) -> "1e-7"
// ToShortest(111111111111111111111.0) -> "111111111111111110000"
// ToShortest(100000000000000000000.0) -> "100000000000000000000"
// ToShortest(1111111111111111111111.0) -> "1.1111111111111111e+21"
//
// Note: the conversion may round the output if the returned string
// is accurate enough to uniquely identify the input-number.
// For example the most precise representation of the double 9e59 equals
// "899999999999999918767229449717619953810131273674690656206848", but
// the converter will return the shorter (but still correct) "9e59".
//
// Returns true if the conversion succeeds. The conversion always succeeds
// except when the input value is special and no infinity_symbol or
// nan_symbol has been given to the constructor.
bool ToShortest(double value, StringBuilder * result_builder) const { return ToShortestIeeeNumber(value, result_builder, SHORTEST); }
// Same as ToShortest, but for single-precision floats.
bool ToShortestSingle(float value, StringBuilder * result_builder) const
{
return ToShortestIeeeNumber(value, result_builder, SHORTEST_SINGLE);
}
// Computes a decimal representation with a fixed number of digits after the
// decimal point. The last emitted digit is rounded.
//
// Examples:
// ToFixed(3.12, 1) -> "3.1"
// ToFixed(3.1415, 3) -> "3.142"
// ToFixed(1234.56789, 4) -> "1234.5679"
// ToFixed(1.23, 5) -> "1.23000"
// ToFixed(0.1, 4) -> "0.1000"
// ToFixed(1e30, 2) -> "1000000000000000019884624838656.00"
// ToFixed(0.1, 30) -> "0.100000000000000005551115123126"
// ToFixed(0.1, 17) -> "0.10000000000000001"
//
// If requested_digits equals 0, then the tail of the result depends on
// the EMIT_TRAILING_DECIMAL_POINT and EMIT_TRAILING_ZERO_AFTER_POINT.
// Examples, for requested_digits == 0,
// let EMIT_TRAILING_DECIMAL_POINT and EMIT_TRAILING_ZERO_AFTER_POINT be
// - false and false: then 123.45 -> 123
// 0.678 -> 1
// - true and false: then 123.45 -> 123.
// 0.678 -> 1.
// - true and true: then 123.45 -> 123.0
// 0.678 -> 1.0
//
// Returns true if the conversion succeeds. The conversion always succeeds
// except for the following cases:
// - the input value is special and no infinity_symbol or nan_symbol has
// been provided to the constructor,
// - 'value' > 10^kMaxFixedDigitsBeforePoint, or
// - 'requested_digits' > kMaxFixedDigitsAfterPoint.
// The last two conditions imply that the result will never contain more than
// 1 + kMaxFixedDigitsBeforePoint + 1 + kMaxFixedDigitsAfterPoint characters
// (one additional character for the sign, and one for the decimal point).
bool ToFixed(double value, int requested_digits, StringBuilder * result_builder) const;
// Computes a representation in exponential format with requested_digits
// after the decimal point. The last emitted digit is rounded.
// If requested_digits equals -1, then the shortest exponential representation
// is computed.
//
// Examples with EMIT_POSITIVE_EXPONENT_SIGN deactivated, and
// exponent_character set to 'e'.
// ToExponential(3.12, 1) -> "3.1e0"
// ToExponential(5.0, 3) -> "5.000e0"
// ToExponential(0.001, 2) -> "1.00e-3"
// ToExponential(3.1415, -1) -> "3.1415e0"
// ToExponential(3.1415, 4) -> "3.1415e0"
// ToExponential(3.1415, 3) -> "3.142e0"
// ToExponential(123456789000000, 3) -> "1.235e14"
// ToExponential(1000000000000000019884624838656.0, -1) -> "1e30"
// ToExponential(1000000000000000019884624838656.0, 32) ->
// "1.00000000000000001988462483865600e30"
// ToExponential(1234, 0) -> "1e3"
//
// Returns true if the conversion succeeds. The conversion always succeeds
// except for the following cases:
// - the input value is special and no infinity_symbol or nan_symbol has
// been provided to the constructor,
// - 'requested_digits' > kMaxExponentialDigits.
// The last condition implies that the result will never contain more than
// kMaxExponentialDigits + 8 characters (the sign, the digit before the
// decimal point, the decimal point, the exponent character, the
// exponent's sign, and at most 3 exponent digits).
bool ToExponential(double value, int requested_digits, StringBuilder * result_builder) const;
// Computes 'precision' leading digits of the given 'value' and returns them
// either in exponential or decimal format, depending on
// max_{leading|trailing}_padding_zeroes_in_precision_mode (given to the
// constructor).
// The last computed digit is rounded.
//
// Example with max_leading_padding_zeroes_in_precision_mode = 6.
// ToPrecision(0.0000012345, 2) -> "0.0000012"
// ToPrecision(0.00000012345, 2) -> "1.2e-7"
// Similarly the converter may add up to
// max_trailing_padding_zeroes_in_precision_mode in precision mode to avoid
// returning an exponential representation. A zero added by the
// EMIT_TRAILING_ZERO_AFTER_POINT flag is counted for this limit.
// Examples for max_trailing_padding_zeroes_in_precision_mode = 1:
// ToPrecision(230.0, 2) -> "230"
// ToPrecision(230.0, 2) -> "230." with EMIT_TRAILING_DECIMAL_POINT.
// ToPrecision(230.0, 2) -> "2.3e2" with EMIT_TRAILING_ZERO_AFTER_POINT.
// Examples for max_trailing_padding_zeroes_in_precision_mode = 3, and no
// EMIT_TRAILING_ZERO_AFTER_POINT:
// ToPrecision(123450.0, 6) -> "123450"
// ToPrecision(123450.0, 5) -> "123450"
// ToPrecision(123450.0, 4) -> "123500"
// ToPrecision(123450.0, 3) -> "123000"
// ToPrecision(123450.0, 2) -> "1.2e5"
//
// Returns true if the conversion succeeds. The conversion always succeeds
// except for the following cases:
// - the input value is special and no infinity_symbol or nan_symbol has
// been provided to the constructor,
// - precision < kMinPericisionDigits
// - precision > kMaxPrecisionDigits
// The last condition implies that the result will never contain more than
// kMaxPrecisionDigits + 7 characters (the sign, the decimal point, the
// exponent character, the exponent's sign, and at most 3 exponent digits).
bool ToPrecision(double value, int precision, StringBuilder * result_builder) const;
enum DtoaMode
{
// Produce the shortest correct representation.
// For example the output of 0.299999999999999988897 is (the less accurate
// but correct) 0.3.
SHORTEST,
// Same as SHORTEST, but for single-precision floats.
SHORTEST_SINGLE,
// Produce a fixed number of digits after the decimal point.
// For instance fixed(0.1, 4) becomes 0.1000
// If the input number is big, the output will be big.
FIXED,
// Fixed number of digits (independent of the decimal point).
PRECISION
};
// The maximal number of digits that are needed to emit a double in base 10.
// A higher precision can be achieved by using more digits, but the shortest
// accurate representation of any double will never use more digits than
// kBase10MaximalLength.
// Note that DoubleToAscii null-terminates its input. So the given buffer
// should be at least kBase10MaximalLength + 1 characters long.
static const int kBase10MaximalLength = 17;
// Converts the given double 'v' to ascii. 'v' must not be NaN, +Infinity, or
// -Infinity. In SHORTEST_SINGLE-mode this restriction also applies to 'v'
// after it has been casted to a single-precision float. That is, in this
// mode static_cast<float>(v) must not be NaN, +Infinity or -Infinity.
//
// The result should be interpreted as buffer * 10^(point-length).
//
// The output depends on the given mode:
// - SHORTEST: produce the least amount of digits for which the internal
// identity requirement is still satisfied. If the digits are printed
// (together with the correct exponent) then reading this number will give
// 'v' again. The buffer will choose the representation that is closest to
// 'v'. If there are two at the same distance, than the one farther away
// from 0 is chosen (halfway cases - ending with 5 - are rounded up).
// In this mode the 'requested_digits' parameter is ignored.
// - SHORTEST_SINGLE: same as SHORTEST but with single-precision.
// - FIXED: produces digits necessary to print a given number with
// 'requested_digits' digits after the decimal point. The produced digits
// might be too short in which case the caller has to fill the remainder
// with '0's.
// Example: toFixed(0.001, 5) is allowed to return buffer="1", point=-2.
// Halfway cases are rounded towards +/-Infinity (away from 0). The call
// toFixed(0.15, 2) thus returns buffer="2", point=0.
// The returned buffer may contain digits that would be truncated from the
// shortest representation of the input.
// - PRECISION: produces 'requested_digits' where the first digit is not '0'.
// Even though the length of produced digits usually equals
// 'requested_digits', the function is allowed to return fewer digits, in
// which case the caller has to fill the missing digits with '0's.
// Halfway cases are again rounded away from 0.
// DoubleToAscii expects the given buffer to be big enough to hold all
// digits and a terminating null-character. In SHORTEST-mode it expects a
// buffer of at least kBase10MaximalLength + 1. In all other modes the
// requested_digits parameter and the padding-zeroes limit the size of the
// output. Don't forget the decimal point, the exponent character and the
// terminating null-character when computing the maximal output size.
// The given length is only used in debug mode to ensure the buffer is big
// enough.
static void
DoubleToAscii(double v, DtoaMode mode, int requested_digits, char * buffer, int buffer_length, bool * sign, int * length, int * point);
private:
// Implementation for ToShortest and ToShortestSingle.
bool ToShortestIeeeNumber(double value, StringBuilder * result_builder, DtoaMode mode) const;
// If the value is a special value (NaN or Infinity) constructs the
// corresponding string using the configured infinity/nan-symbol.
// If either of them is NULL or the value is not special then the
// function returns false.
bool HandleSpecialValues(double value, StringBuilder * result_builder) const;
// Constructs an exponential representation (i.e. 1.234e56).
// The given exponent assumes a decimal point after the first decimal digit.
void CreateExponentialRepresentation(const char * decimal_digits, int length, int exponent, StringBuilder * result_builder) const;
// Creates a decimal representation (i.e 1234.5678).
void CreateDecimalRepresentation(
const char * decimal_digits, int length, int decimal_point, int digits_after_point, StringBuilder * result_builder) const;
const int flags_;
const char * const infinity_symbol_;
const char * const nan_symbol_;
const char exponent_character_;
const int decimal_in_shortest_low_;
const int decimal_in_shortest_high_;
const int max_leading_padding_zeroes_in_precision_mode_;
const int max_trailing_padding_zeroes_in_precision_mode_;
DISALLOW_IMPLICIT_CONSTRUCTORS(DoubleToStringConverter);
};
class StringToDoubleConverter
{
public:
// Enumeration for allowing octals and ignoring junk when converting
// strings to numbers.
enum Flags
{
NO_FLAGS = 0,
ALLOW_HEX = 1,
ALLOW_OCTALS = 2,
ALLOW_TRAILING_JUNK = 4,
ALLOW_LEADING_SPACES = 8,
ALLOW_TRAILING_SPACES = 16,
ALLOW_SPACES_AFTER_SIGN = 32
};
// Flags should be a bit-or combination of the possible Flags-enum.
// - NO_FLAGS: no special flags.
// - ALLOW_HEX: recognizes the prefix "0x". Hex numbers may only be integers.
// Ex: StringToDouble("0x1234") -> 4660.0
// In StringToDouble("0x1234.56") the characters ".56" are trailing
// junk. The result of the call is hence dependent on
// the ALLOW_TRAILING_JUNK flag and/or the junk value.
// With this flag "0x" is a junk-string. Even with ALLOW_TRAILING_JUNK,
// the string will not be parsed as "0" followed by junk.
//
// - ALLOW_OCTALS: recognizes the prefix "0" for octals:
// If a sequence of octal digits starts with '0', then the number is
// read as octal integer. Octal numbers may only be integers.
// Ex: StringToDouble("01234") -> 668.0
// StringToDouble("012349") -> 12349.0 // Not a sequence of octal
// // digits.
// In StringToDouble("01234.56") the characters ".56" are trailing
// junk. The result of the call is hence dependent on
// the ALLOW_TRAILING_JUNK flag and/or the junk value.
// In StringToDouble("01234e56") the characters "e56" are trailing
// junk, too.
// - ALLOW_TRAILING_JUNK: ignore trailing characters that are not part of
// a double literal.
// - ALLOW_LEADING_SPACES: skip over leading spaces.
// - ALLOW_TRAILING_SPACES: ignore trailing spaces.
// - ALLOW_SPACES_AFTER_SIGN: ignore spaces after the sign.
// Ex: StringToDouble("- 123.2") -> -123.2.
// StringToDouble("+ 123.2") -> 123.2
//
// empty_string_value is returned when an empty string is given as input.
// If ALLOW_LEADING_SPACES or ALLOW_TRAILING_SPACES are set, then a string
// containing only spaces is converted to the 'empty_string_value', too.
//
// junk_string_value is returned when
// a) ALLOW_TRAILING_JUNK is not set, and a junk character (a character not
// part of a double-literal) is found.
// b) ALLOW_TRAILING_JUNK is set, but the string does not start with a
// double literal.
//
// infinity_symbol and nan_symbol are strings that are used to detect
// inputs that represent infinity and NaN. They can be null, in which case
// they are ignored.
// The conversion routine first reads any possible signs. Then it compares the
// following character of the input-string with the first character of
// the infinity, and nan-symbol. If either matches, the function assumes, that
// a match has been found, and expects the following input characters to match
// the remaining characters of the special-value symbol.
// This means that the following restrictions apply to special-value symbols:
// - they must not start with signs ('+', or '-'),
// - they must not have the same first character.
// - they must not start with digits.
//
// Examples:
// flags = ALLOW_HEX | ALLOW_TRAILING_JUNK,
// empty_string_value = 0.0,
// junk_string_value = NaN,
// infinity_symbol = "infinity",
// nan_symbol = "nan":
// StringToDouble("0x1234") -> 4660.0.
// StringToDouble("0x1234K") -> 4660.0.
// StringToDouble("") -> 0.0 // empty_string_value.
// StringToDouble(" ") -> NaN // junk_string_value.
// StringToDouble(" 1") -> NaN // junk_string_value.
// StringToDouble("0x") -> NaN // junk_string_value.
// StringToDouble("-123.45") -> -123.45.
// StringToDouble("--123.45") -> NaN // junk_string_value.
// StringToDouble("123e45") -> 123e45.
// StringToDouble("123E45") -> 123e45.
// StringToDouble("123e+45") -> 123e45.
// StringToDouble("123E-45") -> 123e-45.
// StringToDouble("123e") -> 123.0 // trailing junk ignored.
// StringToDouble("123e-") -> 123.0 // trailing junk ignored.
// StringToDouble("+NaN") -> NaN // NaN string literal.
// StringToDouble("-infinity") -> -inf. // infinity literal.
// StringToDouble("Infinity") -> NaN // junk_string_value.
//
// flags = ALLOW_OCTAL | ALLOW_LEADING_SPACES,
// empty_string_value = 0.0,
// junk_string_value = NaN,
// infinity_symbol = NULL,
// nan_symbol = NULL:
// StringToDouble("0x1234") -> NaN // junk_string_value.
// StringToDouble("01234") -> 668.0.
// StringToDouble("") -> 0.0 // empty_string_value.
// StringToDouble(" ") -> 0.0 // empty_string_value.
// StringToDouble(" 1") -> 1.0
// StringToDouble("0x") -> NaN // junk_string_value.
// StringToDouble("0123e45") -> NaN // junk_string_value.
// StringToDouble("01239E45") -> 1239e45.
// StringToDouble("-infinity") -> NaN // junk_string_value.
// StringToDouble("NaN") -> NaN // junk_string_value.
StringToDoubleConverter(
int flags, double empty_string_value, double junk_string_value, const char * infinity_symbol, const char * nan_symbol)
: flags_(flags)
, empty_string_value_(empty_string_value)
, junk_string_value_(junk_string_value)
, infinity_symbol_(infinity_symbol)
, nan_symbol_(nan_symbol)
{
}
// Performs the conversion.
// The output parameter 'processed_characters_count' is set to the number
// of characters that have been processed to read the number.
// Spaces than are processed with ALLOW_{LEADING|TRAILING}_SPACES are included
// in the 'processed_characters_count'. Trailing junk is never included.
double StringToDouble(const char * buffer, int length, int * processed_characters_count) const
{
return StringToIeee(buffer, length, processed_characters_count, true);
}
// Same as StringToDouble but reads a float.
// Note that this is not equivalent to static_cast<float>(StringToDouble(...))
// due to potential double-rounding.
float StringToFloat(const char * buffer, int length, int * processed_characters_count) const
{
return static_cast<float>(StringToIeee(buffer, length, processed_characters_count, false));
}
private:
const int flags_;
const double empty_string_value_;
const double junk_string_value_;
const char * const infinity_symbol_;
const char * const nan_symbol_;
double StringToIeee(const char * buffer, int length, int * processed_characters_count, bool read_as_double) const;
DISALLOW_IMPLICIT_CONSTRUCTORS(StringToDoubleConverter);
};
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_DOUBLE_CONVERSION_H_

665
base/poco/Foundation/src/fast-dtoa.cc
View File

@ -1,665 +0,0 @@
// Copyright 2012 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include "fast-dtoa.h"
#include "cached-powers.h"
#include "diy-fp.h"
#include "ieee.h"
namespace double_conversion {
// The minimal and maximal target exponent define the range of w's binary
// exponent, where 'w' is the result of multiplying the input by a cached power
// of ten.
//
// A different range might be chosen on a different platform, to optimize digit
// generation, but a smaller range requires more powers of ten to be cached.
static const int kMinimalTargetExponent = -60;
static const int kMaximalTargetExponent = -32;
// Adjusts the last digit of the generated number, and screens out generated
// solutions that may be inaccurate. A solution may be inaccurate if it is
// outside the safe interval, or if we cannot prove that it is closer to the
// input than a neighboring representation of the same length.
//
// Input: * buffer containing the digits of too_high / 10^kappa
// * the buffer's length
// * distance_too_high_w == (too_high - w).f() * unit
// * unsafe_interval == (too_high - too_low).f() * unit
// * rest = (too_high - buffer * 10^kappa).f() * unit
// * ten_kappa = 10^kappa * unit
// * unit = the common multiplier
// Output: returns true if the buffer is guaranteed to contain the closest
// representable number to the input.
// Modifies the generated digits in the buffer to approach (round towards) w.
static bool RoundWeed(Vector<char> buffer,
int length,
uint64_t distance_too_high_w,
uint64_t unsafe_interval,
uint64_t rest,
uint64_t ten_kappa,
uint64_t unit) {
uint64_t small_distance = distance_too_high_w - unit;
uint64_t big_distance = distance_too_high_w + unit;
// Let w_low = too_high - big_distance, and
// w_high = too_high - small_distance.
// Note: w_low < w < w_high
//
// The real w (* unit) must lie somewhere inside the interval
// ]w_low; w_high[ (often written as "(w_low; w_high)")
// Basically the buffer currently contains a number in the unsafe interval
// ]too_low; too_high[ with too_low < w < too_high
//
// too_high - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
// ^v 1 unit ^ ^ ^ ^
// boundary_high --------------------- . . . .
// ^v 1 unit . . . .
// - - - - - - - - - - - - - - - - - - - + - - + - - - - - - . .
// . . ^ . .
// . big_distance . . .
// . . . . rest
// small_distance . . . .
// v . . . .
// w_high - - - - - - - - - - - - - - - - - - . . . .
// ^v 1 unit . . . .
// w ---------------------------------------- . . . .
// ^v 1 unit v . . .
// w_low - - - - - - - - - - - - - - - - - - - - - . . .
// . . v
// buffer --------------------------------------------------+-------+--------
// . .
// safe_interval .
// v .
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - .
// ^v 1 unit .
// boundary_low ------------------------- unsafe_interval
// ^v 1 unit v
// too_low - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
//
//
// Note that the value of buffer could lie anywhere inside the range too_low
// to too_high.
//
// boundary_low, boundary_high and w are approximations of the real boundaries
// and v (the input number). They are guaranteed to be precise up to one unit.
// In fact the error is guaranteed to be strictly less than one unit.
//
// Anything that lies outside the unsafe interval is guaranteed not to round
// to v when read again.
// Anything that lies inside the safe interval is guaranteed to round to v
// when read again.
// If the number inside the buffer lies inside the unsafe interval but not
// inside the safe interval then we simply do not know and bail out (returning
// false).
//
// Similarly we have to take into account the imprecision of 'w' when finding
// the closest representation of 'w'. If we have two potential
// representations, and one is closer to both w_low and w_high, then we know
// it is closer to the actual value v.
//
// By generating the digits of too_high we got the largest (closest to
// too_high) buffer that is still in the unsafe interval. In the case where
// w_high < buffer < too_high we try to decrement the buffer.
// This way the buffer approaches (rounds towards) w.
// There are 3 conditions that stop the decrementation process:
// 1) the buffer is already below w_high
// 2) decrementing the buffer would make it leave the unsafe interval
// 3) decrementing the buffer would yield a number below w_high and farther
// away than the current number. In other words:
// (buffer{-1} < w_high) && w_high - buffer{-1} > buffer - w_high
// Instead of using the buffer directly we use its distance to too_high.
// Conceptually rest ~= too_high - buffer
// We need to do the following tests in this order to avoid over- and
// underflows.
ASSERT(rest <= unsafe_interval);
while (rest < small_distance && // Negated condition 1
unsafe_interval - rest >= ten_kappa && // Negated condition 2
(rest + ten_kappa < small_distance || // buffer{-1} > w_high
small_distance - rest >= rest + ten_kappa - small_distance)) {
buffer[length - 1]--;
rest += ten_kappa;
}
// We have approached w+ as much as possible. We now test if approaching w-
// would require changing the buffer. If yes, then we have two possible
// representations close to w, but we cannot decide which one is closer.
if (rest < big_distance &&
unsafe_interval - rest >= ten_kappa &&
(rest + ten_kappa < big_distance ||
big_distance - rest > rest + ten_kappa - big_distance)) {
return false;
}
// Weeding test.
// The safe interval is [too_low + 2 ulp; too_high - 2 ulp]
// Since too_low = too_high - unsafe_interval this is equivalent to
// [too_high - unsafe_interval + 4 ulp; too_high - 2 ulp]
// Conceptually we have: rest ~= too_high - buffer
return (2 * unit <= rest) && (rest <= unsafe_interval - 4 * unit);
}
// Rounds the buffer upwards if the result is closer to v by possibly adding
// 1 to the buffer. If the precision of the calculation is not sufficient to
// round correctly, return false.
// The rounding might shift the whole buffer in which case the kappa is
// adjusted. For example "99", kappa = 3 might become "10", kappa = 4.
//
// If 2*rest > ten_kappa then the buffer needs to be round up.
// rest can have an error of +/- 1 unit. This function accounts for the
// imprecision and returns false, if the rounding direction cannot be
// unambiguously determined.
//
// Precondition: rest < ten_kappa.
static bool RoundWeedCounted(Vector<char> buffer,
int length,
uint64_t rest,
uint64_t ten_kappa,
uint64_t unit,
int* kappa) {
ASSERT(rest < ten_kappa);
// The following tests are done in a specific order to avoid overflows. They
// will work correctly with any uint64 values of rest < ten_kappa and unit.
//
// If the unit is too big, then we don't know which way to round. For example
// a unit of 50 means that the real number lies within rest +/- 50. If
// 10^kappa == 40 then there is no way to tell which way to round.
if (unit >= ten_kappa) return false;
// Even if unit is just half the size of 10^kappa we are already completely
// lost. (And after the previous test we know that the expression will not
// over/underflow.)
if (ten_kappa - unit <= unit) return false;
// If 2 * (rest + unit) <= 10^kappa we can safely round down.
if ((ten_kappa - rest > rest) && (ten_kappa - 2 * rest >= 2 * unit)) {
return true;
}
// If 2 * (rest - unit) >= 10^kappa, then we can safely round up.
if ((rest > unit) && (ten_kappa - (rest - unit) <= (rest - unit))) {
// Increment the last digit recursively until we find a non '9' digit.
buffer[length - 1]++;
for (int i = length - 1; i > 0; --i) {
if (buffer[i] != '0' + 10) break;
buffer[i] = '0';
buffer[i - 1]++;
}
// If the first digit is now '0'+ 10 we had a buffer with all '9's. With the
// exception of the first digit all digits are now '0'. Simply switch the
// first digit to '1' and adjust the kappa. Example: "99" becomes "10" and
// the power (the kappa) is increased.
if (buffer[0] == '0' + 10) {
buffer[0] = '1';
(*kappa) += 1;
}
return true;
}
return false;
}
// Returns the biggest power of ten that is less than or equal to the given
// number. We furthermore receive the maximum number of bits 'number' has.
//
// Returns power == 10^(exponent_plus_one-1) such that
// power <= number < power * 10.
// If number_bits == 0 then 0^(0-1) is returned.
// The number of bits must be <= 32.
// Precondition: number < (1 << (number_bits + 1)).
// Inspired by the method for finding an integer log base 10 from here:
// http://graphics.stanford.edu/~seander/bithacks.html#IntegerLog10
static unsigned int const kSmallPowersOfTen[] =
{0, 1, 10, 100, 1000, 10000, 100000, 1000000, 10000000, 100000000,
1000000000};
static void BiggestPowerTen(uint32_t number,
int number_bits,
uint32_t* power,
int* exponent_plus_one) {
ASSERT(number < (1u << (number_bits + 1)));
// 1233/4096 is approximately 1/lg(10).
int exponent_plus_one_guess = ((number_bits + 1) * 1233 >> 12);
// We increment to skip over the first entry in the kPowersOf10 table.
// Note: kPowersOf10[i] == 10^(i-1).
exponent_plus_one_guess++;
// We don't have any guarantees that 2^number_bits <= number.
if (number < kSmallPowersOfTen[exponent_plus_one_guess] && exponent_plus_one_guess > 0) {
exponent_plus_one_guess--;
}
*power = kSmallPowersOfTen[exponent_plus_one_guess];
*exponent_plus_one = exponent_plus_one_guess;
}
// Generates the digits of input number w.
// w is a floating-point number (DiyFp), consisting of a significand and an
// exponent. Its exponent is bounded by kMinimalTargetExponent and
// kMaximalTargetExponent.
// Hence -60 <= w.e() <= -32.
//
// Returns false if it fails, in which case the generated digits in the buffer
// should not be used.
// Preconditions:
// * low, w and high are correct up to 1 ulp (unit in the last place). That
// is, their error must be less than a unit of their last digits.
// * low.e() == w.e() == high.e()
// * low < w < high, and taking into account their error: low~ <= high~
// * kMinimalTargetExponent <= w.e() <= kMaximalTargetExponent
// Postconditions: returns false if procedure fails.
// otherwise:
// * buffer is not null-terminated, but len contains the number of digits.
// * buffer contains the shortest possible decimal digit-sequence
// such that LOW < buffer * 10^kappa < HIGH, where LOW and HIGH are the
// correct values of low and high (without their error).
// * if more than one decimal representation gives the minimal number of
// decimal digits then the one closest to W (where W is the correct value
// of w) is chosen.
// Remark: this procedure takes into account the imprecision of its input
// numbers. If the precision is not enough to guarantee all the postconditions
// then false is returned. This usually happens rarely (~0.5%).
//
// Say, for the sake of example, that
// w.e() == -48, and w.f() == 0x1234567890abcdef
// w's value can be computed by w.f() * 2^w.e()
// We can obtain w's integral digits by simply shifting w.f() by -w.e().
// -> w's integral part is 0x1234
// w's fractional part is therefore 0x567890abcdef.
// Printing w's integral part is easy (simply print 0x1234 in decimal).
// In order to print its fraction we repeatedly multiply the fraction by 10 and
// get each digit. Example the first digit after the point would be computed by
// (0x567890abcdef * 10) >> 48. -> 3
// The whole thing becomes slightly more complicated because we want to stop
// once we have enough digits. That is, once the digits inside the buffer
// represent 'w' we can stop. Everything inside the interval low - high
// represents w. However we have to pay attention to low, high and w's
// imprecision.
static bool DigitGen(DiyFp low,
DiyFp w,
DiyFp high,
Vector<char> buffer,
int* length,
int* kappa) {
ASSERT(low.e() == w.e() && w.e() == high.e());
ASSERT(low.f() + 1 <= high.f() - 1);
ASSERT(kMinimalTargetExponent <= w.e() && w.e() <= kMaximalTargetExponent);
// low, w and high are imprecise, but by less than one ulp (unit in the last
// place).
// If we remove (resp. add) 1 ulp from low (resp. high) we are certain that
// the new numbers are outside of the interval we want the final
// representation to lie in.
// Inversely adding (resp. removing) 1 ulp from low (resp. high) would yield
// numbers that are certain to lie in the interval. We will use this fact
// later on.
// We will now start by generating the digits within the uncertain
// interval. Later we will weed out representations that lie outside the safe
// interval and thus _might_ lie outside the correct interval.
uint64_t unit = 1;
DiyFp too_low = DiyFp(low.f() - unit, low.e());
DiyFp too_high = DiyFp(high.f() + unit, high.e());
// too_low and too_high are guaranteed to lie outside the interval we want the
// generated number in.
DiyFp unsafe_interval = DiyFp::Minus(too_high, too_low);
// We now cut the input number into two parts: the integral digits and the
// fractionals. We will not write any decimal separator though, but adapt
// kappa instead.
// Reminder: we are currently computing the digits (stored inside the buffer)
// such that: too_low < buffer * 10^kappa < too_high
// We use too_high for the digit_generation and stop as soon as possible.
// If we stop early we effectively round down.
DiyFp one = DiyFp(static_cast<uint64_t>(1) << -w.e(), w.e());
// Division by one is a shift.
uint32_t integrals = static_cast<uint32_t>(too_high.f() >> -one.e());
// Modulo by one is an and.
uint64_t fractionals = too_high.f() & (one.f() - 1);
uint32_t divisor;
int divisor_exponent_plus_one;
BiggestPowerTen(integrals, DiyFp::kSignificandSize - (-one.e()),
&divisor, &divisor_exponent_plus_one);
*kappa = divisor_exponent_plus_one;
*length = 0;
// Loop invariant: buffer = too_high / 10^kappa (integer division)
// The invariant holds for the first iteration: kappa has been initialized
// with the divisor exponent + 1. And the divisor is the biggest power of ten
// that is smaller than integrals.
while (*kappa > 0) {
int digit = integrals / divisor;
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
integrals %= divisor;
(*kappa)--;
// Note that kappa now equals the exponent of the divisor and that the
// invariant thus holds again.
uint64_t rest =
(static_cast<uint64_t>(integrals) << -one.e()) + fractionals;
// Invariant: too_high = buffer * 10^kappa + DiyFp(rest, one.e())
// Reminder: unsafe_interval.e() == one.e()
if (rest < unsafe_interval.f()) {
// Rounding down (by not emitting the remaining digits) yields a number
// that lies within the unsafe interval.
return RoundWeed(buffer, *length, DiyFp::Minus(too_high, w).f(),
unsafe_interval.f(), rest,
static_cast<uint64_t>(divisor) << -one.e(), unit);
}
divisor /= 10;
}
// The integrals have been generated. We are at the point of the decimal
// separator. In the following loop we simply multiply the remaining digits by
// 10 and divide by one. We just need to pay attention to multiply associated
// data (like the interval or 'unit'), too.
// Note that the multiplication by 10 does not overflow, because w.e >= -60
// and thus one.e >= -60.
ASSERT(one.e() >= -60);
ASSERT(fractionals < one.f());
ASSERT(UINT64_2PART_C(0xFFFFFFFF, FFFFFFFF) / 10 >= one.f());
for (;;) {
fractionals *= 10;
unit *= 10;
unsafe_interval.set_f(unsafe_interval.f() * 10);
// Integer division by one.
int digit = static_cast<int>(fractionals >> -one.e());
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
fractionals &= one.f() - 1; // Modulo by one.
(*kappa)--;
if (fractionals < unsafe_interval.f()) {
return RoundWeed(buffer, *length, DiyFp::Minus(too_high, w).f() * unit,
unsafe_interval.f(), fractionals, one.f(), unit);
}
}
}
// Generates (at most) requested_digits digits of input number w.
// w is a floating-point number (DiyFp), consisting of a significand and an
// exponent. Its exponent is bounded by kMinimalTargetExponent and
// kMaximalTargetExponent.
// Hence -60 <= w.e() <= -32.
//
// Returns false if it fails, in which case the generated digits in the buffer
// should not be used.
// Preconditions:
// * w is correct up to 1 ulp (unit in the last place). That
// is, its error must be strictly less than a unit of its last digit.
// * kMinimalTargetExponent <= w.e() <= kMaximalTargetExponent
//
// Postconditions: returns false if procedure fails.
// otherwise:
// * buffer is not null-terminated, but length contains the number of
// digits.
// * the representation in buffer is the most precise representation of
// requested_digits digits.
// * buffer contains at most requested_digits digits of w. If there are less
// than requested_digits digits then some trailing '0's have been removed.
// * kappa is such that
// w = buffer * 10^kappa + eps with |eps| < 10^kappa / 2.
//
// Remark: This procedure takes into account the imprecision of its input
// numbers. If the precision is not enough to guarantee all the postconditions
// then false is returned. This usually happens rarely, but the failure-rate
// increases with higher requested_digits.
static bool DigitGenCounted(DiyFp w,
int requested_digits,
Vector<char> buffer,
int* length,
int* kappa) {
ASSERT(kMinimalTargetExponent <= w.e() && w.e() <= kMaximalTargetExponent);
ASSERT(kMinimalTargetExponent >= -60);
ASSERT(kMaximalTargetExponent <= -32);
// w is assumed to have an error less than 1 unit. Whenever w is scaled we
// also scale its error.
uint64_t w_error = 1;
// We cut the input number into two parts: the integral digits and the
// fractional digits. We don't emit any decimal separator, but adapt kappa
// instead. Example: instead of writing "1.2" we put "12" into the buffer and
// increase kappa by 1.
DiyFp one = DiyFp(static_cast<uint64_t>(1) << -w.e(), w.e());
// Division by one is a shift.
uint32_t integrals = static_cast<uint32_t>(w.f() >> -one.e());
// Modulo by one is an and.
uint64_t fractionals = w.f() & (one.f() - 1);
uint32_t divisor;
int divisor_exponent_plus_one;
BiggestPowerTen(integrals, DiyFp::kSignificandSize - (-one.e()),
&divisor, &divisor_exponent_plus_one);
*kappa = divisor_exponent_plus_one;
*length = 0;
// Loop invariant: buffer = w / 10^kappa (integer division)
// The invariant holds for the first iteration: kappa has been initialized
// with the divisor exponent + 1. And the divisor is the biggest power of ten
// that is smaller than 'integrals'.
while (*kappa > 0) {
int digit = integrals / divisor;
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
requested_digits--;
integrals %= divisor;
(*kappa)--;
// Note that kappa now equals the exponent of the divisor and that the
// invariant thus holds again.
if (requested_digits == 0) break;
divisor /= 10;
}
if (requested_digits == 0) {
uint64_t rest =
(static_cast<uint64_t>(integrals) << -one.e()) + fractionals;
return RoundWeedCounted(buffer, *length, rest,
static_cast<uint64_t>(divisor) << -one.e(), w_error,
kappa);
}
// The integrals have been generated. We are at the point of the decimal
// separator. In the following loop we simply multiply the remaining digits by
// 10 and divide by one. We just need to pay attention to multiply associated
// data (the 'unit'), too.
// Note that the multiplication by 10 does not overflow, because w.e >= -60
// and thus one.e >= -60.
ASSERT(one.e() >= -60);
ASSERT(fractionals < one.f());
ASSERT(UINT64_2PART_C(0xFFFFFFFF, FFFFFFFF) / 10 >= one.f());
while (requested_digits > 0 && fractionals > w_error) {
fractionals *= 10;
w_error *= 10;
// Integer division by one.
int digit = static_cast<int>(fractionals >> -one.e());
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
requested_digits--;
fractionals &= one.f() - 1; // Modulo by one.
(*kappa)--;
}
if (requested_digits != 0) return false;
return RoundWeedCounted(buffer, *length, fractionals, one.f(), w_error,
kappa);
}
// Provides a decimal representation of v.
// Returns true if it succeeds, otherwise the result cannot be trusted.
// There will be *length digits inside the buffer (not null-terminated).
// If the function returns true then
// v == (double) (buffer * 10^decimal_exponent).
// The digits in the buffer are the shortest representation possible: no
// 0.09999999999999999 instead of 0.1. The shorter representation will even be
// chosen even if the longer one would be closer to v.
// The last digit will be closest to the actual v. That is, even if several
// digits might correctly yield 'v' when read again, the closest will be
// computed.
static bool Grisu3(double v,
FastDtoaMode mode,
Vector<char> buffer,
int* length,
int* decimal_exponent) {
DiyFp w = Double(v).AsNormalizedDiyFp();
// boundary_minus and boundary_plus are the boundaries between v and its
// closest floating-point neighbors. Any number strictly between
// boundary_minus and boundary_plus will round to v when convert to a double.
// Grisu3 will never output representations that lie exactly on a boundary.
DiyFp boundary_minus, boundary_plus;
if (mode == FAST_DTOA_SHORTEST) {
Double(v).NormalizedBoundaries(&boundary_minus, &boundary_plus);
} else {
ASSERT(mode == FAST_DTOA_SHORTEST_SINGLE);
float single_v = static_cast<float>(v);
Single(single_v).NormalizedBoundaries(&boundary_minus, &boundary_plus);
}
ASSERT(boundary_plus.e() == w.e());
DiyFp ten_mk; // Cached power of ten: 10^-k
int mk; // -k
int ten_mk_minimal_binary_exponent =
kMinimalTargetExponent - (w.e() + DiyFp::kSignificandSize);
int ten_mk_maximal_binary_exponent =
kMaximalTargetExponent - (w.e() + DiyFp::kSignificandSize);
PowersOfTenCache::GetCachedPowerForBinaryExponentRange(
ten_mk_minimal_binary_exponent,
ten_mk_maximal_binary_exponent,
&ten_mk, &mk);
ASSERT((kMinimalTargetExponent <= w.e() + ten_mk.e() +
DiyFp::kSignificandSize) &&
(kMaximalTargetExponent >= w.e() + ten_mk.e() +
DiyFp::kSignificandSize));
// Note that ten_mk is only an approximation of 10^-k. A DiyFp only contains a
// 64 bit significand and ten_mk is thus only precise up to 64 bits.
// The DiyFp::Times procedure rounds its result, and ten_mk is approximated
// too. The variable scaled_w (as well as scaled_boundary_minus/plus) are now
// off by a small amount.
// In fact: scaled_w - w*10^k < 1ulp (unit in the last place) of scaled_w.
// In other words: let f = scaled_w.f() and e = scaled_w.e(), then
// (f-1) * 2^e < w*10^k < (f+1) * 2^e
DiyFp scaled_w = DiyFp::Times(w, ten_mk);
ASSERT(scaled_w.e() ==
boundary_plus.e() + ten_mk.e() + DiyFp::kSignificandSize);
// In theory it would be possible to avoid some recomputations by computing
// the difference between w and boundary_minus/plus (a power of 2) and to
// compute scaled_boundary_minus/plus by subtracting/adding from
// scaled_w. However the code becomes much less readable and the speed
// enhancements are not terrific.
DiyFp scaled_boundary_minus = DiyFp::Times(boundary_minus, ten_mk);
DiyFp scaled_boundary_plus = DiyFp::Times(boundary_plus, ten_mk);
// DigitGen will generate the digits of scaled_w. Therefore we have
// v == (double) (scaled_w * 10^-mk).
// Set decimal_exponent == -mk and pass it to DigitGen. If scaled_w is not an
// integer than it will be updated. For instance if scaled_w == 1.23 then
// the buffer will be filled with "123" und the decimal_exponent will be
// decreased by 2.
int kappa;
bool result = DigitGen(scaled_boundary_minus, scaled_w, scaled_boundary_plus,
buffer, length, &kappa);
*decimal_exponent = -mk + kappa;
return result;
}
// The "counted" version of grisu3 (see above) only generates requested_digits
// number of digits. This version does not generate the shortest representation,
// and with enough requested digits 0.1 will at some point print as 0.9999999...
// Grisu3 is too imprecise for real halfway cases (1.5 will not work) and
// therefore the rounding strategy for halfway cases is irrelevant.
static bool Grisu3Counted(double v,
int requested_digits,
Vector<char> buffer,
int* length,
int* decimal_exponent) {
DiyFp w = Double(v).AsNormalizedDiyFp();
DiyFp ten_mk; // Cached power of ten: 10^-k
int mk; // -k
int ten_mk_minimal_binary_exponent =
kMinimalTargetExponent - (w.e() + DiyFp::kSignificandSize);
int ten_mk_maximal_binary_exponent =
kMaximalTargetExponent - (w.e() + DiyFp::kSignificandSize);
PowersOfTenCache::GetCachedPowerForBinaryExponentRange(
ten_mk_minimal_binary_exponent,
ten_mk_maximal_binary_exponent,
&ten_mk, &mk);
ASSERT((kMinimalTargetExponent <= w.e() + ten_mk.e() +
DiyFp::kSignificandSize) &&
(kMaximalTargetExponent >= w.e() + ten_mk.e() +
DiyFp::kSignificandSize));
// Note that ten_mk is only an approximation of 10^-k. A DiyFp only contains a
// 64 bit significand and ten_mk is thus only precise up to 64 bits.
// The DiyFp::Times procedure rounds its result, and ten_mk is approximated
// too. The variable scaled_w (as well as scaled_boundary_minus/plus) are now
// off by a small amount.
// In fact: scaled_w - w*10^k < 1ulp (unit in the last place) of scaled_w.
// In other words: let f = scaled_w.f() and e = scaled_w.e(), then
// (f-1) * 2^e < w*10^k < (f+1) * 2^e
DiyFp scaled_w = DiyFp::Times(w, ten_mk);
// We now have (double) (scaled_w * 10^-mk).
// DigitGen will generate the first requested_digits digits of scaled_w and
// return together with a kappa such that scaled_w ~= buffer * 10^kappa. (It
// will not always be exactly the same since DigitGenCounted only produces a
// limited number of digits.)
int kappa;
bool result = DigitGenCounted(scaled_w, requested_digits,
buffer, length, &kappa);
*decimal_exponent = -mk + kappa;
return result;
}
bool FastDtoa(double v,
FastDtoaMode mode,
int requested_digits,
Vector<char> buffer,
int* length,
int* decimal_point) {
ASSERT(v > 0);
ASSERT(!Double(v).IsSpecial());
bool result = false;
int decimal_exponent = 0;
switch (mode) {
case FAST_DTOA_SHORTEST:
case FAST_DTOA_SHORTEST_SINGLE:
result = Grisu3(v, mode, buffer, length, &decimal_exponent);
break;
case FAST_DTOA_PRECISION:
result = Grisu3Counted(v, requested_digits,
buffer, length, &decimal_exponent);
break;
default:
UNREACHABLE();
}
if (result) {
*decimal_point = *length + decimal_exponent;
buffer[*length] = '\0';
}
return result;
}
} // namespace double_conversion

85
base/poco/Foundation/src/fast-dtoa.h
View File

@ -1,85 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_FAST_DTOA_H_
#define DOUBLE_CONVERSION_FAST_DTOA_H_
#include "utils.h"
namespace double_conversion
{
enum FastDtoaMode
{
// Computes the shortest representation of the given input. The returned
// result will be the most accurate number of this length. Longer
// representations might be more accurate.
FAST_DTOA_SHORTEST,
// Same as FAST_DTOA_SHORTEST but for single-precision floats.
FAST_DTOA_SHORTEST_SINGLE,
// Computes a representation where the precision (number of digits) is
// given as input. The precision is independent of the decimal point.
FAST_DTOA_PRECISION
};
// FastDtoa will produce at most kFastDtoaMaximalLength digits. This does not
// include the terminating '\0' character.
static const int kFastDtoaMaximalLength = 17;
// Same for single-precision numbers.
static const int kFastDtoaMaximalSingleLength = 9;
// Provides a decimal representation of v.
// The result should be interpreted as buffer * 10^(point - length).
//
// Precondition:
// * v must be a strictly positive finite double.
//
// Returns true if it succeeds, otherwise the result can not be trusted.
// There will be *length digits inside the buffer followed by a null terminator.
// If the function returns true and mode equals
// - FAST_DTOA_SHORTEST, then
// the parameter requested_digits is ignored.
// The result satisfies
// v == (double) (buffer * 10^(point - length)).
// The digits in the buffer are the shortest representation possible. E.g.
// if 0.099999999999 and 0.1 represent the same double then "1" is returned
// with point = 0.
// The last digit will be closest to the actual v. That is, even if several
// digits might correctly yield 'v' when read again, the buffer will contain
// the one closest to v.
// - FAST_DTOA_PRECISION, then
// the buffer contains requested_digits digits.
// the difference v - (buffer * 10^(point-length)) is closest to zero for
// all possible representations of requested_digits digits.
// If there are two values that are equally close, then FastDtoa returns
// false.
// For both modes the buffer must be large enough to hold the result.
bool FastDtoa(double d, FastDtoaMode mode, int requested_digits, Vector<char> buffer, int * length, int * decimal_point);
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_FAST_DTOA_H_

404
base/poco/Foundation/src/fixed-dtoa.cc
View File

@ -1,404 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <math.h>
#include "fixed-dtoa.h"
#include "ieee.h"
namespace double_conversion {
// Represents a 128bit type. This class should be replaced by a native type on
// platforms that support 128bit integers.
class UInt128 {
public:
UInt128() : high_bits_(0), low_bits_(0) { }
UInt128(uint64_t high, uint64_t low) : high_bits_(high), low_bits_(low) { }
void Multiply(uint32_t multiplicand) {
uint64_t accumulator;
accumulator = (low_bits_ & kMask32) * multiplicand;
uint32_t part = static_cast<uint32_t>(accumulator & kMask32);
accumulator >>= 32;
accumulator = accumulator + (low_bits_ >> 32) * multiplicand;
low_bits_ = (accumulator << 32) + part;
accumulator >>= 32;
accumulator = accumulator + (high_bits_ & kMask32) * multiplicand;
part = static_cast<uint32_t>(accumulator & kMask32);
accumulator >>= 32;
accumulator = accumulator + (high_bits_ >> 32) * multiplicand;
high_bits_ = (accumulator << 32) + part;
ASSERT((accumulator >> 32) == 0);
}
void Shift(int shift_amount) {
ASSERT(-64 <= shift_amount && shift_amount <= 64);
if (shift_amount == 0) {
return;
} else if (shift_amount == -64) {
high_bits_ = low_bits_;
low_bits_ = 0;
} else if (shift_amount == 64) {
low_bits_ = high_bits_;
high_bits_ = 0;
} else if (shift_amount <= 0) {
high_bits_ <<= -shift_amount;
high_bits_ += low_bits_ >> (64 + shift_amount);
low_bits_ <<= -shift_amount;
} else {
low_bits_ >>= shift_amount;
low_bits_ += high_bits_ << (64 - shift_amount);
high_bits_ >>= shift_amount;
}
}
// Modifies *this to *this MOD (2^power).
// Returns *this DIV (2^power).
int DivModPowerOf2(int power) {
if (power >= 64) {
int result = static_cast<int>(high_bits_ >> (power - 64));
high_bits_ -= static_cast<uint64_t>(result) << (power - 64);
return result;
} else {
uint64_t part_low = low_bits_ >> power;
uint64_t part_high = high_bits_ << (64 - power);
int result = static_cast<int>(part_low + part_high);
high_bits_ = 0;
low_bits_ -= part_low << power;
return result;
}
}
bool IsZero() const {
return high_bits_ == 0 && low_bits_ == 0;
}
int BitAt(int position) {
if (position >= 64) {
return static_cast<int>(high_bits_ >> (position - 64)) & 1;
} else {
return static_cast<int>(low_bits_ >> position) & 1;
}
}
private:
static const uint64_t kMask32 = 0xFFFFFFFF;
// Value == (high_bits_ << 64) + low_bits_
uint64_t high_bits_;
uint64_t low_bits_;
};
static const int kDoubleSignificandSize = 53; // Includes the hidden bit.
static void FillDigits32FixedLength(uint32_t number, int requested_length,
Vector<char> buffer, int* length) {
for (int i = requested_length - 1; i >= 0; --i) {
buffer[(*length) + i] = '0' + number % 10;
number /= 10;
}
*length += requested_length;
}
static void FillDigits32(uint32_t number, Vector<char> buffer, int* length) {
int number_length = 0;
// We fill the digits in reverse order and exchange them afterwards.
while (number != 0) {
int digit = number % 10;
number /= 10;
buffer[(*length) + number_length] = static_cast<char>('0' + digit);
number_length++;
}
// Exchange the digits.
int i = *length;
int j = *length + number_length - 1;
while (i < j) {
char tmp = buffer[i];
buffer[i] = buffer[j];
buffer[j] = tmp;
i++;
j--;
}
*length += number_length;
}
static void FillDigits64FixedLength(uint64_t number,
Vector<char> buffer, int* length) {
const uint32_t kTen7 = 10000000;
// For efficiency cut the number into 3 uint32_t parts, and print those.
uint32_t part2 = static_cast<uint32_t>(number % kTen7);
number /= kTen7;
uint32_t part1 = static_cast<uint32_t>(number % kTen7);
uint32_t part0 = static_cast<uint32_t>(number / kTen7);
FillDigits32FixedLength(part0, 3, buffer, length);
FillDigits32FixedLength(part1, 7, buffer, length);
FillDigits32FixedLength(part2, 7, buffer, length);
}
static void FillDigits64(uint64_t number, Vector<char> buffer, int* length) {
const uint32_t kTen7 = 10000000;
// For efficiency cut the number into 3 uint32_t parts, and print those.
uint32_t part2 = static_cast<uint32_t>(number % kTen7);
number /= kTen7;
uint32_t part1 = static_cast<uint32_t>(number % kTen7);
uint32_t part0 = static_cast<uint32_t>(number / kTen7);
if (part0 != 0) {
FillDigits32(part0, buffer, length);
FillDigits32FixedLength(part1, 7, buffer, length);
FillDigits32FixedLength(part2, 7, buffer, length);
} else if (part1 != 0) {
FillDigits32(part1, buffer, length);
FillDigits32FixedLength(part2, 7, buffer, length);
} else {
FillDigits32(part2, buffer, length);
}
}
static void RoundUp(Vector<char> buffer, int* length, int* decimal_point) {
// An empty buffer represents 0.
if (*length == 0) {
buffer[0] = '1';
*decimal_point = 1;
*length = 1;
return;
}
// Round the last digit until we either have a digit that was not '9' or until
// we reached the first digit.
buffer[(*length) - 1]++;
for (int i = (*length) - 1; i > 0; --i) {
if (buffer[i] != '0' + 10) {
return;
}
buffer[i] = '0';
buffer[i - 1]++;
}
// If the first digit is now '0' + 10, we would need to set it to '0' and add
// a '1' in front. However we reach the first digit only if all following
// digits had been '9' before rounding up. Now all trailing digits are '0' and
// we simply switch the first digit to '1' and update the decimal-point
// (indicating that the point is now one digit to the right).
if (buffer[0] == '0' + 10) {
buffer[0] = '1';
(*decimal_point)++;
}
}
// The given fractionals number represents a fixed-point number with binary
// point at bit (-exponent).
// Preconditions:
// -128 <= exponent <= 0.
// 0 <= fractionals * 2^exponent < 1
// The buffer holds the result.
// The function will round its result. During the rounding-process digits not
// generated by this function might be updated, and the decimal-point variable
// might be updated. If this function generates the digits 99 and the buffer
// already contained "199" (thus yielding a buffer of "19999") then a
// rounding-up will change the contents of the buffer to "20000".
static void FillFractionals(uint64_t fractionals, int exponent,
int fractional_count, Vector<char> buffer,
int* length, int* decimal_point) {
ASSERT(-128 <= exponent && exponent <= 0);
// 'fractionals' is a fixed-point number, with binary point at bit
// (-exponent). Inside the function the non-converted remainder of fractionals
// is a fixed-point number, with binary point at bit 'point'.
if (-exponent <= 64) {
// One 64 bit number is sufficient.
ASSERT(fractionals >> 56 == 0);
int point = -exponent;
for (int i = 0; i < fractional_count; ++i) {
if (fractionals == 0) break;
// Instead of multiplying by 10 we multiply by 5 and adjust the point
// location. This way the fractionals variable will not overflow.
// Invariant at the beginning of the loop: fractionals < 2^point.
// Initially we have: point <= 64 and fractionals < 2^56
// After each iteration the point is decremented by one.
// Note that 5^3 = 125 < 128 = 2^7.
// Therefore three iterations of this loop will not overflow fractionals
// (even without the subtraction at the end of the loop body). At this
// time point will satisfy point <= 61 and therefore fractionals < 2^point
// and any further multiplication of fractionals by 5 will not overflow.
fractionals *= 5;
point--;
int digit = static_cast<int>(fractionals >> point);
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
fractionals -= static_cast<uint64_t>(digit) << point;
}
// If the first bit after the point is set we have to round up.
if (((fractionals >> (point - 1)) & 1) == 1) {
RoundUp(buffer, length, decimal_point);
}
} else { // We need 128 bits.
ASSERT(64 < -exponent && -exponent <= 128);
UInt128 fractionals128 = UInt128(fractionals, 0);
fractionals128.Shift(-exponent - 64);
int point = 128;
for (int i = 0; i < fractional_count; ++i) {
if (fractionals128.IsZero()) break;
// As before: instead of multiplying by 10 we multiply by 5 and adjust the
// point location.
// This multiplication will not overflow for the same reasons as before.
fractionals128.Multiply(5);
point--;
int digit = fractionals128.DivModPowerOf2(point);
ASSERT(digit <= 9);
buffer[*length] = static_cast<char>('0' + digit);
(*length)++;
}
if (fractionals128.BitAt(point - 1) == 1) {
RoundUp(buffer, length, decimal_point);
}
}
}
// Removes leading and trailing zeros.
// If leading zeros are removed then the decimal point position is adjusted.
static void TrimZeros(Vector<char> buffer, int* length, int* decimal_point) {
while (*length > 0 && buffer[(*length) - 1] == '0') {
(*length)--;
}
int first_non_zero = 0;
while (first_non_zero < *length && buffer[first_non_zero] == '0') {
first_non_zero++;
}
if (first_non_zero != 0) {
for (int i = first_non_zero; i < *length; ++i) {
buffer[i - first_non_zero] = buffer[i];
}
*length -= first_non_zero;
*decimal_point -= first_non_zero;
}
}
bool FastFixedDtoa(double v,
int fractional_count,
Vector<char> buffer,
int* length,
int* decimal_point) {
const uint32_t kMaxUInt32 = 0xFFFFFFFF;
uint64_t significand = Double(v).Significand();
int exponent = Double(v).Exponent();
// v = significand * 2^exponent (with significand a 53bit integer).
// If the exponent is larger than 20 (i.e. we may have a 73bit number) then we
// don't know how to compute the representation. 2^73 ~= 9.5*10^21.
// If necessary this limit could probably be increased, but we don't need
// more.
if (exponent > 20) return false;
if (fractional_count > 20) return false;
*length = 0;
// At most kDoubleSignificandSize bits of the significand are non-zero.
// Given a 64 bit integer we have 11 0s followed by 53 potentially non-zero
// bits: 0..11*..0xxx..53*..xx
if (exponent + kDoubleSignificandSize > 64) {
// The exponent must be > 11.
//
// We know that v = significand * 2^exponent.
// And the exponent > 11.
// We simplify the task by dividing v by 10^17.
// The quotient delivers the first digits, and the remainder fits into a 64
// bit number.
// Dividing by 10^17 is equivalent to dividing by 5^17*2^17.
const uint64_t kFive17 = UINT64_2PART_C(0xB1, A2BC2EC5); // 5^17
uint64_t divisor = kFive17;
int divisor_power = 17;
uint64_t dividend = significand;
uint32_t quotient;
uint64_t remainder;
// Let v = f * 2^e with f == significand and e == exponent.
// Then need q (quotient) and r (remainder) as follows:
// v = q * 10^17 + r
// f * 2^e = q * 10^17 + r
// f * 2^e = q * 5^17 * 2^17 + r
// If e > 17 then
// f * 2^(e-17) = q * 5^17 + r/2^17
// else
// f = q * 5^17 * 2^(17-e) + r/2^e
if (exponent > divisor_power) {
// We only allow exponents of up to 20 and therefore (17 - e) <= 3
dividend <<= exponent - divisor_power;
quotient = static_cast<uint32_t>(dividend / divisor);
remainder = (dividend % divisor) << divisor_power;
} else {
divisor <<= divisor_power - exponent;
quotient = static_cast<uint32_t>(dividend / divisor);
remainder = (dividend % divisor) << exponent;
}
FillDigits32(quotient, buffer, length);
FillDigits64FixedLength(remainder, buffer, length);
*decimal_point = *length;
} else if (exponent >= 0) {
// 0 <= exponent <= 11
significand <<= exponent;
FillDigits64(significand, buffer, length);
*decimal_point = *length;
} else if (exponent > -kDoubleSignificandSize) {
// We have to cut the number.
uint64_t integrals = significand >> -exponent;
uint64_t fractionals = significand - (integrals << -exponent);
if (integrals > kMaxUInt32) {
FillDigits64(integrals, buffer, length);
} else {
FillDigits32(static_cast<uint32_t>(integrals), buffer, length);
}
*decimal_point = *length;
FillFractionals(fractionals, exponent, fractional_count,
buffer, length, decimal_point);
} else if (exponent < -128) {
// This configuration (with at most 20 digits) means that all digits must be
// 0.
ASSERT(fractional_count <= 20);
buffer[0] = '\0';
*length = 0;
*decimal_point = -fractional_count;
} else {
*decimal_point = 0;
FillFractionals(significand, exponent, fractional_count,
buffer, length, decimal_point);
}
TrimZeros(buffer, length, decimal_point);
buffer[*length] = '\0';
if ((*length) == 0) {
// The string is empty and the decimal_point thus has no importance. Mimic
// Gay's dtoa and and set it to -fractional_count.
*decimal_point = -fractional_count;
}
return true;
}
} // namespace double_conversion

56
base/poco/Foundation/src/fixed-dtoa.h
View File

@ -1,56 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_FIXED_DTOA_H_
#define DOUBLE_CONVERSION_FIXED_DTOA_H_
#include "utils.h"
namespace double_conversion
{
// Produces digits necessary to print a given number with
// 'fractional_count' digits after the decimal point.
// The buffer must be big enough to hold the result plus one terminating null
// character.
//
// The produced digits might be too short in which case the caller has to fill
// the gaps with '0's.
// Example: FastFixedDtoa(0.001, 5, ...) is allowed to return buffer = "1", and
// decimal_point = -2.
// Halfway cases are rounded towards +/-Infinity (away from 0). The call
// FastFixedDtoa(0.15, 2, ...) thus returns buffer = "2", decimal_point = 0.
// The returned buffer may contain digits that would be truncated from the
// shortest representation of the input.
//
// This method only works for some parameters. If it can't handle the input it
// returns false. The output is null-terminated when the function succeeds.
bool FastFixedDtoa(double v, int fractional_count, Vector<char> buffer, int * length, int * decimal_point);
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_FIXED_DTOA_H_

194
base/poco/Foundation/src/gzguts.h
View File

@ -1,194 +0,0 @@
/* gzguts.h -- zlib internal header definitions for gz* operations
* Copyright (C) 2004, 2005, 2010, 2011, 2012, 2013 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
#ifdef _LARGEFILE64_SOURCE
# ifndef _LARGEFILE_SOURCE
# define _LARGEFILE_SOURCE 1
# endif
# ifdef _FILE_OFFSET_BITS
# undef _FILE_OFFSET_BITS
# endif
#endif
#ifdef HAVE_HIDDEN
# define ZLIB_INTERNAL __attribute__((visibility("hidden")))
#else
# define ZLIB_INTERNAL
#endif
#include <stdio.h>
#include "zlib.h"
#ifdef STDC
# include <limits.h>
# include <stdlib.h>
# include <string.h>
#endif
#ifndef _POSIX_SOURCE
# define _POSIX_SOURCE
#endif
#include <fcntl.h>
# if defined(__TURBOC__) || defined(_MSC_VER) || defined(_WIN32)
# include <io.h>
# endif
#if defined(_WIN32) || defined(__CYGWIN__)
# define WIDECHAR
#endif
#ifdef WINAPI_FAMILY
# define open _open
# define read _read
# define write _write
# define close _close
#endif
#ifdef NO_DEFLATE /* for compatibility with old definition */
# define NO_GZCOMPRESS
#endif
#if defined(STDC99) || (defined(__TURBOC__) && __TURBOC__ >= 0x550)
# ifndef HAVE_VSNPRINTF
# define HAVE_VSNPRINTF
# endif
#endif
#ifndef HAVE_VSNPRINTF
# ifdef __TURBOC__
# define NO_vsnprintf
# endif
# ifdef WIN32
/* In Win32, vsnprintf is available as the "non-ANSI" _vsnprintf. */
# if !defined(vsnprintf) && !defined(NO_vsnprintf)
# define vsnprintf _vsnprintf
# endif
# endif
# ifdef __SASC
# define NO_vsnprintf
# endif
# ifdef VMS
# define NO_vsnprintf
# endif
# ifdef __OS400__
# define NO_vsnprintf
# endif
# ifdef __MVS__
# define NO_vsnprintf
# endif
#endif
/* unlike snprintf (which is required in C99), _snprintf does not guarantee
null termination of the result -- however this is only used in gzlib.c where
the result is assured to fit in the space provided */
#ifndef local
# define local static
#endif
/* since "static" is used to mean two completely different things in C, we
define "local" for the non-static meaning of "static", for readability
(compile with -Dlocal if your debugger can't find static symbols) */
/* gz* functions always use library allocation functions */
#ifndef STDC
extern voidp malloc OF((uInt size));
extern void free OF((voidpf ptr));
#endif
/* get errno and strerror definition */
#if defined UNDER_CE
# include <windows.h>
# define zstrerror() gz_strwinerror((DWORD)GetLastError())
#else
# ifndef NO_STRERROR
# include <errno.h>
# define zstrerror() strerror(errno)
# else
# define zstrerror() "stdio error (consult errno)"
# endif
#endif
/* provide prototypes for these when building zlib without LFS */
#if !defined(_LARGEFILE64_SOURCE) || _LFS64_LARGEFILE - 0 == 0
ZEXTERN gzFile ZEXPORT gzopen64 OF((const char *, const char *));
ZEXTERN z_off64_t ZEXPORT gzseek64 OF((gzFile, z_off64_t, int));
ZEXTERN z_off64_t ZEXPORT gztell64 OF((gzFile));
ZEXTERN z_off64_t ZEXPORT gzoffset64 OF((gzFile));
#endif
/* default memLevel */
#if MAX_MEM_LEVEL >= 8
# define DEF_MEM_LEVEL 8
#else
# define DEF_MEM_LEVEL MAX_MEM_LEVEL
#endif
/* default i/o buffer size -- double this for output when reading (this and
twice this must be able to fit in an unsigned type) */
#define GZBUFSIZE 8192
/* gzip modes, also provide a little integrity check on the passed structure */
#define GZ_NONE 0
#define GZ_READ 7247
#define GZ_WRITE 31153
#define GZ_APPEND 1 /* mode set to GZ_WRITE after the file is opened */
/* values for gz_state how */
#define LOOK 0 /* look for a gzip header */
#define COPY 1 /* copy input directly */
#define GZIP 2 /* decompress a gzip stream */
/* internal gzip file state data structure */
typedef struct
{
/* exposed contents for gzgetc() macro */
struct gzFile_s x; /* "x" for exposed */
/* x.have: number of bytes available at x.next */
/* x.next: next output data to deliver or write */
/* x.pos: current position in uncompressed data */
/* used for both reading and writing */
int mode; /* see gzip modes above */
int fd; /* file descriptor */
char * path; /* path or fd for error messages */
unsigned size; /* buffer size, zero if not allocated yet */
unsigned want; /* requested buffer size, default is GZBUFSIZE */
unsigned char * in; /* input buffer (double-sized when writing) */
unsigned char * out; /* output buffer (double-sized when reading) */
int direct; /* 0 if processing gzip, 1 if transparent */
/* just for reading */
int how; /* 0: get header, 1: copy, 2: decompress */
z_off64_t start; /* where the gzip data started, for rewinding */
int eof; /* true if end of input file reached */
int past; /* true if read requested past end */
/* just for writing */
int level; /* compression level */
int strategy; /* compression strategy */
/* seek request */
z_off64_t skip; /* amount to skip (already rewound if backwards) */
int seek; /* true if seek request pending */
/* error information */
int err; /* error code */
char * msg; /* error message */
/* zlib inflate or deflate stream */
z_stream strm; /* stream structure in-place (not a pointer) */
} gz_state;
typedef gz_state FAR * gz_statep;
/* shared functions */
void ZLIB_INTERNAL gz_error OF((gz_statep, int, const char *));
#if defined UNDER_CE
char ZLIB_INTERNAL * gz_strwinerror OF((DWORD error));
#endif
/* GT_OFF(x), where x is an unsigned value, is true if x > maximum z_off64_t
value -- needed when comparing unsigned to z_off64_t, which is signed
(possible z_off64_t types off_t, off64_t, and long are all signed) */
#ifdef INT_MAX
# define GT_OFF(x) (sizeof(int) == sizeof(z_off64_t) && (x) > INT_MAX)
#else
unsigned ZLIB_INTERNAL gz_intmax OF((void));
# define GT_OFF(x) (sizeof(int) == sizeof(z_off64_t) && (x) > gz_intmax())
#endif

458
base/poco/Foundation/src/ieee.h
View File

@ -1,458 +0,0 @@
// Copyright 2012 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_DOUBLE_H_
#define DOUBLE_CONVERSION_DOUBLE_H_
#include "diy-fp.h"
namespace double_conversion
{
// We assume that doubles and uint64_t have the same endianness.
static uint64_t double_to_uint64(double d)
{
return BitCast<uint64_t>(d);
}
static double uint64_to_double(uint64_t d64)
{
return BitCast<double>(d64);
}
static uint32_t float_to_uint32(float f)
{
return BitCast<uint32_t>(f);
}
static float uint32_to_float(uint32_t d32)
{
return BitCast<float>(d32);
}
// Helper functions for doubles.
class Double
{
public:
static const uint64_t kSignMask = UINT64_2PART_C(0x80000000, 00000000);
static const uint64_t kExponentMask = UINT64_2PART_C(0x7FF00000, 00000000);
static const uint64_t kSignificandMask = UINT64_2PART_C(0x000FFFFF, FFFFFFFF);
static const uint64_t kHiddenBit = UINT64_2PART_C(0x00100000, 00000000);
static const int kPhysicalSignificandSize = 52; // Excludes the hidden bit.
static const int kSignificandSize = 53;
Double() : d64_(0) { }
explicit Double(double d) : d64_(double_to_uint64(d)) { }
explicit Double(uint64_t d64) : d64_(d64) { }
explicit Double(DiyFp diy_fp) : d64_(DiyFpToUint64(diy_fp)) { }
// The value encoded by this Double must be greater or equal to +0.0.
// It must not be special (infinity, or NaN).
DiyFp AsDiyFp() const
{
ASSERT(Sign() > 0);
ASSERT(!IsSpecial());
return DiyFp(Significand(), Exponent());
}
// The value encoded by this Double must be strictly greater than 0.
DiyFp AsNormalizedDiyFp() const
{
ASSERT(value() > 0.0);
uint64_t f = Significand();
int e = Exponent();
// The current double could be a denormal.
while ((f & kHiddenBit) == 0)
{
f <<= 1;
e--;
}
// Do the final shifts in one go.
f <<= DiyFp::kSignificandSize - kSignificandSize;
e -= DiyFp::kSignificandSize - kSignificandSize;
return DiyFp(f, e);
}
// Returns the double's bit as uint64.
uint64_t AsUint64() const { return d64_; }
// Returns the next greater double. Returns +infinity on input +infinity.
double NextDouble() const
{
if (d64_ == kInfinity)
return Double(kInfinity).value();
if (Sign() < 0 && Significand() == 0)
{
// -0.0
return 0.0;
}
if (Sign() < 0)
{
return Double(d64_ - 1).value();
}
else
{
return Double(d64_ + 1).value();
}
}
double PreviousDouble() const
{
if (d64_ == (kInfinity | kSignMask))
return -Double::Infinity();
if (Sign() < 0)
{
return Double(d64_ + 1).value();
}
else
{
if (Significand() == 0)
return -0.0;
return Double(d64_ - 1).value();
}
}
int Exponent() const
{
if (IsDenormal())
return kDenormalExponent;
uint64_t d64 = AsUint64();
int biased_e = static_cast<int>((d64 & kExponentMask) >> kPhysicalSignificandSize);
return biased_e - kExponentBias;
}
uint64_t Significand() const
{
uint64_t d64 = AsUint64();
uint64_t significand = d64 & kSignificandMask;
if (!IsDenormal())
{
return significand + kHiddenBit;
}
else
{
return significand;
}
}
// Returns true if the double is a denormal.
bool IsDenormal() const
{
uint64_t d64 = AsUint64();
return (d64 & kExponentMask) == 0;
}
// We consider denormals not to be special.
// Hence only Infinity and NaN are special.
bool IsSpecial() const
{
uint64_t d64 = AsUint64();
return (d64 & kExponentMask) == kExponentMask;
}
bool IsNan() const
{
uint64_t d64 = AsUint64();
return ((d64 & kExponentMask) == kExponentMask) && ((d64 & kSignificandMask) != 0);
}
bool IsInfinite() const
{
uint64_t d64 = AsUint64();
return ((d64 & kExponentMask) == kExponentMask) && ((d64 & kSignificandMask) == 0);
}
int Sign() const
{
uint64_t d64 = AsUint64();
return (d64 & kSignMask) == 0 ? 1 : -1;
}
// Precondition: the value encoded by this Double must be greater or equal
// than +0.0.
DiyFp UpperBoundary() const
{
ASSERT(Sign() > 0);
return DiyFp(Significand() * 2 + 1, Exponent() - 1);
}
// Computes the two boundaries of this.
// The bigger boundary (m_plus) is normalized. The lower boundary has the same
// exponent as m_plus.
// Precondition: the value encoded by this Double must be greater than 0.
void NormalizedBoundaries(DiyFp * out_m_minus, DiyFp * out_m_plus) const
{
ASSERT(value() > 0.0);
DiyFp v = this->AsDiyFp();
DiyFp m_plus = DiyFp::Normalize(DiyFp((v.f() << 1) + 1, v.e() - 1));
DiyFp m_minus;
if (LowerBoundaryIsCloser())
{
m_minus = DiyFp((v.f() << 2) - 1, v.e() - 2);
}
else
{
m_minus = DiyFp((v.f() << 1) - 1, v.e() - 1);
}
m_minus.set_f(m_minus.f() << (m_minus.e() - m_plus.e()));
m_minus.set_e(m_plus.e());
*out_m_plus = m_plus;
*out_m_minus = m_minus;
}
bool LowerBoundaryIsCloser() const
{
// The boundary is closer if the significand is of the form f == 2^p-1 then
// the lower boundary is closer.
// Think of v = 1000e10 and v- = 9999e9.
// Then the boundary (== (v - v-)/2) is not just at a distance of 1e9 but
// at a distance of 1e8.
// The only exception is for the smallest normal: the largest denormal is
// at the same distance as its successor.
// Note: denormals have the same exponent as the smallest normals.
bool physical_significand_is_zero = ((AsUint64() & kSignificandMask) == 0);
return physical_significand_is_zero && (Exponent() != kDenormalExponent);
}
double value() const { return uint64_to_double(d64_); }
// Returns the significand size for a given order of magnitude.
// If v = f*2^e with 2^p-1 <= f <= 2^p then p+e is v's order of magnitude.
// This function returns the number of significant binary digits v will have
// once it's encoded into a double. In almost all cases this is equal to
// kSignificandSize. The only exceptions are denormals. They start with
// leading zeroes and their effective significand-size is hence smaller.
static int SignificandSizeForOrderOfMagnitude(int order)
{
if (order >= (kDenormalExponent + kSignificandSize))
{
return kSignificandSize;
}
if (order <= kDenormalExponent)
return 0;
return order - kDenormalExponent;
}
static double Infinity() { return Double(kInfinity).value(); }
static double NaN() { return Double(kNaN).value(); }
private:
static const int kExponentBias = 0x3FF + kPhysicalSignificandSize;
static const int kDenormalExponent = -kExponentBias + 1;
static const int kMaxExponent = 0x7FF - kExponentBias;
static const uint64_t kInfinity = UINT64_2PART_C(0x7FF00000, 00000000);
static const uint64_t kNaN = UINT64_2PART_C(0x7FF80000, 00000000);
const uint64_t d64_;
static uint64_t DiyFpToUint64(DiyFp diy_fp)
{
uint64_t significand = diy_fp.f();
int exponent = diy_fp.e();
while (significand > kHiddenBit + kSignificandMask)
{
significand >>= 1;
exponent++;
}
if (exponent >= kMaxExponent)
{
return kInfinity;
}
if (exponent < kDenormalExponent)
{
return 0;
}
while (exponent > kDenormalExponent && (significand & kHiddenBit) == 0)
{
significand <<= 1;
exponent--;
}
uint64_t biased_exponent;
if (exponent == kDenormalExponent && (significand & kHiddenBit) == 0)
{
biased_exponent = 0;
}
else
{
biased_exponent = static_cast<uint64_t>(exponent + kExponentBias);
}
return (significand & kSignificandMask) | (biased_exponent << kPhysicalSignificandSize);
}
DISALLOW_COPY_AND_ASSIGN(Double);
};
class Single
{
public:
static const uint32_t kSignMask = 0x80000000;
static const uint32_t kExponentMask = 0x7F800000;
static const uint32_t kSignificandMask = 0x007FFFFF;
static const uint32_t kHiddenBit = 0x00800000;
static const int kPhysicalSignificandSize = 23; // Excludes the hidden bit.
static const int kSignificandSize = 24;
Single() : d32_(0) { }
explicit Single(float f) : d32_(float_to_uint32(f)) { }
explicit Single(uint32_t d32) : d32_(d32) { }
// The value encoded by this Single must be greater or equal to +0.0.
// It must not be special (infinity, or NaN).
DiyFp AsDiyFp() const
{
ASSERT(Sign() > 0);
ASSERT(!IsSpecial());
return DiyFp(Significand(), Exponent());
}
// Returns the single's bit as uint64.
uint32_t AsUint32() const { return d32_; }
int Exponent() const
{
if (IsDenormal())
return kDenormalExponent;
uint32_t d32 = AsUint32();
int biased_e = static_cast<int>((d32 & kExponentMask) >> kPhysicalSignificandSize);
return biased_e - kExponentBias;
}
uint32_t Significand() const
{
uint32_t d32 = AsUint32();
uint32_t significand = d32 & kSignificandMask;
if (!IsDenormal())
{
return significand + kHiddenBit;
}
else
{
return significand;
}
}
// Returns true if the single is a denormal.
bool IsDenormal() const
{
uint32_t d32 = AsUint32();
return (d32 & kExponentMask) == 0;
}
// We consider denormals not to be special.
// Hence only Infinity and NaN are special.
bool IsSpecial() const
{
uint32_t d32 = AsUint32();
return (d32 & kExponentMask) == kExponentMask;
}
bool IsNan() const
{
uint32_t d32 = AsUint32();
return ((d32 & kExponentMask) == kExponentMask) && ((d32 & kSignificandMask) != 0);
}
bool IsInfinite() const
{
uint32_t d32 = AsUint32();
return ((d32 & kExponentMask) == kExponentMask) && ((d32 & kSignificandMask) == 0);
}
int Sign() const
{
uint32_t d32 = AsUint32();
return (d32 & kSignMask) == 0 ? 1 : -1;
}
// Computes the two boundaries of this.
// The bigger boundary (m_plus) is normalized. The lower boundary has the same
// exponent as m_plus.
// Precondition: the value encoded by this Single must be greater than 0.
void NormalizedBoundaries(DiyFp * out_m_minus, DiyFp * out_m_plus) const
{
ASSERT(value() > 0.0);
DiyFp v = this->AsDiyFp();
DiyFp m_plus = DiyFp::Normalize(DiyFp((v.f() << 1) + 1, v.e() - 1));
DiyFp m_minus;
if (LowerBoundaryIsCloser())
{
m_minus = DiyFp((v.f() << 2) - 1, v.e() - 2);
}
else
{
m_minus = DiyFp((v.f() << 1) - 1, v.e() - 1);
}
m_minus.set_f(m_minus.f() << (m_minus.e() - m_plus.e()));
m_minus.set_e(m_plus.e());
*out_m_plus = m_plus;
*out_m_minus = m_minus;
}
// Precondition: the value encoded by this Single must be greater or equal
// than +0.0.
DiyFp UpperBoundary() const
{
ASSERT(Sign() > 0);
return DiyFp(Significand() * 2 + 1, Exponent() - 1);
}
bool LowerBoundaryIsCloser() const
{
// The boundary is closer if the significand is of the form f == 2^p-1 then
// the lower boundary is closer.
// Think of v = 1000e10 and v- = 9999e9.
// Then the boundary (== (v - v-)/2) is not just at a distance of 1e9 but
// at a distance of 1e8.
// The only exception is for the smallest normal: the largest denormal is
// at the same distance as its successor.
// Note: denormals have the same exponent as the smallest normals.
bool physical_significand_is_zero = ((AsUint32() & kSignificandMask) == 0);
return physical_significand_is_zero && (Exponent() != kDenormalExponent);
}
float value() const { return uint32_to_float(d32_); }
static float Infinity() { return Single(kInfinity).value(); }
static float NaN() { return Single(kNaN).value(); }
private:
static const int kExponentBias = 0x7F + kPhysicalSignificandSize;
static const int kDenormalExponent = -kExponentBias + 1;
static const int kMaxExponent = 0xFF - kExponentBias;
static const uint32_t kInfinity = 0x7F800000;
static const uint32_t kNaN = 0x7FC00000;
const uint32_t d32_;
DISALLOW_COPY_AND_ASSIGN(Single);
};
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_DOUBLE_H_

640
base/poco/Foundation/src/infback.c
View File

@ -1,640 +0,0 @@
/* infback.c -- inflate using a call-back interface
* Copyright (C) 1995-2016 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/*
This code is largely copied from inflate.c. Normally either infback.o or
inflate.o would be linked into an application--not both. The interface
with inffast.c is retained so that optimized assembler-coded versions of
inflate_fast() can be used with either inflate.c or infback.c.
*/
#include "zutil.h"
#include "inftrees.h"
#include "inflate.h"
#include "inffast.h"
/* function prototypes */
local void fixedtables OF((struct inflate_state FAR *state));
/*
strm provides memory allocation functions in zalloc and zfree, or
Z_NULL to use the library memory allocation functions.
windowBits is in the range 8..15, and window is a user-supplied
window and output buffer that is 2**windowBits bytes.
*/
int ZEXPORT inflateBackInit_(strm, windowBits, window, version, stream_size)
z_streamp strm;
int windowBits;
unsigned char FAR *window;
const char *version;
int stream_size;
{
struct inflate_state FAR *state;
if (version == Z_NULL || version[0] != ZLIB_VERSION[0] ||
stream_size != (int)(sizeof(z_stream)))
return Z_VERSION_ERROR;
if (strm == Z_NULL || window == Z_NULL ||
windowBits < 8 || windowBits > 15)
return Z_STREAM_ERROR;
strm->msg = Z_NULL; /* in case we return an error */
if (strm->zalloc == (alloc_func)0) {
#ifdef Z_SOLO
return Z_STREAM_ERROR;
#else
strm->zalloc = zcalloc;
strm->opaque = (voidpf)0;
#endif
}
if (strm->zfree == (free_func)0)
#ifdef Z_SOLO
return Z_STREAM_ERROR;
#else
strm->zfree = zcfree;
#endif
state = (struct inflate_state FAR *)ZALLOC(strm, 1,
sizeof(struct inflate_state));
if (state == Z_NULL) return Z_MEM_ERROR;
Tracev((stderr, "inflate: allocated\n"));
strm->state = (struct internal_state FAR *)state;
state->dmax = 32768U;
state->wbits = (uInt)windowBits;
state->wsize = 1U << windowBits;
state->window = window;
state->wnext = 0;
state->whave = 0;
return Z_OK;
}
/*
Return state with length and distance decoding tables and index sizes set to
fixed code decoding. Normally this returns fixed tables from inffixed.h.
If BUILDFIXED is defined, then instead this routine builds the tables the
first time it's called, and returns those tables the first time and
thereafter. This reduces the size of the code by about 2K bytes, in
exchange for a little execution time. However, BUILDFIXED should not be
used for threaded applications, since the rewriting of the tables and virgin
may not be thread-safe.
*/
local void fixedtables(state)
struct inflate_state FAR *state;
{
#ifdef BUILDFIXED
static int virgin = 1;
static code *lenfix, *distfix;
static code fixed[544];
/* build fixed huffman tables if first call (may not be thread safe) */
if (virgin) {
unsigned sym, bits;
static code *next;
/* literal/length table */
sym = 0;
while (sym < 144) state->lens[sym++] = 8;
while (sym < 256) state->lens[sym++] = 9;
while (sym < 280) state->lens[sym++] = 7;
while (sym < 288) state->lens[sym++] = 8;
next = fixed;
lenfix = next;
bits = 9;
inflate_table(LENS, state->lens, 288, &(next), &(bits), state->work);
/* distance table */
sym = 0;
while (sym < 32) state->lens[sym++] = 5;
distfix = next;
bits = 5;
inflate_table(DISTS, state->lens, 32, &(next), &(bits), state->work);
/* do this just once */
virgin = 0;
}
#else /* !BUILDFIXED */
# include "inffixed.h"
#endif /* BUILDFIXED */
state->lencode = lenfix;
state->lenbits = 9;
state->distcode = distfix;
state->distbits = 5;
}
/* Macros for inflateBack(): */
/* Load returned state from inflate_fast() */
#define LOAD() \
do { \
put = strm->next_out; \
left = strm->avail_out; \
next = strm->next_in; \
have = strm->avail_in; \
hold = state->hold; \
bits = state->bits; \
} while (0)
/* Set state from registers for inflate_fast() */
#define RESTORE() \
do { \
strm->next_out = put; \
strm->avail_out = left; \
strm->next_in = next; \
strm->avail_in = have; \
state->hold = hold; \
state->bits = bits; \
} while (0)
/* Clear the input bit accumulator */
#define INITBITS() \
do { \
hold = 0; \
bits = 0; \
} while (0)
/* Assure that some input is available. If input is requested, but denied,
then return a Z_BUF_ERROR from inflateBack(). */
#define PULL() \
do { \
if (have == 0) { \
have = in(in_desc, &next); \
if (have == 0) { \
next = Z_NULL; \
ret = Z_BUF_ERROR; \
goto inf_leave; \
} \
} \
} while (0)
/* Get a byte of input into the bit accumulator, or return from inflateBack()
with an error if there is no input available. */
#define PULLBYTE() \
do { \
PULL(); \
have--; \
hold += (unsigned long)(*next++) << bits; \
bits += 8; \
} while (0)
/* Assure that there are at least n bits in the bit accumulator. If there is
not enough available input to do that, then return from inflateBack() with
an error. */
#define NEEDBITS(n) \
do { \
while (bits < (unsigned)(n)) \
PULLBYTE(); \
} while (0)
/* Return the low n bits of the bit accumulator (n < 16) */
#define BITS(n) \
((unsigned)hold & ((1U << (n)) - 1))
/* Remove n bits from the bit accumulator */
#define DROPBITS(n) \
do { \
hold >>= (n); \
bits -= (unsigned)(n); \
} while (0)
/* Remove zero to seven bits as needed to go to a byte boundary */
#define BYTEBITS() \
do { \
hold >>= bits & 7; \
bits -= bits & 7; \
} while (0)
/* Assure that some output space is available, by writing out the window
if it's full. If the write fails, return from inflateBack() with a
Z_BUF_ERROR. */
#define ROOM() \
do { \
if (left == 0) { \
put = state->window; \
left = state->wsize; \
state->whave = left; \
if (out(out_desc, put, left)) { \
ret = Z_BUF_ERROR; \
goto inf_leave; \
} \
} \
} while (0)
/*
strm provides the memory allocation functions and window buffer on input,
and provides information on the unused input on return. For Z_DATA_ERROR
returns, strm will also provide an error message.
in() and out() are the call-back input and output functions. When
inflateBack() needs more input, it calls in(). When inflateBack() has
filled the window with output, or when it completes with data in the
window, it calls out() to write out the data. The application must not
change the provided input until in() is called again or inflateBack()
returns. The application must not change the window/output buffer until
inflateBack() returns.
in() and out() are called with a descriptor parameter provided in the
inflateBack() call. This parameter can be a structure that provides the
information required to do the read or write, as well as accumulated
information on the input and output such as totals and check values.
in() should return zero on failure. out() should return non-zero on
failure. If either in() or out() fails, than inflateBack() returns a
Z_BUF_ERROR. strm->next_in can be checked for Z_NULL to see whether it
was in() or out() that caused in the error. Otherwise, inflateBack()
returns Z_STREAM_END on success, Z_DATA_ERROR for an deflate format
error, or Z_MEM_ERROR if it could not allocate memory for the state.
inflateBack() can also return Z_STREAM_ERROR if the input parameters
are not correct, i.e. strm is Z_NULL or the state was not initialized.
*/
int ZEXPORT inflateBack(strm, in, in_desc, out, out_desc)
z_streamp strm;
in_func in;
void FAR *in_desc;
out_func out;
void FAR *out_desc;
{
struct inflate_state FAR *state;
z_const unsigned char FAR *next; /* next input */
unsigned char FAR *put; /* next output */
unsigned have, left; /* available input and output */
unsigned long hold; /* bit buffer */
unsigned bits; /* bits in bit buffer */
unsigned copy; /* number of stored or match bytes to copy */
unsigned char FAR *from; /* where to copy match bytes from */
code here; /* current decoding table entry */
code last; /* parent table entry */
unsigned len; /* length to copy for repeats, bits to drop */
int ret; /* return code */
static const unsigned short order[19] = /* permutation of code lengths */
{16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15};
/* Check that the strm exists and that the state was initialized */
if (strm == Z_NULL || strm->state == Z_NULL)
return Z_STREAM_ERROR;
state = (struct inflate_state FAR *)strm->state;
/* Reset the state */
strm->msg = Z_NULL;
state->mode = TYPE;
state->last = 0;
state->whave = 0;
next = strm->next_in;
have = next != Z_NULL ? strm->avail_in : 0;
hold = 0;
bits = 0;
put = state->window;
left = state->wsize;
/* Inflate until end of block marked as last */
for (;;)
switch (state->mode) {
case TYPE:
/* determine and dispatch block type */
if (state->last) {
BYTEBITS();
state->mode = DONE;
break;
}
NEEDBITS(3);
state->last = BITS(1);
DROPBITS(1);
switch (BITS(2)) {
case 0: /* stored block */
Tracev((stderr, "inflate: stored block%s\n",
state->last ? " (last)" : ""));
state->mode = STORED;
break;
case 1: /* fixed block */
fixedtables(state);
Tracev((stderr, "inflate: fixed codes block%s\n",
state->last ? " (last)" : ""));
state->mode = LEN; /* decode codes */
break;
case 2: /* dynamic block */
Tracev((stderr, "inflate: dynamic codes block%s\n",
state->last ? " (last)" : ""));
state->mode = TABLE;
break;
case 3:
strm->msg = (char *)"invalid block type";
state->mode = BAD;
}
DROPBITS(2);
break;
case STORED:
/* get and verify stored block length */
BYTEBITS(); /* go to byte boundary */
NEEDBITS(32);
if ((hold & 0xffff) != ((hold >> 16) ^ 0xffff)) {
strm->msg = (char *)"invalid stored block lengths";
state->mode = BAD;
break;
}
state->length = (unsigned)hold & 0xffff;
Tracev((stderr, "inflate: stored length %u\n",
state->length));
INITBITS();
/* copy stored block from input to output */
while (state->length != 0) {
copy = state->length;
PULL();
ROOM();
if (copy > have) copy = have;
if (copy > left) copy = left;
zmemcpy(put, next, copy);
have -= copy;
next += copy;
left -= copy;
put += copy;
state->length -= copy;
}
Tracev((stderr, "inflate: stored end\n"));
state->mode = TYPE;
break;
case TABLE:
/* get dynamic table entries descriptor */
NEEDBITS(14);
state->nlen = BITS(5) + 257;
DROPBITS(5);
state->ndist = BITS(5) + 1;
DROPBITS(5);
state->ncode = BITS(4) + 4;
DROPBITS(4);
#ifndef PKZIP_BUG_WORKAROUND
if (state->nlen > 286 || state->ndist > 30) {
strm->msg = (char *)"too many length or distance symbols";
state->mode = BAD;
break;
}
#endif
Tracev((stderr, "inflate: table sizes ok\n"));
/* get code length code lengths (not a typo) */
state->have = 0;
while (state->have < state->ncode) {
NEEDBITS(3);
state->lens[order[state->have++]] = (unsigned short)BITS(3);
DROPBITS(3);
}
while (state->have < 19)
state->lens[order[state->have++]] = 0;
state->next = state->codes;
state->lencode = (code const FAR *)(state->next);
state->lenbits = 7;
ret = inflate_table(CODES, state->lens, 19, &(state->next),
&(state->lenbits), state->work);
if (ret) {
strm->msg = (char *)"invalid code lengths set";
state->mode = BAD;
break;
}
Tracev((stderr, "inflate: code lengths ok\n"));
/* get length and distance code code lengths */
state->have = 0;
while (state->have < state->nlen + state->ndist) {
for (;;) {
here = state->lencode[BITS(state->lenbits)];
if ((unsigned)(here.bits) <= bits) break;
PULLBYTE();
}
if (here.val < 16) {
DROPBITS(here.bits);
state->lens[state->have++] = here.val;
}
else {
if (here.val == 16) {
NEEDBITS(here.bits + 2);
DROPBITS(here.bits);
if (state->have == 0) {
strm->msg = (char *)"invalid bit length repeat";
state->mode = BAD;
break;
}
len = (unsigned)(state->lens[state->have - 1]);
copy = 3 + BITS(2);
DROPBITS(2);
}
else if (here.val == 17) {
NEEDBITS(here.bits + 3);
DROPBITS(here.bits);
len = 0;
copy = 3 + BITS(3);
DROPBITS(3);
}
else {
NEEDBITS(here.bits + 7);
DROPBITS(here.bits);
len = 0;
copy = 11 + BITS(7);
DROPBITS(7);
}
if (state->have + copy > state->nlen + state->ndist) {
strm->msg = (char *)"invalid bit length repeat";
state->mode = BAD;
break;
}
while (copy--)
state->lens[state->have++] = (unsigned short)len;
}
}
/* handle error breaks in while */
if (state->mode == BAD) break;
/* check for end-of-block code (better have one) */
if (state->lens[256] == 0) {
strm->msg = (char *)"invalid code -- missing end-of-block";
state->mode = BAD;
break;
}
/* build code tables -- note: do not change the lenbits or distbits
values here (9 and 6) without reading the comments in inftrees.h
concerning the ENOUGH constants, which depend on those values */
state->next = state->codes;
state->lencode = (code const FAR *)(state->next);
state->lenbits = 9;
ret = inflate_table(LENS, state->lens, state->nlen, &(state->next),
&(state->lenbits), state->work);
if (ret) {
strm->msg = (char *)"invalid literal/lengths set";
state->mode = BAD;
break;
}
state->distcode = (code const FAR *)(state->next);
state->distbits = 6;
ret = inflate_table(DISTS, state->lens + state->nlen, state->ndist,
&(state->next), &(state->distbits), state->work);
if (ret) {
strm->msg = (char *)"invalid distances set";
state->mode = BAD;
break;
}
Tracev((stderr, "inflate: codes ok\n"));
state->mode = LEN;
case LEN:
/* use inflate_fast() if we have enough input and output */
if (have >= 6 && left >= 258) {
RESTORE();
if (state->whave < state->wsize)
state->whave = state->wsize - left;
inflate_fast(strm, state->wsize);
LOAD();
break;
}
/* get a literal, length, or end-of-block code */
for (;;) {
here = state->lencode[BITS(state->lenbits)];
if ((unsigned)(here.bits) <= bits) break;
PULLBYTE();
}
if (here.op && (here.op & 0xf0) == 0) {
last = here;
for (;;) {
here = state->lencode[last.val +
(BITS(last.bits + last.op) >> last.bits)];
if ((unsigned)(last.bits + here.bits) <= bits) break;
PULLBYTE();
}
DROPBITS(last.bits);
}
DROPBITS(here.bits);
state->length = (unsigned)here.val;
/* process literal */
if (here.op == 0) {
Tracevv((stderr, here.val >= 0x20 && here.val < 0x7f ?
"inflate: literal '%c'\n" :
"inflate: literal 0x%02x\n", here.val));
ROOM();
*put++ = (unsigned char)(state->length);
left--;
state->mode = LEN;
break;
}
/* process end of block */
if (here.op & 32) {
Tracevv((stderr, "inflate: end of block\n"));
state->mode = TYPE;
break;
}
/* invalid code */
if (here.op & 64) {
strm->msg = (char *)"invalid literal/length code";
state->mode = BAD;
break;
}
/* length code -- get extra bits, if any */
state->extra = (unsigned)(here.op) & 15;
if (state->extra != 0) {
NEEDBITS(state->extra);
state->length += BITS(state->extra);
DROPBITS(state->extra);
}
Tracevv((stderr, "inflate: length %u\n", state->length));
/* get distance code */
for (;;) {
here = state->distcode[BITS(state->distbits)];
if ((unsigned)(here.bits) <= bits) break;
PULLBYTE();
}
if ((here.op & 0xf0) == 0) {
last = here;
for (;;) {
here = state->distcode[last.val +
(BITS(last.bits + last.op) >> last.bits)];
if ((unsigned)(last.bits + here.bits) <= bits) break;
PULLBYTE();
}
DROPBITS(last.bits);
}
DROPBITS(here.bits);
if (here.op & 64) {
strm->msg = (char *)"invalid distance code";
state->mode = BAD;
break;
}
state->offset = (unsigned)here.val;
/* get distance extra bits, if any */
state->extra = (unsigned)(here.op) & 15;
if (state->extra != 0) {
NEEDBITS(state->extra);
state->offset += BITS(state->extra);
DROPBITS(state->extra);
}
if (state->offset > state->wsize - (state->whave < state->wsize ?
left : 0)) {
strm->msg = (char *)"invalid distance too far back";
state->mode = BAD;
break;
}
Tracevv((stderr, "inflate: distance %u\n", state->offset));
/* copy match from window to output */
do {
ROOM();
copy = state->wsize - state->offset;
if (copy < left) {
from = put + copy;
copy = left - copy;
}
else {
from = put - state->offset;
copy = left;
}
if (copy > state->length) copy = state->length;
state->length -= copy;
left -= copy;
do {
*put++ = *from++;
} while (--copy);
} while (state->length != 0);
break;
case DONE:
/* inflate stream terminated properly -- write leftover output */
ret = Z_STREAM_END;
if (left < state->wsize) {
if (out(out_desc, state->window, state->wsize - left))
ret = Z_BUF_ERROR;
}
goto inf_leave;
case BAD:
ret = Z_DATA_ERROR;
goto inf_leave;
default: /* can't happen, but makes compilers happy */
ret = Z_STREAM_ERROR;
goto inf_leave;
}
/* Return unused input */
inf_leave:
strm->next_in = next;
strm->avail_in = have;
return ret;
}
int ZEXPORT inflateBackEnd(strm)
z_streamp strm;
{
if (strm == Z_NULL || strm->state == Z_NULL || strm->zfree == (free_func)0)
return Z_STREAM_ERROR;
ZFREE(strm, strm->state);
strm->state = Z_NULL;
Tracev((stderr, "inflate: end\n"));
return Z_OK;
}

323
base/poco/Foundation/src/inffast.c
View File

@ -1,323 +0,0 @@
/* inffast.c -- fast decoding
* Copyright (C) 1995-2017 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
#include "zutil.h"
#include "inftrees.h"
#include "inflate.h"
#include "inffast.h"
#ifdef ASMINF
# pragma message("Assembler code may have bugs -- use at your own risk")
#else
/*
Decode literal, length, and distance codes and write out the resulting
literal and match bytes until either not enough input or output is
available, an end-of-block is encountered, or a data error is encountered.
When large enough input and output buffers are supplied to inflate(), for
example, a 16K input buffer and a 64K output buffer, more than 95% of the
inflate execution time is spent in this routine.
Entry assumptions:
state->mode == LEN
strm->avail_in >= 6
strm->avail_out >= 258
start >= strm->avail_out
state->bits < 8
On return, state->mode is one of:
LEN -- ran out of enough output space or enough available input
TYPE -- reached end of block code, inflate() to interpret next block
BAD -- error in block data
Notes:
- The maximum input bits used by a length/distance pair is 15 bits for the
length code, 5 bits for the length extra, 15 bits for the distance code,
and 13 bits for the distance extra. This totals 48 bits, or six bytes.
Therefore if strm->avail_in >= 6, then there is enough input to avoid
checking for available input while decoding.
- The maximum bytes that a single length/distance pair can output is 258
bytes, which is the maximum length that can be coded. inflate_fast()
requires strm->avail_out >= 258 for each loop to avoid checking for
output space.
*/
void ZLIB_INTERNAL inflate_fast(strm, start)
z_streamp strm;
unsigned start; /* inflate()'s starting value for strm->avail_out */
{
struct inflate_state FAR *state;
z_const unsigned char FAR *in; /* local strm->next_in */
z_const unsigned char FAR *last; /* have enough input while in < last */
unsigned char FAR *out; /* local strm->next_out */
unsigned char FAR *beg; /* inflate()'s initial strm->next_out */
unsigned char FAR *end; /* while out < end, enough space available */
#ifdef INFLATE_STRICT
unsigned dmax; /* maximum distance from zlib header */
#endif
unsigned wsize; /* window size or zero if not using window */
unsigned whave; /* valid bytes in the window */
unsigned wnext; /* window write index */
unsigned char FAR *window; /* allocated sliding window, if wsize != 0 */
unsigned long hold; /* local strm->hold */
unsigned bits; /* local strm->bits */
code const FAR *lcode; /* local strm->lencode */
code const FAR *dcode; /* local strm->distcode */
unsigned lmask; /* mask for first level of length codes */
unsigned dmask; /* mask for first level of distance codes */
code here; /* retrieved table entry */
unsigned op; /* code bits, operation, extra bits, or */
/* window position, window bytes to copy */
unsigned len; /* match length, unused bytes */
unsigned dist; /* match distance */
unsigned char FAR *from; /* where to copy match from */
/* copy state to local variables */
state = (struct inflate_state FAR *)strm->state;
in = strm->next_in;
last = in + (strm->avail_in - 5);
out = strm->next_out;
beg = out - (start - strm->avail_out);
end = out + (strm->avail_out - 257);
#ifdef INFLATE_STRICT
dmax = state->dmax;
#endif
wsize = state->wsize;
whave = state->whave;
wnext = state->wnext;
window = state->window;
hold = state->hold;
bits = state->bits;
lcode = state->lencode;
dcode = state->distcode;
lmask = (1U << state->lenbits) - 1;
dmask = (1U << state->distbits) - 1;
/* decode literals and length/distances until end-of-block or not enough
input data or output space */
do {
if (bits < 15) {
hold += (unsigned long)(*in++) << bits;
bits += 8;
hold += (unsigned long)(*in++) << bits;
bits += 8;
}
here = lcode[hold & lmask];
dolen:
op = (unsigned)(here.bits);
hold >>= op;
bits -= op;
op = (unsigned)(here.op);
if (op == 0) { /* literal */
Tracevv((stderr, here.val >= 0x20 && here.val < 0x7f ?
"inflate: literal '%c'\n" :
"inflate: literal 0x%02x\n", here.val));
*out++ = (unsigned char)(here.val);
}
else if (op & 16) { /* length base */
len = (unsigned)(here.val);
op &= 15; /* number of extra bits */
if (op) {
if (bits < op) {
hold += (unsigned long)(*in++) << bits;
bits += 8;
}
len += (unsigned)hold & ((1U << op) - 1);
hold >>= op;
bits -= op;
}
Tracevv((stderr, "inflate: length %u\n", len));
if (bits < 15) {
hold += (unsigned long)(*in++) << bits;
bits += 8;
hold += (unsigned long)(*in++) << bits;
bits += 8;
}
here = dcode[hold & dmask];
dodist:
op = (unsigned)(here.bits);
hold >>= op;
bits -= op;
op = (unsigned)(here.op);
if (op & 16) { /* distance base */
dist = (unsigned)(here.val);
op &= 15; /* number of extra bits */
if (bits < op) {
hold += (unsigned long)(*in++) << bits;
bits += 8;
if (bits < op) {
hold += (unsigned long)(*in++) << bits;
bits += 8;
}
}
dist += (unsigned)hold & ((1U << op) - 1);
#ifdef INFLATE_STRICT
if (dist > dmax) {
strm->msg = (char *)"invalid distance too far back";
state->mode = BAD;
break;
}
#endif
hold >>= op;
bits -= op;
Tracevv((stderr, "inflate: distance %u\n", dist));
op = (unsigned)(out - beg); /* max distance in output */
if (dist > op) { /* see if copy from window */
op = dist - op; /* distance back in window */
if (op > whave) {
if (state->sane) {
strm->msg =
(char *)"invalid distance too far back";
state->mode = BAD;
break;
}
#ifdef INFLATE_ALLOW_INVALID_DISTANCE_TOOFAR_ARRR
if (len <= op - whave) {
do {
*out++ = 0;
} while (--len);
continue;
}
len -= op - whave;
do {
*out++ = 0;
} while (--op > whave);
if (op == 0) {
from = out - dist;
do {
*out++ = *from++;
} while (--len);
continue;
}
#endif
}
from = window;
if (wnext == 0) { /* very common case */
from += wsize - op;
if (op < len) { /* some from window */
len -= op;
do {
*out++ = *from++;
} while (--op);
from = out - dist; /* rest from output */
}
}
else if (wnext < op) { /* wrap around window */
from += wsize + wnext - op;
op -= wnext;
if (op < len) { /* some from end of window */
len -= op;
do {
*out++ = *from++;
} while (--op);
from = window;
if (wnext < len) { /* some from start of window */
op = wnext;
len -= op;
do {
*out++ = *from++;
} while (--op);
from = out - dist; /* rest from output */
}
}
}
else { /* contiguous in window */
from += wnext - op;
if (op < len) { /* some from window */
len -= op;
do {
*out++ = *from++;
} while (--op);
from = out - dist; /* rest from output */
}
}
while (len > 2) {
*out++ = *from++;
*out++ = *from++;
*out++ = *from++;
len -= 3;
}
if (len) {
*out++ = *from++;
if (len > 1)
*out++ = *from++;
}
}
else {
from = out - dist; /* copy direct from output */
do { /* minimum length is three */
*out++ = *from++;
*out++ = *from++;
*out++ = *from++;
len -= 3;
} while (len > 2);
if (len) {
*out++ = *from++;
if (len > 1)
*out++ = *from++;
}
}
}
else if ((op & 64) == 0) { /* 2nd level distance code */
here = dcode[here.val + (hold & ((1U << op) - 1))];
goto dodist;
}
else {
strm->msg = (char *)"invalid distance code";
state->mode = BAD;
break;
}
}
else if ((op & 64) == 0) { /* 2nd level length code */
here = lcode[here.val + (hold & ((1U << op) - 1))];
goto dolen;
}
else if (op & 32) { /* end-of-block */
Tracevv((stderr, "inflate: end of block\n"));
state->mode = TYPE;
break;
}
else {
strm->msg = (char *)"invalid literal/length code";
state->mode = BAD;
break;
}
} while (in < last && out < end);
/* return unused bytes (on entry, bits < 8, so in won't go too far back) */
len = bits >> 3;
in -= len;
bits -= len << 3;
hold &= (1U << bits) - 1;
/* update state and return */
strm->next_in = in;
strm->next_out = out;
strm->avail_in = (unsigned)(in < last ? 5 + (last - in) : 5 - (in - last));
strm->avail_out = (unsigned)(out < end ?
257 + (end - out) : 257 - (out - end));
state->hold = hold;
state->bits = bits;
return;
}
/*
inflate_fast() speedups that turned out slower (on a PowerPC G3 750CXe):
- Using bit fields for code structure
- Different op definition to avoid & for extra bits (do & for table bits)
- Three separate decoding do-loops for direct, window, and wnext == 0
- Special case for distance > 1 copies to do overlapped load and store copy
- Explicit branch predictions (based on measured branch probabilities)
- Deferring match copy and interspersed it with decoding subsequent codes
- Swapping literal/length else
- Swapping window/direct else
- Larger unrolled copy loops (three is about right)
- Moving len -= 3 statement into middle of loop
*/
#endif /* !ASMINF */

11
base/poco/Foundation/src/inffast.h
View File

@ -1,11 +0,0 @@
/* inffast.h -- header to use inffast.c
* Copyright (C) 1995-2003, 2010 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* WARNING: this file should *not* be used by applications. It is
part of the implementation of the compression library and is
subject to change. Applications should only use zlib.h.
*/
void ZLIB_INTERNAL inflate_fast OF((z_streamp strm, unsigned start));

68
base/poco/Foundation/src/inffixed.h
View File

@ -1,68 +0,0 @@
/* inffixed.h -- table for decoding fixed codes
* Generated automatically by makefixed().
*/
/* WARNING: this file should *not* be used by applications.
It is part of the implementation of this library and is
subject to change. Applications should only use zlib.h.
*/
static const code lenfix[512] = {
{96, 7, 0}, {0, 8, 80}, {0, 8, 16}, {20, 8, 115}, {18, 7, 31}, {0, 8, 112}, {0, 8, 48}, {0, 9, 192}, {16, 7, 10}, {0, 8, 96},
{0, 8, 32}, {0, 9, 160}, {0, 8, 0}, {0, 8, 128}, {0, 8, 64}, {0, 9, 224}, {16, 7, 6}, {0, 8, 88}, {0, 8, 24}, {0, 9, 144},
{19, 7, 59}, {0, 8, 120}, {0, 8, 56}, {0, 9, 208}, {17, 7, 17}, {0, 8, 104}, {0, 8, 40}, {0, 9, 176}, {0, 8, 8}, {0, 8, 136},
{0, 8, 72}, {0, 9, 240}, {16, 7, 4}, {0, 8, 84}, {0, 8, 20}, {21, 8, 227}, {19, 7, 43}, {0, 8, 116}, {0, 8, 52}, {0, 9, 200},
{17, 7, 13}, {0, 8, 100}, {0, 8, 36}, {0, 9, 168}, {0, 8, 4}, {0, 8, 132}, {0, 8, 68}, {0, 9, 232}, {16, 7, 8}, {0, 8, 92},
{0, 8, 28}, {0, 9, 152}, {20, 7, 83}, {0, 8, 124}, {0, 8, 60}, {0, 9, 216}, {18, 7, 23}, {0, 8, 108}, {0, 8, 44}, {0, 9, 184},
{0, 8, 12}, {0, 8, 140}, {0, 8, 76}, {0, 9, 248}, {16, 7, 3}, {0, 8, 82}, {0, 8, 18}, {21, 8, 163}, {19, 7, 35}, {0, 8, 114},
{0, 8, 50}, {0, 9, 196}, {17, 7, 11}, {0, 8, 98}, {0, 8, 34}, {0, 9, 164}, {0, 8, 2}, {0, 8, 130}, {0, 8, 66}, {0, 9, 228},
{16, 7, 7}, {0, 8, 90}, {0, 8, 26}, {0, 9, 148}, {20, 7, 67}, {0, 8, 122}, {0, 8, 58}, {0, 9, 212}, {18, 7, 19}, {0, 8, 106},
{0, 8, 42}, {0, 9, 180}, {0, 8, 10}, {0, 8, 138}, {0, 8, 74}, {0, 9, 244}, {16, 7, 5}, {0, 8, 86}, {0, 8, 22}, {64, 8, 0},
{19, 7, 51}, {0, 8, 118}, {0, 8, 54}, {0, 9, 204}, {17, 7, 15}, {0, 8, 102}, {0, 8, 38}, {0, 9, 172}, {0, 8, 6}, {0, 8, 134},
{0, 8, 70}, {0, 9, 236}, {16, 7, 9}, {0, 8, 94}, {0, 8, 30}, {0, 9, 156}, {20, 7, 99}, {0, 8, 126}, {0, 8, 62}, {0, 9, 220},
{18, 7, 27}, {0, 8, 110}, {0, 8, 46}, {0, 9, 188}, {0, 8, 14}, {0, 8, 142}, {0, 8, 78}, {0, 9, 252}, {96, 7, 0}, {0, 8, 81},
{0, 8, 17}, {21, 8, 131}, {18, 7, 31}, {0, 8, 113}, {0, 8, 49}, {0, 9, 194}, {16, 7, 10}, {0, 8, 97}, {0, 8, 33}, {0, 9, 162},
{0, 8, 1}, {0, 8, 129}, {0, 8, 65}, {0, 9, 226}, {16, 7, 6}, {0, 8, 89}, {0, 8, 25}, {0, 9, 146}, {19, 7, 59}, {0, 8, 121},
{0, 8, 57}, {0, 9, 210}, {17, 7, 17}, {0, 8, 105}, {0, 8, 41}, {0, 9, 178}, {0, 8, 9}, {0, 8, 137}, {0, 8, 73}, {0, 9, 242},
{16, 7, 4}, {0, 8, 85}, {0, 8, 21}, {16, 8, 258}, {19, 7, 43}, {0, 8, 117}, {0, 8, 53}, {0, 9, 202}, {17, 7, 13}, {0, 8, 101},
{0, 8, 37}, {0, 9, 170}, {0, 8, 5}, {0, 8, 133}, {0, 8, 69}, {0, 9, 234}, {16, 7, 8}, {0, 8, 93}, {0, 8, 29}, {0, 9, 154},
{20, 7, 83}, {0, 8, 125}, {0, 8, 61}, {0, 9, 218}, {18, 7, 23}, {0, 8, 109}, {0, 8, 45}, {0, 9, 186}, {0, 8, 13}, {0, 8, 141},
{0, 8, 77}, {0, 9, 250}, {16, 7, 3}, {0, 8, 83}, {0, 8, 19}, {21, 8, 195}, {19, 7, 35}, {0, 8, 115}, {0, 8, 51}, {0, 9, 198},
{17, 7, 11}, {0, 8, 99}, {0, 8, 35}, {0, 9, 166}, {0, 8, 3}, {0, 8, 131}, {0, 8, 67}, {0, 9, 230}, {16, 7, 7}, {0, 8, 91},
{0, 8, 27}, {0, 9, 150}, {20, 7, 67}, {0, 8, 123}, {0, 8, 59}, {0, 9, 214}, {18, 7, 19}, {0, 8, 107}, {0, 8, 43}, {0, 9, 182},
{0, 8, 11}, {0, 8, 139}, {0, 8, 75}, {0, 9, 246}, {16, 7, 5}, {0, 8, 87}, {0, 8, 23}, {64, 8, 0}, {19, 7, 51}, {0, 8, 119},
{0, 8, 55}, {0, 9, 206}, {17, 7, 15}, {0, 8, 103}, {0, 8, 39}, {0, 9, 174}, {0, 8, 7}, {0, 8, 135}, {0, 8, 71}, {0, 9, 238},
{16, 7, 9}, {0, 8, 95}, {0, 8, 31}, {0, 9, 158}, {20, 7, 99}, {0, 8, 127}, {0, 8, 63}, {0, 9, 222}, {18, 7, 27}, {0, 8, 111},
{0, 8, 47}, {0, 9, 190}, {0, 8, 15}, {0, 8, 143}, {0, 8, 79}, {0, 9, 254}, {96, 7, 0}, {0, 8, 80}, {0, 8, 16}, {20, 8, 115},
{18, 7, 31}, {0, 8, 112}, {0, 8, 48}, {0, 9, 193}, {16, 7, 10}, {0, 8, 96}, {0, 8, 32}, {0, 9, 161}, {0, 8, 0}, {0, 8, 128},
{0, 8, 64}, {0, 9, 225}, {16, 7, 6}, {0, 8, 88}, {0, 8, 24}, {0, 9, 145}, {19, 7, 59}, {0, 8, 120}, {0, 8, 56}, {0, 9, 209},
{17, 7, 17}, {0, 8, 104}, {0, 8, 40}, {0, 9, 177}, {0, 8, 8}, {0, 8, 136}, {0, 8, 72}, {0, 9, 241}, {16, 7, 4}, {0, 8, 84},
{0, 8, 20}, {21, 8, 227}, {19, 7, 43}, {0, 8, 116}, {0, 8, 52}, {0, 9, 201}, {17, 7, 13}, {0, 8, 100}, {0, 8, 36}, {0, 9, 169},
{0, 8, 4}, {0, 8, 132}, {0, 8, 68}, {0, 9, 233}, {16, 7, 8}, {0, 8, 92}, {0, 8, 28}, {0, 9, 153}, {20, 7, 83}, {0, 8, 124},
{0, 8, 60}, {0, 9, 217}, {18, 7, 23}, {0, 8, 108}, {0, 8, 44}, {0, 9, 185}, {0, 8, 12}, {0, 8, 140}, {0, 8, 76}, {0, 9, 249},
{16, 7, 3}, {0, 8, 82}, {0, 8, 18}, {21, 8, 163}, {19, 7, 35}, {0, 8, 114}, {0, 8, 50}, {0, 9, 197}, {17, 7, 11}, {0, 8, 98},
{0, 8, 34}, {0, 9, 165}, {0, 8, 2}, {0, 8, 130}, {0, 8, 66}, {0, 9, 229}, {16, 7, 7}, {0, 8, 90}, {0, 8, 26}, {0, 9, 149},
{20, 7, 67}, {0, 8, 122}, {0, 8, 58}, {0, 9, 213}, {18, 7, 19}, {0, 8, 106}, {0, 8, 42}, {0, 9, 181}, {0, 8, 10}, {0, 8, 138},
{0, 8, 74}, {0, 9, 245}, {16, 7, 5}, {0, 8, 86}, {0, 8, 22}, {64, 8, 0}, {19, 7, 51}, {0, 8, 118}, {0, 8, 54}, {0, 9, 205},
{17, 7, 15}, {0, 8, 102}, {0, 8, 38}, {0, 9, 173}, {0, 8, 6}, {0, 8, 134}, {0, 8, 70}, {0, 9, 237}, {16, 7, 9}, {0, 8, 94},
{0, 8, 30}, {0, 9, 157}, {20, 7, 99}, {0, 8, 126}, {0, 8, 62}, {0, 9, 221}, {18, 7, 27}, {0, 8, 110}, {0, 8, 46}, {0, 9, 189},
{0, 8, 14}, {0, 8, 142}, {0, 8, 78}, {0, 9, 253}, {96, 7, 0}, {0, 8, 81}, {0, 8, 17}, {21, 8, 131}, {18, 7, 31}, {0, 8, 113},
{0, 8, 49}, {0, 9, 195}, {16, 7, 10}, {0, 8, 97}, {0, 8, 33}, {0, 9, 163}, {0, 8, 1}, {0, 8, 129}, {0, 8, 65}, {0, 9, 227},
{16, 7, 6}, {0, 8, 89}, {0, 8, 25}, {0, 9, 147}, {19, 7, 59}, {0, 8, 121}, {0, 8, 57}, {0, 9, 211}, {17, 7, 17}, {0, 8, 105},
{0, 8, 41}, {0, 9, 179}, {0, 8, 9}, {0, 8, 137}, {0, 8, 73}, {0, 9, 243}, {16, 7, 4}, {0, 8, 85}, {0, 8, 21}, {16, 8, 258},
{19, 7, 43}, {0, 8, 117}, {0, 8, 53}, {0, 9, 203}, {17, 7, 13}, {0, 8, 101}, {0, 8, 37}, {0, 9, 171}, {0, 8, 5}, {0, 8, 133},
{0, 8, 69}, {0, 9, 235}, {16, 7, 8}, {0, 8, 93}, {0, 8, 29}, {0, 9, 155}, {20, 7, 83}, {0, 8, 125}, {0, 8, 61}, {0, 9, 219},
{18, 7, 23}, {0, 8, 109}, {0, 8, 45}, {0, 9, 187}, {0, 8, 13}, {0, 8, 141}, {0, 8, 77}, {0, 9, 251}, {16, 7, 3}, {0, 8, 83},
{0, 8, 19}, {21, 8, 195}, {19, 7, 35}, {0, 8, 115}, {0, 8, 51}, {0, 9, 199}, {17, 7, 11}, {0, 8, 99}, {0, 8, 35}, {0, 9, 167},
{0, 8, 3}, {0, 8, 131}, {0, 8, 67}, {0, 9, 231}, {16, 7, 7}, {0, 8, 91}, {0, 8, 27}, {0, 9, 151}, {20, 7, 67}, {0, 8, 123},
{0, 8, 59}, {0, 9, 215}, {18, 7, 19}, {0, 8, 107}, {0, 8, 43}, {0, 9, 183}, {0, 8, 11}, {0, 8, 139}, {0, 8, 75}, {0, 9, 247},
{16, 7, 5}, {0, 8, 87}, {0, 8, 23}, {64, 8, 0}, {19, 7, 51}, {0, 8, 119}, {0, 8, 55}, {0, 9, 207}, {17, 7, 15}, {0, 8, 103},
{0, 8, 39}, {0, 9, 175}, {0, 8, 7}, {0, 8, 135}, {0, 8, 71}, {0, 9, 239}, {16, 7, 9}, {0, 8, 95}, {0, 8, 31}, {0, 9, 159},
{20, 7, 99}, {0, 8, 127}, {0, 8, 63}, {0, 9, 223}, {18, 7, 27}, {0, 8, 111}, {0, 8, 47}, {0, 9, 191}, {0, 8, 15}, {0, 8, 143},
{0, 8, 79}, {0, 9, 255}};
static const code distfix[32]
= {{16, 5, 1}, {23, 5, 257}, {19, 5, 17}, {27, 5, 4097}, {17, 5, 5}, {25, 5, 1025}, {21, 5, 65}, {29, 5, 16385},
{16, 5, 3}, {24, 5, 513}, {20, 5, 33}, {28, 5, 8193}, {18, 5, 9}, {26, 5, 2049}, {22, 5, 129}, {64, 5, 0},
{16, 5, 2}, {23, 5, 385}, {19, 5, 25}, {27, 5, 6145}, {17, 5, 7}, {25, 5, 1537}, {21, 5, 97}, {29, 5, 24577},
{16, 5, 4}, {24, 5, 769}, {20, 5, 49}, {28, 5, 12289}, {18, 5, 13}, {26, 5, 3073}, {22, 5, 193}, {64, 5, 0}};

1561
base/poco/Foundation/src/inflate.c
View File

File diff suppressed because it is too large Load Diff

127
base/poco/Foundation/src/inflate.h
View File

@ -1,127 +0,0 @@
/* inflate.h -- internal inflate state definition
* Copyright (C) 1995-2016 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* WARNING: this file should *not* be used by applications. It is
part of the implementation of the compression library and is
subject to change. Applications should only use zlib.h.
*/
/* define NO_GZIP when compiling if you want to disable gzip header and
trailer decoding by inflate(). NO_GZIP would be used to avoid linking in
the crc code when it is not needed. For shared libraries, gzip decoding
should be left enabled. */
#ifndef NO_GZIP
# define GUNZIP
#endif
/* Possible inflate modes between inflate() calls */
typedef enum
{
HEAD = 16180, /* i: waiting for magic header */
FLAGS, /* i: waiting for method and flags (gzip) */
TIME, /* i: waiting for modification time (gzip) */
OS, /* i: waiting for extra flags and operating system (gzip) */
EXLEN, /* i: waiting for extra length (gzip) */
EXTRA, /* i: waiting for extra bytes (gzip) */
NAME, /* i: waiting for end of file name (gzip) */
COMMENT, /* i: waiting for end of comment (gzip) */
HCRC, /* i: waiting for header crc (gzip) */
DICTID, /* i: waiting for dictionary check value */
DICT, /* waiting for inflateSetDictionary() call */
TYPE, /* i: waiting for type bits, including last-flag bit */
TYPEDO, /* i: same, but skip check to exit inflate on new block */
STORED, /* i: waiting for stored size (length and complement) */
COPY_, /* i/o: same as COPY below, but only first time in */
COPY, /* i/o: waiting for input or output to copy stored block */
TABLE, /* i: waiting for dynamic block table lengths */
LENLENS, /* i: waiting for code length code lengths */
CODELENS, /* i: waiting for length/lit and distance code lengths */
LEN_, /* i: same as LEN below, but only first time in */
LEN, /* i: waiting for length/lit/eob code */
LENEXT, /* i: waiting for length extra bits */
DIST, /* i: waiting for distance code */
DISTEXT, /* i: waiting for distance extra bits */
MATCH, /* o: waiting for output space to copy string */
LIT, /* o: waiting for output space to write literal */
CHECK, /* i: waiting for 32-bit check value */
LENGTH, /* i: waiting for 32-bit length (gzip) */
DONE, /* finished check, done -- remain here until reset */
BAD, /* got a data error -- remain here until reset */
MEM, /* got an inflate() memory error -- remain here until reset */
SYNC /* looking for synchronization bytes to restart inflate() */
} inflate_mode;
/*
State transitions between above modes -
(most modes can go to BAD or MEM on error -- not shown for clarity)
Process header:
HEAD -> (gzip) or (zlib) or (raw)
(gzip) -> FLAGS -> TIME -> OS -> EXLEN -> EXTRA -> NAME -> COMMENT ->
HCRC -> TYPE
(zlib) -> DICTID or TYPE
DICTID -> DICT -> TYPE
(raw) -> TYPEDO
Read deflate blocks:
TYPE -> TYPEDO -> STORED or TABLE or LEN_ or CHECK
STORED -> COPY_ -> COPY -> TYPE
TABLE -> LENLENS -> CODELENS -> LEN_
LEN_ -> LEN
Read deflate codes in fixed or dynamic block:
LEN -> LENEXT or LIT or TYPE
LENEXT -> DIST -> DISTEXT -> MATCH -> LEN
LIT -> LEN
Process trailer:
CHECK -> LENGTH -> DONE
*/
/* State maintained between inflate() calls -- approximately 7K bytes, not
including the allocated sliding window, which is up to 32K bytes. */
struct inflate_state
{
z_streamp strm; /* pointer back to this zlib stream */
inflate_mode mode; /* current inflate mode */
int last; /* true if processing last block */
int wrap; /* bit 0 true for zlib, bit 1 true for gzip,
bit 2 true to validate check value */
int havedict; /* true if dictionary provided */
int flags; /* gzip header method and flags (0 if zlib) */
unsigned dmax; /* zlib header max distance (INFLATE_STRICT) */
unsigned long check; /* protected copy of check value */
unsigned long total; /* protected copy of output count */
gz_headerp head; /* where to save gzip header information */
/* sliding window */
unsigned wbits; /* log base 2 of requested window size */
unsigned wsize; /* window size or zero if not using window */
unsigned whave; /* valid bytes in the window */
unsigned wnext; /* window write index */
unsigned char FAR * window; /* allocated sliding window, if needed */
/* bit accumulator */
unsigned long hold; /* input bit accumulator */
unsigned bits; /* number of bits in "in" */
/* for string and stored block copying */
unsigned length; /* literal or length of data to copy */
unsigned offset; /* distance back to copy string from */
/* for table and code decoding */
unsigned extra; /* extra bits needed */
/* fixed and dynamic code tables */
code const FAR * lencode; /* starting table for length/literal codes */
code const FAR * distcode; /* starting table for distance codes */
unsigned lenbits; /* index bits for lencode */
unsigned distbits; /* index bits for distcode */
/* dynamic table building */
unsigned ncode; /* number of code length code lengths */
unsigned nlen; /* number of length code lengths */
unsigned ndist; /* number of distance code lengths */
unsigned have; /* number of code lengths in lens[] */
code FAR * next; /* next available space in codes[] */
unsigned short lens[320]; /* temporary storage for code lengths */
unsigned short work[288]; /* work area for code table building */
code codes[ENOUGH]; /* space for code tables */
int sane; /* if false, allow invalid distance too far */
int back; /* bits back of last unprocessed length/lit */
unsigned was; /* initial length of match */
};

304
base/poco/Foundation/src/inftrees.c
View File

@ -1,304 +0,0 @@
/* inftrees.c -- generate Huffman trees for efficient decoding
* Copyright (C) 1995-2017 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
#include "zutil.h"
#include "inftrees.h"
#define MAXBITS 15
const char inflate_copyright[] =
" inflate 1.2.11 Copyright 1995-2017 Mark Adler ";
/*
If you use the zlib library in a product, an acknowledgment is welcome
in the documentation of your product. If for some reason you cannot
include such an acknowledgment, I would appreciate that you keep this
copyright string in the executable of your product.
*/
/*
Build a set of tables to decode the provided canonical Huffman code.
The code lengths are lens[0..codes-1]. The result starts at *table,
whose indices are 0..2^bits-1. work is a writable array of at least
lens shorts, which is used as a work area. type is the type of code
to be generated, CODES, LENS, or DISTS. On return, zero is success,
-1 is an invalid code, and +1 means that ENOUGH isn't enough. table
on return points to the next available entry's address. bits is the
requested root table index bits, and on return it is the actual root
table index bits. It will differ if the request is greater than the
longest code or if it is less than the shortest code.
*/
int ZLIB_INTERNAL inflate_table(type, lens, codes, table, bits, work)
codetype type;
unsigned short FAR *lens;
unsigned codes;
code FAR * FAR *table;
unsigned FAR *bits;
unsigned short FAR *work;
{
unsigned len; /* a code's length in bits */
unsigned sym; /* index of code symbols */
unsigned min, max; /* minimum and maximum code lengths */
unsigned root; /* number of index bits for root table */
unsigned curr; /* number of index bits for current table */
unsigned drop; /* code bits to drop for sub-table */
int left; /* number of prefix codes available */
unsigned used; /* code entries in table used */
unsigned huff; /* Huffman code */
unsigned incr; /* for incrementing code, index */
unsigned fill; /* index for replicating entries */
unsigned low; /* low bits for current root entry */
unsigned mask; /* mask for low root bits */
code here; /* table entry for duplication */
code FAR *next; /* next available space in table */
const unsigned short FAR *base; /* base value table to use */
const unsigned short FAR *extra; /* extra bits table to use */
unsigned match; /* use base and extra for symbol >= match */
unsigned short count[MAXBITS+1]; /* number of codes of each length */
unsigned short offs[MAXBITS+1]; /* offsets in table for each length */
static const unsigned short lbase[31] = { /* Length codes 257..285 base */
3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31,
35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0};
static const unsigned short lext[31] = { /* Length codes 257..285 extra */
16, 16, 16, 16, 16, 16, 16, 16, 17, 17, 17, 17, 18, 18, 18, 18,
19, 19, 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 16, 77, 202};
static const unsigned short dbase[32] = { /* Distance codes 0..29 base */
1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193,
257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145,
8193, 12289, 16385, 24577, 0, 0};
static const unsigned short dext[32] = { /* Distance codes 0..29 extra */
16, 16, 16, 16, 17, 17, 18, 18, 19, 19, 20, 20, 21, 21, 22, 22,
23, 23, 24, 24, 25, 25, 26, 26, 27, 27,
28, 28, 29, 29, 64, 64};
/*
Process a set of code lengths to create a canonical Huffman code. The
code lengths are lens[0..codes-1]. Each length corresponds to the
symbols 0..codes-1. The Huffman code is generated by first sorting the
symbols by length from short to long, and retaining the symbol order
for codes with equal lengths. Then the code starts with all zero bits
for the first code of the shortest length, and the codes are integer
increments for the same length, and zeros are appended as the length
increases. For the deflate format, these bits are stored backwards
from their more natural integer increment ordering, and so when the
decoding tables are built in the large loop below, the integer codes
are incremented backwards.
This routine assumes, but does not check, that all of the entries in
lens[] are in the range 0..MAXBITS. The caller must assure this.
1..MAXBITS is interpreted as that code length. zero means that that
symbol does not occur in this code.
The codes are sorted by computing a count of codes for each length,
creating from that a table of starting indices for each length in the
sorted table, and then entering the symbols in order in the sorted
table. The sorted table is work[], with that space being provided by
the caller.
The length counts are used for other purposes as well, i.e. finding
the minimum and maximum length codes, determining if there are any
codes at all, checking for a valid set of lengths, and looking ahead
at length counts to determine sub-table sizes when building the
decoding tables.
*/
/* accumulate lengths for codes (assumes lens[] all in 0..MAXBITS) */
for (len = 0; len <= MAXBITS; len++)
count[len] = 0;
for (sym = 0; sym < codes; sym++)
count[lens[sym]]++;
/* bound code lengths, force root to be within code lengths */
root = *bits;
for (max = MAXBITS; max >= 1; max--)
if (count[max] != 0) break;
if (root > max) root = max;
if (max == 0) { /* no symbols to code at all */
here.op = (unsigned char)64; /* invalid code marker */
here.bits = (unsigned char)1;
here.val = (unsigned short)0;
*(*table)++ = here; /* make a table to force an error */
*(*table)++ = here;
*bits = 1;
return 0; /* no symbols, but wait for decoding to report error */
}
for (min = 1; min < max; min++)
if (count[min] != 0) break;
if (root < min) root = min;
/* check for an over-subscribed or incomplete set of lengths */
left = 1;
for (len = 1; len <= MAXBITS; len++) {
left <<= 1;
left -= count[len];
if (left < 0) return -1; /* over-subscribed */
}
if (left > 0 && (type == CODES || max != 1))
return -1; /* incomplete set */
/* generate offsets into symbol table for each length for sorting */
offs[1] = 0;
for (len = 1; len < MAXBITS; len++)
offs[len + 1] = offs[len] + count[len];
/* sort symbols by length, by symbol order within each length */
for (sym = 0; sym < codes; sym++)
if (lens[sym] != 0) work[offs[lens[sym]]++] = (unsigned short)sym;
/*
Create and fill in decoding tables. In this loop, the table being
filled is at next and has curr index bits. The code being used is huff
with length len. That code is converted to an index by dropping drop
bits off of the bottom. For codes where len is less than drop + curr,
those top drop + curr - len bits are incremented through all values to
fill the table with replicated entries.
root is the number of index bits for the root table. When len exceeds
root, sub-tables are created pointed to by the root entry with an index
of the low root bits of huff. This is saved in low to check for when a
new sub-table should be started. drop is zero when the root table is
being filled, and drop is root when sub-tables are being filled.
When a new sub-table is needed, it is necessary to look ahead in the
code lengths to determine what size sub-table is needed. The length
counts are used for this, and so count[] is decremented as codes are
entered in the tables.
used keeps track of how many table entries have been allocated from the
provided *table space. It is checked for LENS and DIST tables against
the constants ENOUGH_LENS and ENOUGH_DISTS to guard against changes in
the initial root table size constants. See the comments in inftrees.h
for more information.
sym increments through all symbols, and the loop terminates when
all codes of length max, i.e. all codes, have been processed. This
routine permits incomplete codes, so another loop after this one fills
in the rest of the decoding tables with invalid code markers.
*/
/* set up for code type */
switch (type) {
case CODES:
base = extra = work; /* dummy value--not used */
match = 20;
break;
case LENS:
base = lbase;
extra = lext;
match = 257;
break;
default: /* DISTS */
base = dbase;
extra = dext;
match = 0;
}
/* initialize state for loop */
huff = 0; /* starting code */
sym = 0; /* starting code symbol */
len = min; /* starting code length */
next = *table; /* current table to fill in */
curr = root; /* current table index bits */
drop = 0; /* current bits to drop from code for index */
low = (unsigned)(-1); /* trigger new sub-table when len > root */
used = 1U << root; /* use root table entries */
mask = used - 1; /* mask for comparing low */
/* check available table space */
if ((type == LENS && used > ENOUGH_LENS) ||
(type == DISTS && used > ENOUGH_DISTS))
return 1;
/* process all codes and make table entries */
for (;;) {
/* create table entry */
here.bits = (unsigned char)(len - drop);
if (work[sym] + 1U < match) {
here.op = (unsigned char)0;
here.val = work[sym];
}
else if (work[sym] >= match) {
here.op = (unsigned char)(extra[work[sym] - match]);
here.val = base[work[sym] - match];
}
else {
here.op = (unsigned char)(32 + 64); /* end of block */
here.val = 0;
}
/* replicate for those indices with low len bits equal to huff */
incr = 1U << (len - drop);
fill = 1U << curr;
min = fill; /* save offset to next table */
do {
fill -= incr;
next[(huff >> drop) + fill] = here;
} while (fill != 0);
/* backwards increment the len-bit code huff */
incr = 1U << (len - 1);
while (huff & incr)
incr >>= 1;
if (incr != 0) {
huff &= incr - 1;
huff += incr;
}
else
huff = 0;
/* go to next symbol, update count, len */
sym++;
if (--(count[len]) == 0) {
if (len == max) break;
len = lens[work[sym]];
}
/* create new sub-table if needed */
if (len > root && (huff & mask) != low) {
/* if first time, transition to sub-tables */
if (drop == 0)
drop = root;
/* increment past last table */
next += min; /* here min is 1 << curr */
/* determine length of next table */
curr = len - drop;
left = (int)(1 << curr);
while (curr + drop < max) {
left -= count[curr + drop];
if (left <= 0) break;
curr++;
left <<= 1;
}
/* check for enough space */
used += 1U << curr;
if ((type == LENS && used > ENOUGH_LENS) ||
(type == DISTS && used > ENOUGH_DISTS))
return 1;
/* point entry in root table to sub-table */
low = huff & mask;
(*table)[low].op = (unsigned char)curr;
(*table)[low].bits = (unsigned char)root;
(*table)[low].val = (unsigned short)(next - *table);
}
}
/* fill in remaining table entry if code is incomplete (guaranteed to have
at most one remaining entry, since if the code is incomplete, the
maximum code length that was allowed to get this far is one bit) */
if (huff != 0) {
here.op = (unsigned char)64; /* invalid code marker */
here.bits = (unsigned char)(len - drop);
here.val = (unsigned short)0;
next[huff] = here;
}
/* set return parameters */
*table += used;
*bits = root;
return 0;
}

63
base/poco/Foundation/src/inftrees.h
View File

@ -1,63 +0,0 @@
/* inftrees.h -- header to use inftrees.c
* Copyright (C) 1995-2005, 2010 Mark Adler
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* WARNING: this file should *not* be used by applications. It is
part of the implementation of the compression library and is
subject to change. Applications should only use zlib.h.
*/
/* Structure for decoding tables. Each entry provides either the
information needed to do the operation requested by the code that
indexed that table entry, or it provides a pointer to another
table that indexes more bits of the code. op indicates whether
the entry is a pointer to another table, a literal, a length or
distance, an end-of-block, or an invalid code. For a table
pointer, the low four bits of op is the number of index bits of
that table. For a length or distance, the low four bits of op
is the number of extra bits to get after the code. bits is
the number of bits in this code or part of the code to drop off
of the bit buffer. val is the actual byte to output in the case
of a literal, the base length or distance, or the offset from
the current table to the next table. Each entry is four bytes. */
typedef struct
{
unsigned char op; /* operation, extra bits, table bits */
unsigned char bits; /* bits in this part of the code */
unsigned short val; /* offset in table or code value */
} code;
/* op values as set by inflate_table():
00000000 - literal
0000tttt - table link, tttt != 0 is the number of table index bits
0001eeee - length or distance, eeee is the number of extra bits
01100000 - end of block
01000000 - invalid code
*/
/* Maximum size of the dynamic table. The maximum number of code structures is
1444, which is the sum of 852 for literal/length codes and 592 for distance
codes. These values were found by exhaustive searches using the program
examples/enough.c found in the zlib distribution. The arguments to that
program are the number of symbols, the initial root table size, and the
maximum bit length of a code. "enough 286 9 15" for literal/length codes
returns returns 852, and "enough 30 6 15" for distance codes returns 592.
The initial root table size (9 or 6) is found in the fifth argument of the
inflate_table() calls in inflate.c and infback.c. If the root table size is
changed, then these maximum sizes would be need to be recalculated and
updated. */
#define ENOUGH_LENS 852
#define ENOUGH_DISTS 592
#define ENOUGH (ENOUGH_LENS + ENOUGH_DISTS)
/* Type of code to build for inflate_table() */
typedef enum
{
CODES,
LENS,
DISTS
} codetype;
int ZLIB_INTERNAL inflate_table
OF((codetype type, unsigned short FAR * lens, unsigned codes, code FAR * FAR * table, unsigned FAR * bits, unsigned short FAR * work));

87
base/poco/Foundation/src/pocomsg.mc
View File

@ -1,87 +0,0 @@
;//
;// pocomsg.mc[.h]
;//
;// The Poco message source/header file.
;//
;// NOTE: pocomsg.h is automatically generated from pocomsg.mc.
;// Never edit pocomsg.h directly!
;//
;// Copyright (c) 2004-2006, Applied Informatics Software Engineering GmbH.
;// and Contributors.
;//
;// Permission is hereby granted, free of charge, to any person or organization
;// obtaining a copy of the software and accompanying documentation covered by
;// this license (the "Software") to use, reproduce, display, distribute,
;// execute, and transmit the Software, and to prepare derivative works of the
;// Software, and to permit third-parties to whom the Software is furnished to
;// do so, all subject to the following:
;//
;// The copyright notices in the Software and this entire statement, including
;// the above license grant, this restriction and the following disclaimer,
;// must be included in all copies of the Software, in whole or in part, and
;// all derivative works of the Software, unless such copies or derivative
;// works are solely in the form of machine-executable object code generated by
;// a source language processor.
;//
;// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
;// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
;// FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
;// SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
;// FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
;// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
;// DEALINGS IN THE SOFTWARE.
;//
;//
;// Categories
;//
MessageId=0x1
SymbolicName=POCO_CTG_FATAL
Language=English
Fatal
.
MessageId=0x2
SymbolicName=POCO_CTG_CRITICAL
Language=English
Critical
.
MessageId=0x3
SymbolicName=POCO_CTG_ERROR
Language=English
Error
.
MessageId=0x4
SymbolicName=POCO_CTG_WARNING
Language=English
Warning
.
MessageId=0x5
SymbolicName=POCO_CTG_NOTICE
Language=English
Notice
.
MessageId=0x6
SymbolicName=POCO_CTG_INFORMATION
Language=English
Information
.
MessageId=0x7
SymbolicName=POCO_CTG_DEBUG
Language=English
Debug
.
MessageId=0x8
SymbolicName=POCO_CTG_TRACE
Language=English
Trace
.
;//
;// Event Identifiers
;//
MessageId=0x1000
SymbolicName=POCO_MSG_LOG
Language=English
%1
.

556
base/poco/Foundation/src/strtod.cc
View File

@ -1,556 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <stdarg.h>
#include <limits.h>
#include "strtod.h"
#include "bignum.h"
#include "cached-powers.h"
#include "ieee.h"
namespace double_conversion {
// 2^53 = 9007199254740992.
// Any integer with at most 15 decimal digits will hence fit into a double
// (which has a 53bit significand) without loss of precision.
static const int kMaxExactDoubleIntegerDecimalDigits = 15;
// 2^64 = 18446744073709551616 > 10^19
static const int kMaxUint64DecimalDigits = 19;
// Max double: 1.7976931348623157 x 10^308
// Min non-zero double: 4.9406564584124654 x 10^-324
// Any x >= 10^309 is interpreted as +infinity.
// Any x <= 10^-324 is interpreted as 0.
// Note that 2.5e-324 (despite being smaller than the min double) will be read
// as non-zero (equal to the min non-zero double).
static const int kMaxDecimalPower = 309;
static const int kMinDecimalPower = -324;
// 2^64 = 18446744073709551616
static const uint64_t kMaxUint64 = UINT64_2PART_C(0xFFFFFFFF, FFFFFFFF);
static const double exact_powers_of_ten[] = {
1.0, // 10^0
10.0,
100.0,
1000.0,
10000.0,
100000.0,
1000000.0,
10000000.0,
100000000.0,
1000000000.0,
10000000000.0, // 10^10
100000000000.0,
1000000000000.0,
10000000000000.0,
100000000000000.0,
1000000000000000.0,
10000000000000000.0,
100000000000000000.0,
1000000000000000000.0,
10000000000000000000.0,
100000000000000000000.0, // 10^20
1000000000000000000000.0,
// 10^22 = 0x21e19e0c9bab2400000 = 0x878678326eac9 * 2^22
10000000000000000000000.0
};
static const int kExactPowersOfTenSize = ARRAY_SIZE(exact_powers_of_ten);
// Maximum number of significant digits in the decimal representation.
// In fact the value is 772 (see conversions.cc), but to give us some margin
// we round up to 780.
static const int kMaxSignificantDecimalDigits = 780;
static Vector<const char> TrimLeadingZeros(Vector<const char> buffer) {
for (int i = 0; i < buffer.length(); i++) {
if (buffer[i] != '0') {
return buffer.SubVector(i, buffer.length());
}
}
return Vector<const char>(buffer.start(), 0);
}
static Vector<const char> TrimTrailingZeros(Vector<const char> buffer) {
for (int i = buffer.length() - 1; i >= 0; --i) {
if (buffer[i] != '0') {
return buffer.SubVector(0, i + 1);
}
}
return Vector<const char>(buffer.start(), 0);
}
static void CutToMaxSignificantDigits(Vector<const char> buffer,
int exponent,
char* significant_buffer,
int* significant_exponent) {
for (int i = 0; i < kMaxSignificantDecimalDigits - 1; ++i) {
significant_buffer[i] = buffer[i];
}
// The input buffer has been trimmed. Therefore the last digit must be
// different from '0'.
ASSERT(buffer[buffer.length() - 1] != '0');
// Set the last digit to be non-zero. This is sufficient to guarantee
// correct rounding.
significant_buffer[kMaxSignificantDecimalDigits - 1] = '1';
*significant_exponent =
exponent + (buffer.length() - kMaxSignificantDecimalDigits);
}
// Trims the buffer and cuts it to at most kMaxSignificantDecimalDigits.
// If possible the input-buffer is reused, but if the buffer needs to be
// modified (due to cutting), then the input needs to be copied into the
// buffer_copy_space.
static void TrimAndCut(Vector<const char> buffer, int exponent,
char* buffer_copy_space, int space_size,
Vector<const char>* trimmed, int* updated_exponent) {
Vector<const char> left_trimmed = TrimLeadingZeros(buffer);
Vector<const char> right_trimmed = TrimTrailingZeros(left_trimmed);
exponent += left_trimmed.length() - right_trimmed.length();
if (right_trimmed.length() > kMaxSignificantDecimalDigits) {
(void) space_size; // Mark variable as used.
ASSERT(space_size >= kMaxSignificantDecimalDigits);
CutToMaxSignificantDigits(right_trimmed, exponent,
buffer_copy_space, updated_exponent);
*trimmed = Vector<const char>(buffer_copy_space,
kMaxSignificantDecimalDigits);
} else {
*trimmed = right_trimmed;
*updated_exponent = exponent;
}
}
// Reads digits from the buffer and converts them to a uint64.
// Reads in as many digits as fit into a uint64.
// When the string starts with "1844674407370955161" no further digit is read.
// Since 2^64 = 18446744073709551616 it would still be possible read another
// digit if it was less or equal than 6, but this would complicate the code.
static uint64_t ReadUint64(Vector<const char> buffer,
int* number_of_read_digits) {
uint64_t result = 0;
int i = 0;
while (i < buffer.length() && result <= (kMaxUint64 / 10 - 1)) {
int digit = buffer[i++] - '0';
ASSERT(0 <= digit && digit <= 9);
result = 10 * result + digit;
}
*number_of_read_digits = i;
return result;
}
// Reads a DiyFp from the buffer.
// The returned DiyFp is not necessarily normalized.
// If remaining_decimals is zero then the returned DiyFp is accurate.
// Otherwise it has been rounded and has error of at most 1/2 ulp.
static void ReadDiyFp(Vector<const char> buffer,
DiyFp* result,
int* remaining_decimals) {
int read_digits;
uint64_t significand = ReadUint64(buffer, &read_digits);
if (buffer.length() == read_digits) {
*result = DiyFp(significand, 0);
*remaining_decimals = 0;
} else {
// Round the significand.
if (buffer[read_digits] >= '5') {
significand++;
}
// Compute the binary exponent.
int exponent = 0;
*result = DiyFp(significand, exponent);
*remaining_decimals = buffer.length() - read_digits;
}
}
static bool DoubleStrtod(Vector<const char> trimmed,
int exponent,
double* result) {
#if !defined(DOUBLE_CONVERSION_CORRECT_DOUBLE_OPERATIONS)
// On x86 the floating-point stack can be 64 or 80 bits wide. If it is
// 80 bits wide (as is the case on Linux) then double-rounding occurs and the
// result is not accurate.
// We know that Windows32 uses 64 bits and is therefore accurate.
// Note that the ARM simulator is compiled for 32bits. It therefore exhibits
// the same problem.
return false;
#endif
if (trimmed.length() <= kMaxExactDoubleIntegerDecimalDigits) {
int read_digits;
// The trimmed input fits into a double.
// If the 10^exponent (resp. 10^-exponent) fits into a double too then we
// can compute the result-double simply by multiplying (resp. dividing) the
// two numbers.
// This is possible because IEEE guarantees that floating-point operations
// return the best possible approximation.
if (exponent < 0 && -exponent < kExactPowersOfTenSize) {
// 10^-exponent fits into a double.
*result = static_cast<double>(ReadUint64(trimmed, &read_digits));
ASSERT(read_digits == trimmed.length());
*result /= exact_powers_of_ten[-exponent];
return true;
}
if (0 <= exponent && exponent < kExactPowersOfTenSize) {
// 10^exponent fits into a double.
*result = static_cast<double>(ReadUint64(trimmed, &read_digits));
ASSERT(read_digits == trimmed.length());
*result *= exact_powers_of_ten[exponent];
return true;
}
int remaining_digits =
kMaxExactDoubleIntegerDecimalDigits - trimmed.length();
if ((0 <= exponent) &&
(exponent - remaining_digits < kExactPowersOfTenSize)) {
// The trimmed string was short and we can multiply it with
// 10^remaining_digits. As a result the remaining exponent now fits
// into a double too.
*result = static_cast<double>(ReadUint64(trimmed, &read_digits));
ASSERT(read_digits == trimmed.length());
*result *= exact_powers_of_ten[remaining_digits];
*result *= exact_powers_of_ten[exponent - remaining_digits];
return true;
}
}
return false;
}
// Returns 10^exponent as an exact DiyFp.
// The given exponent must be in the range [1; kDecimalExponentDistance[.
static DiyFp AdjustmentPowerOfTen(int exponent) {
ASSERT(0 < exponent);
ASSERT(exponent < PowersOfTenCache::kDecimalExponentDistance);
// Simply hardcode the remaining powers for the given decimal exponent
// distance.
ASSERT(PowersOfTenCache::kDecimalExponentDistance == 8);
switch (exponent) {
case 1: return DiyFp(UINT64_2PART_C(0xa0000000, 00000000), -60);
case 2: return DiyFp(UINT64_2PART_C(0xc8000000, 00000000), -57);
case 3: return DiyFp(UINT64_2PART_C(0xfa000000, 00000000), -54);
case 4: return DiyFp(UINT64_2PART_C(0x9c400000, 00000000), -50);
case 5: return DiyFp(UINT64_2PART_C(0xc3500000, 00000000), -47);
case 6: return DiyFp(UINT64_2PART_C(0xf4240000, 00000000), -44);
case 7: return DiyFp(UINT64_2PART_C(0x98968000, 00000000), -40);
default:
UNREACHABLE();
return DiyFp(0, 0);
}
}
// If the function returns true then the result is the correct double.
// Otherwise it is either the correct double or the double that is just below
// the correct double.
static bool DiyFpStrtod(Vector<const char> buffer,
int exponent,
double* result) {
DiyFp input;
int remaining_decimals;
ReadDiyFp(buffer, &input, &remaining_decimals);
// Since we may have dropped some digits the input is not accurate.
// If remaining_decimals is different than 0 than the error is at most
// .5 ulp (unit in the last place).
// We don't want to deal with fractions and therefore keep a common
// denominator.
const int kDenominatorLog = 3;
const int kDenominator = 1 << kDenominatorLog;
// Move the remaining decimals into the exponent.
exponent += remaining_decimals;
int error = (remaining_decimals == 0 ? 0 : kDenominator / 2);
int old_e = input.e();
input.Normalize();
error <<= old_e - input.e();
ASSERT(exponent <= PowersOfTenCache::kMaxDecimalExponent);
if (exponent < PowersOfTenCache::kMinDecimalExponent) {
*result = 0.0;
return true;
}
DiyFp cached_power;
int cached_decimal_exponent;
PowersOfTenCache::GetCachedPowerForDecimalExponent(exponent,
&cached_power,
&cached_decimal_exponent);
if (cached_decimal_exponent != exponent) {
int adjustment_exponent = exponent - cached_decimal_exponent;
DiyFp adjustment_power = AdjustmentPowerOfTen(adjustment_exponent);
input.Multiply(adjustment_power);
if (kMaxUint64DecimalDigits - buffer.length() >= adjustment_exponent) {
// The product of input with the adjustment power fits into a 64 bit
// integer.
ASSERT(DiyFp::kSignificandSize == 64);
} else {
// The adjustment power is exact. There is hence only an error of 0.5.
error += kDenominator / 2;
}
}
input.Multiply(cached_power);
// The error introduced by a multiplication of a*b equals
// error_a + error_b + error_a*error_b/2^64 + 0.5
// Substituting a with 'input' and b with 'cached_power' we have
// error_b = 0.5 (all cached powers have an error of less than 0.5 ulp),
// error_ab = 0 or 1 / kDenominator > error_a*error_b/ 2^64
int error_b = kDenominator / 2;
int error_ab = (error == 0 ? 0 : 1); // We round up to 1.
int fixed_error = kDenominator / 2;
error += error_b + error_ab + fixed_error;
old_e = input.e();
input.Normalize();
error <<= old_e - input.e();
// See if the double's significand changes if we add/subtract the error.
int order_of_magnitude = DiyFp::kSignificandSize + input.e();
int effective_significand_size =
Double::SignificandSizeForOrderOfMagnitude(order_of_magnitude);
int precision_digits_count =
DiyFp::kSignificandSize - effective_significand_size;
if (precision_digits_count + kDenominatorLog >= DiyFp::kSignificandSize) {
// This can only happen for very small denormals. In this case the
// half-way multiplied by the denominator exceeds the range of an uint64.
// Simply shift everything to the right.
int shift_amount = (precision_digits_count + kDenominatorLog) -
DiyFp::kSignificandSize + 1;
input.set_f(input.f() >> shift_amount);
input.set_e(input.e() + shift_amount);
// We add 1 for the lost precision of error, and kDenominator for
// the lost precision of input.f().
error = (error >> shift_amount) + 1 + kDenominator;
precision_digits_count -= shift_amount;
}
// We use uint64_ts now. This only works if the DiyFp uses uint64_ts too.
ASSERT(DiyFp::kSignificandSize == 64);
ASSERT(precision_digits_count < 64);
uint64_t one64 = 1;
uint64_t precision_bits_mask = (one64 << precision_digits_count) - 1;
uint64_t precision_bits = input.f() & precision_bits_mask;
uint64_t half_way = one64 << (precision_digits_count - 1);
precision_bits *= kDenominator;
half_way *= kDenominator;
DiyFp rounded_input(input.f() >> precision_digits_count,
input.e() + precision_digits_count);
if (precision_bits >= half_way + error) {
rounded_input.set_f(rounded_input.f() + 1);
}
// If the last_bits are too close to the half-way case than we are too
// inaccurate and round down. In this case we return false so that we can
// fall back to a more precise algorithm.
*result = Double(rounded_input).value();
if (half_way - error < precision_bits && precision_bits < half_way + error) {
// Too imprecise. The caller will have to fall back to a slower version.
// However the returned number is guaranteed to be either the correct
// double, or the next-lower double.
return false;
} else {
return true;
}
}
// Returns
// - -1 if buffer*10^exponent < diy_fp.
// - 0 if buffer*10^exponent == diy_fp.
// - +1 if buffer*10^exponent > diy_fp.
// Preconditions:
// buffer.length() + exponent <= kMaxDecimalPower + 1
// buffer.length() + exponent > kMinDecimalPower
// buffer.length() <= kMaxDecimalSignificantDigits
static int CompareBufferWithDiyFp(Vector<const char> buffer,
int exponent,
DiyFp diy_fp) {
ASSERT(buffer.length() + exponent <= kMaxDecimalPower + 1);
ASSERT(buffer.length() + exponent > kMinDecimalPower);
ASSERT(buffer.length() <= kMaxSignificantDecimalDigits);
// Make sure that the Bignum will be able to hold all our numbers.
// Our Bignum implementation has a separate field for exponents. Shifts will
// consume at most one bigit (< 64 bits).
// ln(10) == 3.3219...
ASSERT(((kMaxDecimalPower + 1) * 333 / 100) < Bignum::kMaxSignificantBits);
Bignum buffer_bignum;
Bignum diy_fp_bignum;
buffer_bignum.AssignDecimalString(buffer);
diy_fp_bignum.AssignUInt64(diy_fp.f());
if (exponent >= 0) {
buffer_bignum.MultiplyByPowerOfTen(exponent);
} else {
diy_fp_bignum.MultiplyByPowerOfTen(-exponent);
}
if (diy_fp.e() > 0) {
diy_fp_bignum.ShiftLeft(diy_fp.e());
} else {
buffer_bignum.ShiftLeft(-diy_fp.e());
}
return Bignum::Compare(buffer_bignum, diy_fp_bignum);
}
// Returns true if the guess is the correct double.
// Returns false, when guess is either correct or the next-lower double.
static bool ComputeGuess(Vector<const char> trimmed, int exponent,
double* guess) {
if (trimmed.length() == 0) {
*guess = 0.0;
return true;
}
if (exponent + trimmed.length() - 1 >= kMaxDecimalPower) {
*guess = Double::Infinity();
return true;
}
if (exponent + trimmed.length() <= kMinDecimalPower) {
*guess = 0.0;
return true;
}
if (DoubleStrtod(trimmed, exponent, guess) ||
DiyFpStrtod(trimmed, exponent, guess)) {
return true;
}
if (*guess == Double::Infinity()) {
return true;
}
return false;
}
double Strtod(Vector<const char> buffer, int exponent) {
char copy_buffer[kMaxSignificantDecimalDigits];
Vector<const char> trimmed;
int updated_exponent;
TrimAndCut(buffer, exponent, copy_buffer, kMaxSignificantDecimalDigits,
&trimmed, &updated_exponent);
exponent = updated_exponent;
double guess;
bool is_correct = ComputeGuess(trimmed, exponent, &guess);
if (is_correct) return guess;
DiyFp upper_boundary = Double(guess).UpperBoundary();
int comparison = CompareBufferWithDiyFp(trimmed, exponent, upper_boundary);
if (comparison < 0) {
return guess;
} else if (comparison > 0) {
return Double(guess).NextDouble();
} else if ((Double(guess).Significand() & 1) == 0) {
// Round towards even.
return guess;
} else {
return Double(guess).NextDouble();
}
}
float Strtof(Vector<const char> buffer, int exponent) {
char copy_buffer[kMaxSignificantDecimalDigits];
Vector<const char> trimmed;
int updated_exponent;
TrimAndCut(buffer, exponent, copy_buffer, kMaxSignificantDecimalDigits,
&trimmed, &updated_exponent);
exponent = updated_exponent;
double double_guess;
bool is_correct = ComputeGuess(trimmed, exponent, &double_guess);
float float_guess = static_cast<float>(double_guess);
if (float_guess == double_guess) {
// This shortcut triggers for integer values.
return float_guess;
}
// We must catch double-rounding. Say the double has been rounded up, and is
// now a boundary of a float, and rounds up again. This is why we have to
// look at previous too.
// Example (in decimal numbers):
// input: 12349
// high-precision (4 digits): 1235
// low-precision (3 digits):
// when read from input: 123
// when rounded from high precision: 124.
// To do this we simply look at the neighbors of the correct result and see
// if they would round to the same float. If the guess is not correct we have
// to look at four values (since two different doubles could be the correct
// double).
double double_next = Double(double_guess).NextDouble();
double double_previous = Double(double_guess).PreviousDouble();
float f1 = static_cast<float>(double_previous);
float f2 = float_guess;
float f3 = static_cast<float>(double_next);
float f4;
if (is_correct) {
f4 = f3;
} else {
double double_next2 = Double(double_next).NextDouble();
f4 = static_cast<float>(double_next2);
}
(void) f2; // Mark variable as used.
ASSERT(f1 <= f2 && f2 <= f3 && f3 <= f4);
// If the guess doesn't lie near a single-precision boundary we can simply
// return its float-value.
if (f1 == f4) {
return float_guess;
}
ASSERT((f1 != f2 && f2 == f3 && f3 == f4) ||
(f1 == f2 && f2 != f3 && f3 == f4) ||
(f1 == f2 && f2 == f3 && f3 != f4));
// guess and next are the two possible candidates (in the same way that
// double_guess was the lower candidate for a double-precision guess).
float guess = f1;
float next = f4;
DiyFp upper_boundary;
if (guess == 0.0f) {
float min_float = 1e-45f;
upper_boundary = Double(static_cast<double>(min_float) / 2).AsDiyFp();
} else {
upper_boundary = Single(guess).UpperBoundary();
}
int comparison = CompareBufferWithDiyFp(trimmed, exponent, upper_boundary);
if (comparison < 0) {
return guess;
} else if (comparison > 0) {
return next;
} else if ((Single(guess).Significand() & 1) == 0) {
// Round towards even.
return guess;
} else {
return next;
}
}
} // namespace double_conversion

46
base/poco/Foundation/src/strtod.h
View File

@ -1,46 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_STRTOD_H_
#define DOUBLE_CONVERSION_STRTOD_H_
#include "utils.h"
namespace double_conversion
{
// The buffer must only contain digits in the range [0-9]. It must not
// contain a dot or a sign. It must not start with '0', and must not be empty.
double Strtod(Vector<const char> buffer, int exponent);
// The buffer must only contain digits in the range [0-9]. It must not
// contain a dot or a sign. It must not start with '0', and must not be empty.
float Strtof(Vector<const char> buffer, int exponent);
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_STRTOD_H_

1203
base/poco/Foundation/src/trees.c
View File

File diff suppressed because it is too large Load Diff

74
base/poco/Foundation/src/trees.h
View File

@ -1,74 +0,0 @@
/* header created automatically with -DGEN_TREES_H */
local const ct_data static_ltree[L_CODES + 2]
= {{{12}, {8}}, {{140}, {8}}, {{76}, {8}}, {{204}, {8}}, {{44}, {8}}, {{172}, {8}}, {{108}, {8}}, {{236}, {8}}, {{28}, {8}},
{{156}, {8}}, {{92}, {8}}, {{220}, {8}}, {{60}, {8}}, {{188}, {8}}, {{124}, {8}}, {{252}, {8}}, {{2}, {8}}, {{130}, {8}},
{{66}, {8}}, {{194}, {8}}, {{34}, {8}}, {{162}, {8}}, {{98}, {8}}, {{226}, {8}}, {{18}, {8}}, {{146}, {8}}, {{82}, {8}},
{{210}, {8}}, {{50}, {8}}, {{178}, {8}}, {{114}, {8}}, {{242}, {8}}, {{10}, {8}}, {{138}, {8}}, {{74}, {8}}, {{202}, {8}},
{{42}, {8}}, {{170}, {8}}, {{106}, {8}}, {{234}, {8}}, {{26}, {8}}, {{154}, {8}}, {{90}, {8}}, {{218}, {8}}, {{58}, {8}},
{{186}, {8}}, {{122}, {8}}, {{250}, {8}}, {{6}, {8}}, {{134}, {8}}, {{70}, {8}}, {{198}, {8}}, {{38}, {8}}, {{166}, {8}},
{{102}, {8}}, {{230}, {8}}, {{22}, {8}}, {{150}, {8}}, {{86}, {8}}, {{214}, {8}}, {{54}, {8}}, {{182}, {8}}, {{118}, {8}},
{{246}, {8}}, {{14}, {8}}, {{142}, {8}}, {{78}, {8}}, {{206}, {8}}, {{46}, {8}}, {{174}, {8}}, {{110}, {8}}, {{238}, {8}},
{{30}, {8}}, {{158}, {8}}, {{94}, {8}}, {{222}, {8}}, {{62}, {8}}, {{190}, {8}}, {{126}, {8}}, {{254}, {8}}, {{1}, {8}},
{{129}, {8}}, {{65}, {8}}, {{193}, {8}}, {{33}, {8}}, {{161}, {8}}, {{97}, {8}}, {{225}, {8}}, {{17}, {8}}, {{145}, {8}},
{{81}, {8}}, {{209}, {8}}, {{49}, {8}}, {{177}, {8}}, {{113}, {8}}, {{241}, {8}}, {{9}, {8}}, {{137}, {8}}, {{73}, {8}},
{{201}, {8}}, {{41}, {8}}, {{169}, {8}}, {{105}, {8}}, {{233}, {8}}, {{25}, {8}}, {{153}, {8}}, {{89}, {8}}, {{217}, {8}},
{{57}, {8}}, {{185}, {8}}, {{121}, {8}}, {{249}, {8}}, {{5}, {8}}, {{133}, {8}}, {{69}, {8}}, {{197}, {8}}, {{37}, {8}},
{{165}, {8}}, {{101}, {8}}, {{229}, {8}}, {{21}, {8}}, {{149}, {8}}, {{85}, {8}}, {{213}, {8}}, {{53}, {8}}, {{181}, {8}},
{{117}, {8}}, {{245}, {8}}, {{13}, {8}}, {{141}, {8}}, {{77}, {8}}, {{205}, {8}}, {{45}, {8}}, {{173}, {8}}, {{109}, {8}},
{{237}, {8}}, {{29}, {8}}, {{157}, {8}}, {{93}, {8}}, {{221}, {8}}, {{61}, {8}}, {{189}, {8}}, {{125}, {8}}, {{253}, {8}},
{{19}, {9}}, {{275}, {9}}, {{147}, {9}}, {{403}, {9}}, {{83}, {9}}, {{339}, {9}}, {{211}, {9}}, {{467}, {9}}, {{51}, {9}},
{{307}, {9}}, {{179}, {9}}, {{435}, {9}}, {{115}, {9}}, {{371}, {9}}, {{243}, {9}}, {{499}, {9}}, {{11}, {9}}, {{267}, {9}},
{{139}, {9}}, {{395}, {9}}, {{75}, {9}}, {{331}, {9}}, {{203}, {9}}, {{459}, {9}}, {{43}, {9}}, {{299}, {9}}, {{171}, {9}},
{{427}, {9}}, {{107}, {9}}, {{363}, {9}}, {{235}, {9}}, {{491}, {9}}, {{27}, {9}}, {{283}, {9}}, {{155}, {9}}, {{411}, {9}},
{{91}, {9}}, {{347}, {9}}, {{219}, {9}}, {{475}, {9}}, {{59}, {9}}, {{315}, {9}}, {{187}, {9}}, {{443}, {9}}, {{123}, {9}},
{{379}, {9}}, {{251}, {9}}, {{507}, {9}}, {{7}, {9}}, {{263}, {9}}, {{135}, {9}}, {{391}, {9}}, {{71}, {9}}, {{327}, {9}},
{{199}, {9}}, {{455}, {9}}, {{39}, {9}}, {{295}, {9}}, {{167}, {9}}, {{423}, {9}}, {{103}, {9}}, {{359}, {9}}, {{231}, {9}},
{{487}, {9}}, {{23}, {9}}, {{279}, {9}}, {{151}, {9}}, {{407}, {9}}, {{87}, {9}}, {{343}, {9}}, {{215}, {9}}, {{471}, {9}},
{{55}, {9}}, {{311}, {9}}, {{183}, {9}}, {{439}, {9}}, {{119}, {9}}, {{375}, {9}}, {{247}, {9}}, {{503}, {9}}, {{15}, {9}},
{{271}, {9}}, {{143}, {9}}, {{399}, {9}}, {{79}, {9}}, {{335}, {9}}, {{207}, {9}}, {{463}, {9}}, {{47}, {9}}, {{303}, {9}},
{{175}, {9}}, {{431}, {9}}, {{111}, {9}}, {{367}, {9}}, {{239}, {9}}, {{495}, {9}}, {{31}, {9}}, {{287}, {9}}, {{159}, {9}},
{{415}, {9}}, {{95}, {9}}, {{351}, {9}}, {{223}, {9}}, {{479}, {9}}, {{63}, {9}}, {{319}, {9}}, {{191}, {9}}, {{447}, {9}},
{{127}, {9}}, {{383}, {9}}, {{255}, {9}}, {{511}, {9}}, {{0}, {7}}, {{64}, {7}}, {{32}, {7}}, {{96}, {7}}, {{16}, {7}},
{{80}, {7}}, {{48}, {7}}, {{112}, {7}}, {{8}, {7}}, {{72}, {7}}, {{40}, {7}}, {{104}, {7}}, {{24}, {7}}, {{88}, {7}},
{{56}, {7}}, {{120}, {7}}, {{4}, {7}}, {{68}, {7}}, {{36}, {7}}, {{100}, {7}}, {{20}, {7}}, {{84}, {7}}, {{52}, {7}},
{{116}, {7}}, {{3}, {8}}, {{131}, {8}}, {{67}, {8}}, {{195}, {8}}, {{35}, {8}}, {{163}, {8}}, {{99}, {8}}, {{227}, {8}}};
local const ct_data static_dtree[D_CODES]
= {{{0}, {5}}, {{16}, {5}}, {{8}, {5}}, {{24}, {5}}, {{4}, {5}}, {{20}, {5}}, {{12}, {5}}, {{28}, {5}}, {{2}, {5}}, {{18}, {5}},
{{10}, {5}}, {{26}, {5}}, {{6}, {5}}, {{22}, {5}}, {{14}, {5}}, {{30}, {5}}, {{1}, {5}}, {{17}, {5}}, {{9}, {5}}, {{25}, {5}},
{{5}, {5}}, {{21}, {5}}, {{13}, {5}}, {{29}, {5}}, {{3}, {5}}, {{19}, {5}}, {{11}, {5}}, {{27}, {5}}, {{7}, {5}}, {{23}, {5}}};
const uch ZLIB_INTERNAL _dist_code[DIST_CODE_LEN]
= {0, 1, 2, 3, 4, 4, 5, 5, 6, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8, 8, 8, 8, 8, 8, 9, 9, 9, 9, 9, 9, 9, 9,
10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
0, 0, 16, 17, 18, 18, 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 22, 22, 22, 22, 22, 22, 22, 22, 23, 23, 23, 23, 23, 23, 23, 23,
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,
28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,
29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29};
const uch ZLIB_INTERNAL _length_code[MAX_MATCH - MIN_MATCH + 1]
= {0, 1, 2, 3, 4, 5, 6, 7, 8, 8, 9, 9, 10, 10, 11, 11, 12, 12, 12, 12, 13, 13, 13, 13, 14, 14, 14, 14, 15, 15, 15, 15,
16, 16, 16, 16, 16, 16, 16, 16, 17, 17, 17, 17, 17, 17, 17, 17, 18, 18, 18, 18, 18, 18, 18, 18, 19, 19, 19, 19, 19, 19, 19, 19,
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 28};
local const int base_length[LENGTH_CODES]
= {0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 20, 24, 28, 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 0};
local const int base_dist[D_CODES] = {0, 1, 2, 3, 4, 6, 8, 12, 16, 24, 32, 48, 64, 96, 128,
192, 256, 384, 512, 768, 1024, 1536, 2048, 3072, 4096, 6144, 8192, 12288, 16384, 24576};

311
base/poco/Foundation/src/utils.h
View File

@ -1,311 +0,0 @@
// Copyright 2010 the V8 project authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef DOUBLE_CONVERSION_UTILS_H_
#define DOUBLE_CONVERSION_UTILS_H_
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#ifndef ASSERT
# define ASSERT(condition) assert(condition);
#endif
#ifndef UNIMPLEMENTED
# define UNIMPLEMENTED() (abort())
#endif
#ifndef UNREACHABLE
# define UNREACHABLE() (abort())
#endif
// Double operations detection based on target architecture.
// Linux uses a 80bit wide floating point stack on x86. This induces double
// rounding, which in turn leads to wrong results.
// An easy way to test if the floating-point operations are correct is to
// evaluate: 89255.0/1e22. If the floating-point stack is 64 bits wide then
// the result is equal to 89255e-22.
// The best way to test this, is to create a division-function and to compare
// the output of the division with the expected result. (Inlining must be
// disabled.)
// On Linux,x86 89255e-22 != Div_double(89255.0/1e22)
#if defined(_M_X64) || defined(__x86_64__) || defined(__ARMEL__) || defined(_M_ARM) || defined(__arm__) || defined(__arm64__) \
|| defined(__avr32__) || defined(__hppa__) || defined(__ia64__) || defined(__mips__) || defined(__powerpc__) || defined(__ppc__) \
|| defined(__ppc64__) || defined(__sparc__) || defined(__sparc) || defined(__s390__) || defined(__SH4__) || defined(__alpha__) \
|| defined(_MIPS_ARCH_MIPS32R2) || defined(__riscv) || defined(__AARCH64EL__) || defined(nios2) || defined(__nios2) \
|| defined(__nios2__)
# define DOUBLE_CONVERSION_CORRECT_DOUBLE_OPERATIONS 1
#elif defined(_M_IX86) || defined(__i386__) || defined(__i386)
# undef DOUBLE_CONVERSION_CORRECT_DOUBLE_OPERATIONS
#else
# error Target architecture was not detected as supported by Double-Conversion.
#endif
#if defined(__GNUC__)
# define DOUBLE_CONVERSION_UNUSED __attribute__((unused))
#else
# define DOUBLE_CONVERSION_UNUSED
#endif
# include <stdint.h>
// The following macro works on both 32 and 64-bit platforms.
// Usage: instead of writing 0x1234567890123456
// write UINT64_2PART_C(0x12345678,90123456);
#define UINT64_2PART_C(a, b) (((static_cast<uint64_t>(a) << 32) + 0x##b##u))
// The expression ARRAY_SIZE(a) is a compile-time constant of type
// size_t which represents the number of elements of the given
// array. You should only use ARRAY_SIZE on statically allocated
// arrays.
#ifndef ARRAY_SIZE
# define ARRAY_SIZE(a) ((sizeof(a) / sizeof(*(a))) / static_cast<size_t>(!(sizeof(a) % sizeof(*(a)))))
#endif
// A macro to disallow the evil copy constructor and operator= functions
// This should be used in the private: declarations for a class
#ifndef DISALLOW_COPY_AND_ASSIGN
# define DISALLOW_COPY_AND_ASSIGN(TypeName) \
TypeName(const TypeName &); \
void operator=(const TypeName &)
#endif
// A macro to disallow all the implicit constructors, namely the
// default constructor, copy constructor and operator= functions.
//
// This should be used in the private: declarations for a class
// that wants to prevent anyone from instantiating it. This is
// especially useful for classes containing only static methods.
#ifndef DISALLOW_IMPLICIT_CONSTRUCTORS
# define DISALLOW_IMPLICIT_CONSTRUCTORS(TypeName) \
TypeName(); \
DISALLOW_COPY_AND_ASSIGN(TypeName)
#endif
namespace double_conversion
{
static const int kCharSize = sizeof(char);
// Returns the maximum of the two parameters.
template <typename T>
static T Max(T a, T b)
{
return a < b ? b : a;
}
// Returns the minimum of the two parameters.
template <typename T>
static T Min(T a, T b)
{
return a < b ? a : b;
}
inline int StrLength(const char * string)
{
size_t length = strlen(string);
ASSERT(length == static_cast<size_t>(static_cast<int>(length)));
return static_cast<int>(length);
}
// This is a simplified version of V8's Vector class.
template <typename T>
class Vector
{
public:
Vector() : start_(NULL), length_(0) { }
Vector(T * data, int length) : start_(data), length_(length) { ASSERT(length == 0 || (length > 0 && data != NULL)); }
// Returns a vector using the same backing storage as this one,
// spanning from and including 'from', to but not including 'to'.
Vector<T> SubVector(int from, int to)
{
ASSERT(to <= length_);
ASSERT(from < to);
ASSERT(0 <= from);
return Vector<T>(start() + from, to - from);
}
// Returns the length of the vector.
int length() const { return length_; }
// Returns whether or not the vector is empty.
bool is_empty() const { return length_ == 0; }
// Returns the pointer to the start of the data in the vector.
T * start() const { return start_; }
// Access individual vector elements - checks bounds in debug mode.
T & operator[](int index) const
{
ASSERT(0 <= index && index < length_);
return start_[index];
}
T & first() { return start_[0]; }
T & last() { return start_[length_ - 1]; }
private:
T * start_;
int length_;
};
// Helper class for building result strings in a character buffer. The
// purpose of the class is to use safe operations that checks the
// buffer bounds on all operations in debug mode.
class StringBuilder
{
public:
StringBuilder(char * buffer, int size) : buffer_(buffer, size), position_(0) { }
~StringBuilder()
{
if (!is_finalized())
Finalize();
}
int size() const { return buffer_.length(); }
// Get the current position in the builder.
int position() const
{
ASSERT(!is_finalized());
return position_;
}
// Reset the position.
void Reset() { position_ = 0; }
// Add a single character to the builder. It is not allowed to add
// 0-characters; use the Finalize() method to terminate the string
// instead.
void AddCharacter(char c)
{
ASSERT(c != '\0');
ASSERT(!is_finalized() && position_ < buffer_.length());
buffer_[position_++] = c;
}
// Add an entire string to the builder. Uses strlen() internally to
// compute the length of the input string.
void AddString(const char * s) { AddSubstring(s, StrLength(s)); }
// Add the first 'n' characters of the given string 's' to the
// builder. The input string must have enough characters.
void AddSubstring(const char * s, int n)
{
ASSERT(!is_finalized() && position_ + n < buffer_.length());
ASSERT(static_cast<size_t>(n) <= strlen(s));
memmove(&buffer_[position_], s, n * kCharSize);
position_ += n;
}
// Add character padding to the builder. If count is non-positive,
// nothing is added to the builder.
void AddPadding(char c, int count)
{
for (int i = 0; i < count; i++)
{
AddCharacter(c);
}
}
// Finalize the string by 0-terminating it and returning the buffer.
char * Finalize()
{
ASSERT(!is_finalized() && position_ < buffer_.length());
buffer_[position_] = '\0';
// Make sure nobody managed to add a 0-character to the
// buffer while building the string.
ASSERT(strlen(buffer_.start()) == static_cast<size_t>(position_));
position_ = -1;
ASSERT(is_finalized());
return buffer_.start();
}
private:
Vector<char> buffer_;
int position_;
bool is_finalized() const { return position_ < 0; }
DISALLOW_IMPLICIT_CONSTRUCTORS(StringBuilder);
};
// The type-based aliasing rule allows the compiler to assume that pointers of
// different types (for some definition of different) never alias each other.
// Thus the following code does not work:
//
// float f = foo();
// int fbits = *(int*)(&f);
//
// The compiler 'knows' that the int pointer can't refer to f since the types
// don't match, so the compiler may cache f in a register, leaving random data
// in fbits. Using C++ style casts makes no difference, however a pointer to
// char data is assumed to alias any other pointer. This is the 'memcpy
// exception'.
//
// Bit_cast uses the memcpy exception to move the bits from a variable of one
// type of a variable of another type. Of course the end result is likely to
// be implementation dependent. Most compilers (gcc-4.2 and MSVC 2005)
// will completely optimize BitCast away.
//
// There is an additional use for BitCast.
// Recent gccs will warn when they see casts that may result in breakage due to
// the type-based aliasing rule. If you have checked that there is no breakage
// you can use BitCast to cast one pointer type to another. This confuses gcc
// enough that it can no longer see that you have cast one pointer type to
// another thus avoiding the warning.
template <class Dest, class Source>
inline Dest BitCast(const Source & source)
{
// Compile time assertion: sizeof(Dest) == sizeof(Source)
// A compile error here means your Dest and Source have different sizes.
DOUBLE_CONVERSION_UNUSED
typedef char VerifySizesAreEqual[sizeof(Dest) == sizeof(Source) ? 1 : -1];
Dest dest;
memmove(&dest, &source, sizeof(dest));
return dest;
}
template <class Dest, class Source>
inline Dest BitCast(Source * source)
{
return BitCast<Dest>(reinterpret_cast<uintptr_t>(source));
}
} // namespace double_conversion
#endif // DOUBLE_CONVERSION_UTILS_H_

324
base/poco/Foundation/src/zutil.c
View File

@ -1,324 +0,0 @@
/* zutil.c -- target dependent utility functions for the compression library
* Copyright (C) 1995-2005, 2010, 2011, 2012 Jean-loup Gailly.
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* @(#) $Id: //poco/1.4/Foundation/src/zutil.c#3 $ */
#include "zutil.h"
#ifndef Z_SOLO
# include "gzguts.h"
#endif
#ifndef NO_DUMMY_DECL
struct internal_state {int dummy;}; /* for buggy compilers */
#endif
z_const char * const z_errmsg[10] = {
"need dictionary", /* Z_NEED_DICT 2 */
"stream end", /* Z_STREAM_END 1 */
"", /* Z_OK 0 */
"file error", /* Z_ERRNO (-1) */
"stream error", /* Z_STREAM_ERROR (-2) */
"data error", /* Z_DATA_ERROR (-3) */
"insufficient memory", /* Z_MEM_ERROR (-4) */
"buffer error", /* Z_BUF_ERROR (-5) */
"incompatible version",/* Z_VERSION_ERROR (-6) */
""};
const char * ZEXPORT zlibVersion()
{
return ZLIB_VERSION;
}
uLong ZEXPORT zlibCompileFlags()
{
uLong flags;
flags = 0;
switch ((int)(sizeof(uInt))) {
case 2: break;
case 4: flags += 1; break;
case 8: flags += 2; break;
default: flags += 3;
}
switch ((int)(sizeof(uLong))) {
case 2: break;
case 4: flags += 1 << 2; break;
case 8: flags += 2 << 2; break;
default: flags += 3 << 2;
}
switch ((int)(sizeof(voidpf))) {
case 2: break;
case 4: flags += 1 << 4; break;
case 8: flags += 2 << 4; break;
default: flags += 3 << 4;
}
switch ((int)(sizeof(z_off_t))) {
case 2: break;
case 4: flags += 1 << 6; break;
case 8: flags += 2 << 6; break;
default: flags += 3 << 6;
}
#ifdef ZLIB_DEBUG
flags += 1 << 8;
#endif
#if defined(ASMV) || defined(ASMINF)
flags += 1 << 9;
#endif
#ifdef ZLIB_WINAPI
flags += 1 << 10;
#endif
#ifdef BUILDFIXED
flags += 1 << 12;
#endif
#ifdef DYNAMIC_CRC_TABLE
flags += 1 << 13;
#endif
#ifdef NO_GZCOMPRESS
flags += 1L << 16;
#endif
#ifdef NO_GZIP
flags += 1L << 17;
#endif
#ifdef PKZIP_BUG_WORKAROUND
flags += 1L << 20;
#endif
#ifdef FASTEST
flags += 1L << 21;
#endif
#if defined(STDC) || defined(Z_HAVE_STDARG_H)
# ifdef NO_vsnprintf
flags += 1L << 25;
# ifdef HAS_vsprintf_void
flags += 1L << 26;
# endif
# else
# ifdef HAS_vsnprintf_void
flags += 1L << 26;
# endif
# endif
#else
flags += 1L << 24;
# ifdef NO_snprintf
flags += 1L << 25;
# ifdef HAS_sprintf_void
flags += 1L << 26;
# endif
# else
# ifdef HAS_snprintf_void
flags += 1L << 26;
# endif
# endif
#endif
return flags;
}
#ifdef ZLIB_DEBUG
# ifndef verbose
# define verbose 0
# endif
int ZLIB_INTERNAL z_verbose = verbose;
void ZLIB_INTERNAL z_error (m)
char *m;
{
fprintf(stderr, "%s\n", m);
exit(1);
}
#endif
/* exported to allow conversion of error code to string for compress() and
* uncompress()
*/
const char * ZEXPORT zError(err)
int err;
{
return ERR_MSG(err);
}
#if defined(_WIN32_WCE) && _WIN32_WCE < 0x800
/* The Microsoft C Run-Time Library for Windows CE doesn't have
* errno. We define it as a global variable to simplify porting.
* Its value is always 0 and should not be used.
*/
int errno = 0;
#endif
#ifndef HAVE_MEMCPY
void ZLIB_INTERNAL zmemcpy(dest, source, len)
Bytef* dest;
const Bytef* source;
uInt len;
{
if (len == 0) return;
do {
*dest++ = *source++; /* ??? to be unrolled */
} while (--len != 0);
}
int ZLIB_INTERNAL zmemcmp(s1, s2, len)
const Bytef* s1;
const Bytef* s2;
uInt len;
{
uInt j;
for (j = 0; j < len; j++) {
if (s1[j] != s2[j]) return 2*(s1[j] > s2[j])-1;
}
return 0;
}
void ZLIB_INTERNAL zmemzero(dest, len)
Bytef* dest;
uInt len;
{
if (len == 0) return;
do {
*dest++ = 0; /* ??? to be unrolled */
} while (--len != 0);
}
#endif
#ifndef Z_SOLO
#ifdef SYS16BIT
#ifdef __TURBOC__
/* Turbo C in 16-bit mode */
# define MY_ZCALLOC
/* Turbo C malloc() does not allow dynamic allocation of 64K bytes
* and farmalloc(64K) returns a pointer with an offset of 8, so we
* must fix the pointer. Warning: the pointer must be put back to its
* original form in order to free it, use zcfree().
*/
#define MAX_PTR 10
/* 10*64K = 640K */
local int next_ptr = 0;
typedef struct ptr_table_s {
voidpf org_ptr;
voidpf new_ptr;
} ptr_table;
local ptr_table table[MAX_PTR];
/* This table is used to remember the original form of pointers
* to large buffers (64K). Such pointers are normalized with a zero offset.
* Since MS-DOS is not a preemptive multitasking OS, this table is not
* protected from concurrent access. This hack doesn't work anyway on
* a protected system like OS/2. Use Microsoft C instead.
*/
voidpf ZLIB_INTERNAL zcalloc (voidpf opaque, unsigned items, unsigned size)
{
voidpf buf = opaque; /* just to make some compilers happy */
ulg bsize = (ulg)items*size;
/* If we allocate less than 65520 bytes, we assume that farmalloc
* will return a usable pointer which doesn't have to be normalized.
*/
if (bsize < 65520L) {
buf = farmalloc(bsize);
if (*(ush*)&buf != 0) return buf;
} else {
buf = farmalloc(bsize + 16L);
}
if (buf == NULL || next_ptr >= MAX_PTR) return NULL;
table[next_ptr].org_ptr = buf;
/* Normalize the pointer to seg:0 */
*((ush*)&buf+1) += ((ush)((uch*)buf-0) + 15) >> 4;
*(ush*)&buf = 0;
table[next_ptr++].new_ptr = buf;
return buf;
}
void ZLIB_INTERNAL zcfree (voidpf opaque, voidpf ptr)
{
int n;
if (*(ush*)&ptr != 0) { /* object < 64K */
farfree(ptr);
return;
}
/* Find the original pointer */
for (n = 0; n < next_ptr; n++) {
if (ptr != table[n].new_ptr) continue;
farfree(table[n].org_ptr);
while (++n < next_ptr) {
table[n-1] = table[n];
}
next_ptr--;
return;
}
ptr = opaque; /* just to make some compilers happy */
Assert(0, "zcfree: ptr not found");
}
#endif /* __TURBOC__ */
#ifdef M_I86
/* Microsoft C in 16-bit mode */
# define MY_ZCALLOC
#if (!defined(_MSC_VER) || (_MSC_VER <= 600))
# define _halloc halloc
# define _hfree hfree
#endif
voidpf ZLIB_INTERNAL zcalloc (voidpf opaque, uInt items, uInt size)
{
if (opaque) opaque = 0; /* to make compiler happy */
return _halloc((long)items, size);
}
void ZLIB_INTERNAL zcfree (voidpf opaque, voidpf ptr)
{
if (opaque) opaque = 0; /* to make compiler happy */
_hfree(ptr);
}
#endif /* M_I86 */
#endif /* SYS16BIT */
#ifndef MY_ZCALLOC /* Any system without a special alloc function */
#ifndef STDC
extern voidp malloc OF((uInt size));
extern voidp calloc OF((uInt items, uInt size));
extern void free OF((voidpf ptr));
#endif
voidpf ZLIB_INTERNAL zcalloc (opaque, items, size)
voidpf opaque;
unsigned items;
unsigned size;
{
if (opaque) items += size - size; /* make compiler happy */
return sizeof(uInt) > 2 ? (voidpf)malloc(items * size) :
(voidpf)calloc(items, size);
}
void ZLIB_INTERNAL zcfree (opaque, ptr)
voidpf opaque;
voidpf ptr;
{
free(ptr);
if (opaque) return; /* make compiler happy */
}
#endif /* MY_ZCALLOC */
#endif /* !Z_SOLO */

237
base/poco/Foundation/src/zutil.h
View File

@ -1,237 +0,0 @@
/* zutil.h -- internal interface and configuration of the compression library
* Copyright (C) 1995-2013 Jean-loup Gailly.
* For conditions of distribution and use, see copyright notice in zlib.h
*/
/* WARNING: this file should *not* be used by applications. It is
part of the implementation of the compression library and is
subject to change. Applications should only use zlib.h.
*/
/* @(#) $Id: //poco/1.4/Foundation/src/zutil.h#3 $ */
#ifndef ZUTIL_H
#define ZUTIL_H
#ifdef HAVE_HIDDEN
# define ZLIB_INTERNAL __attribute__((visibility("hidden")))
#else
# define ZLIB_INTERNAL
#endif
#include "zlib.h"
#if defined(STDC) && !defined(Z_SOLO)
# include <stddef.h>
# include <stdlib.h>
# include <string.h>
#endif
#ifdef Z_SOLO
typedef long ptrdiff_t; /* guess -- will be caught if guess is wrong */
#endif
#ifndef local
# define local static
#endif
/* compile with -Dlocal if your debugger can't find static symbols */
typedef unsigned char uch;
typedef uch FAR uchf;
typedef unsigned short ush;
typedef ush FAR ushf;
typedef unsigned long ulg;
extern z_const char * const z_errmsg[10]; /* indexed by 2-zlib_error */
/* (size given to avoid silly warnings with Visual C++) */
#define ERR_MSG(err) z_errmsg[Z_NEED_DICT - (err)]
#define ERR_RETURN(strm, err) return (strm->msg = ERR_MSG(err), (err))
/* To be used only when the state is known to be valid */
/* common constants */
#ifndef DEF_WBITS
# define DEF_WBITS MAX_WBITS
#endif
/* default windowBits for decompression. MAX_WBITS is for compression only */
#if MAX_MEM_LEVEL >= 8
# define DEF_MEM_LEVEL 8
#else
# define DEF_MEM_LEVEL MAX_MEM_LEVEL
#endif
/* default memLevel */
#define STORED_BLOCK 0
#define STATIC_TREES 1
#define DYN_TREES 2
/* The three kinds of block type */
#define MIN_MATCH 3
#define MAX_MATCH 258
/* The minimum and maximum match lengths */
#define PRESET_DICT 0x20 /* preset dictionary flag in zlib header */
/* target dependencies */
#ifdef AMIGA
# define OS_CODE 0x01
#endif
#if defined(VAXC) || defined(VMS)
# define OS_CODE 0x02
# define F_OPEN(name, mode) fopen((name), (mode), "mbc=60", "ctx=stm", "rfm=fix", "mrs=512")
#endif
#if defined(ATARI) || defined(atarist)
# define OS_CODE 0x05
#endif
#ifdef OS2
# define OS_CODE 0x06
# if defined(M_I86) && !defined(Z_SOLO)
# include <malloc.h>
# endif
#endif
#if defined(MACOS) || defined(TARGET_OS_MAC)
# define OS_CODE 0x07
# ifndef Z_SOLO
# if defined(__MWERKS__) && __dest_os != __be_os && __dest_os != __win32_os
# include <unix.h> /* for fdopen */
# else
# ifndef fdopen
# define fdopen(fd, mode) NULL /* No fdopen() */
# endif
# endif
# endif
#endif
#ifdef TOPS20
# define OS_CODE 0x0a
#endif
#ifdef WIN32
# define OS_CODE 0x0b
#endif
#ifdef __50SERIES /* Prime/PRIMOS */
# define OS_CODE 0x0f
#endif
#if defined(_BEOS_) || defined(RISCOS)
# define fdopen(fd, mode) NULL /* No fdopen() */
#endif
/* provide prototypes for these when building zlib without LFS */
#if !defined(_WIN32) && (!defined(_LARGEFILE64_SOURCE) || _LFS64_LARGEFILE - 0 == 0)
ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off_t));
ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off_t));
#endif
/* common defaults */
#ifndef OS_CODE
# define OS_CODE 0x03 /* assume Unix */
#endif
#ifndef F_OPEN
# define F_OPEN(name, mode) fopen((name), (mode))
#endif
/* functions */
#if defined(pyr) || defined(Z_SOLO)
# define NO_MEMCPY
#endif
#if defined(SMALL_MEDIUM) && !defined(_MSC_VER) && !defined(__SC__)
/* Use our own functions for small and medium model with MSC <= 5.0.
* You may have to use the same strategy for Borland C (untested).
* The __SC__ check is for Symantec.
*/
# define NO_MEMCPY
#endif
#if defined(STDC) && !defined(HAVE_MEMCPY) && !defined(NO_MEMCPY)
# define HAVE_MEMCPY
#endif
#ifdef HAVE_MEMCPY
# ifdef SMALL_MEDIUM /* MS-DOS small or medium model */
# define zmemcpy _fmemcpy
# define zmemcmp _fmemcmp
# define zmemzero(dest, len) _fmemset(dest, 0, len)
# else
# define zmemcpy memcpy
# define zmemcmp memcmp
# define zmemzero(dest, len) memset(dest, 0, len)
# endif
#else
void ZLIB_INTERNAL zmemcpy OF((Bytef * dest, const Bytef * source, uInt len));
int ZLIB_INTERNAL zmemcmp OF((const Bytef * s1, const Bytef * s2, uInt len));
void ZLIB_INTERNAL zmemzero OF((Bytef * dest, uInt len));
#endif
/* Diagnostic functions */
#ifdef ZLIB_DEBUG
# include <stdio.h>
extern int ZLIB_INTERNAL z_verbose;
extern void ZLIB_INTERNAL z_error OF((char * m));
# define Assert(cond, msg) \
{ \
if (!(cond)) \
z_error(msg); \
}
# define Trace(x) \
{ \
if (z_verbose >= 0) \
fprintf x; \
}
# define Tracev(x) \
{ \
if (z_verbose > 0) \
fprintf x; \
}
# define Tracevv(x) \
{ \
if (z_verbose > 1) \
fprintf x; \
}
# define Tracec(c, x) \
{ \
if (z_verbose > 0 && (c)) \
fprintf x; \
}
# define Tracecv(c, x) \
{ \
if (z_verbose > 1 && (c)) \
fprintf x; \
}
#else
# define Assert(cond, msg)
# define Trace(x)
# define Tracev(x)
# define Tracevv(x)
# define Tracec(c, x)
# define Tracecv(c, x)
#endif
#ifndef Z_SOLO
voidpf ZLIB_INTERNAL zcalloc OF((voidpf opaque, unsigned items, unsigned size));
void ZLIB_INTERNAL zcfree OF((voidpf opaque, voidpf ptr));
#endif
#define ZALLOC(strm, items, size) (*((strm)->zalloc))((strm)->opaque, (items), (size))
#define ZFREE(strm, addr) (*((strm)->zfree))((strm)->opaque, (voidpf)(addr))
#define TRY_FREE(s, p) \
{ \
if (p) \
ZFREE(s, p); \
}
/* Reverse the bytes in a 32-bit value */
#define ZSWAP32(q) ((((q) >> 24) & 0xff) + (((q) >> 8) & 0xff00) + (((q)&0xff00) << 8) + (((q)&0xff) << 24))
#endif /* ZUTIL_H */

2
contrib/llvm-project vendored

@ -1 +1 @@
Subproject commit e0accd517933ebb44aff84bc8db448ffd8ef1929
Subproject commit 2aedf7598a4040b23881dbe05b6afaca25a337ef

2
contrib/zstd vendored

@ -1 +1 @@
Subproject commit 945f27758c0fd67b636103a38dbf050266c6b90a
Subproject commit 63779c798237346c2b245c546c40b72a5a5913fe

4
docker/images.json
View File

@ -151,5 +151,9 @@
"name": "clickhouse/docs-builder",
"dependent": [
]
},
"docker/test/sqllogic": {
"name": "clickhouse/sqllogic-test",
"dependent": []
}
}

9
docker/test/fasttest/run.sh
View File

@ -194,7 +194,12 @@ function build
{
(
cd "$FASTTEST_BUILD"
time ninja clickhouse-bundle 2>&1 | ts '%Y-%m-%d %H:%M:%S' | tee "$FASTTEST_OUTPUT/build_log.txt"
TIMEFORMAT=$'\nreal\t%3R\nuser\t%3U\nsys\t%3S'
( time ninja clickhouse-bundle) |& ts '%Y-%m-%d %H:%M:%S' | tee "$FASTTEST_OUTPUT/build_log.txt"
BUILD_SECONDS_ELAPSED=$(awk '/^....-..-.. ..:..:.. real\t[0-9]/ {print $4}' < "$FASTTEST_OUTPUT/build_log.txt")
echo "build_clickhouse_fasttest_binary: [ OK ] $BUILD_SECONDS_ELAPSED sec." \
| ts '%Y-%m-%d %H:%M:%S' \
| tee "$FASTTEST_OUTPUT/test_result.txt"
if [ "$COPY_CLICKHOUSE_BINARY_TO_OUTPUT" -eq "1" ]; then
cp programs/clickhouse "$FASTTEST_OUTPUT/clickhouse"
@ -251,7 +256,7 @@ function run_tests
)
time clickhouse-test "${test_opts[@]}" -- "$FASTTEST_FOCUS" 2>&1 \
| ts '%Y-%m-%d %H:%M:%S' \
| tee "$FASTTEST_OUTPUT/test_result.txt"
| tee -a "$FASTTEST_OUTPUT/test_result.txt"
set -e
clickhouse stop --pid-path "$FASTTEST_DATA"

14
docker/test/integration/runner/Dockerfile
View File

@ -47,10 +47,9 @@ ENV TZ=Etc/UTC
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
ENV DOCKER_CHANNEL stable
RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add -
RUN add-apt-repository "deb https://download.docker.com/linux/ubuntu $(lsb_release -c -s) ${DOCKER_CHANNEL}"
RUN apt-get update \
RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add - \
&& add-apt-repository "deb https://download.docker.com/linux/ubuntu $(lsb_release -c -s) ${DOCKER_CHANNEL}" \
&& apt-get update \
&& env DEBIAN_FRONTEND=noninteractive apt-get install --yes \
docker-ce \
&& rm -rf \
@ -61,7 +60,7 @@ RUN apt-get update \
RUN dockerd --version; docker --version
RUN python3 -m pip install \
RUN python3 -m pip install --no-cache-dir \
PyMySQL \
aerospike==4.0.0 \
avro==1.10.2 \
@ -104,8 +103,9 @@ COPY dockerd-entrypoint.sh /usr/local/bin/
COPY compose/ /compose/
COPY misc/ /misc/
RUN wget https://dlcdn.apache.org/spark/spark-3.3.2/spark-3.3.2-bin-hadoop3.tgz \
&& tar xzvf spark-3.3.2-bin-hadoop3.tgz -C /
RUN curl -fsSL -O https://dlcdn.apache.org/spark/spark-3.3.2/spark-3.3.2-bin-hadoop3.tgz \
&& tar xzvf spark-3.3.2-bin-hadoop3.tgz -C / \
&& rm spark-3.3.2-bin-hadoop3.tgz
# download spark and packages
# if you change packages, don't forget to update them in tests/integration/helpers/cluster.py

4
docker/test/integration/runner/dockerd-entrypoint.sh
View File

@ -16,7 +16,9 @@ echo '{
# and on hung you can simply press Ctrl-C and it will spawn a python pdb,
# but on SIGINT dockerd will exit, so ignore it to preserve the daemon.
trap '' INT
dockerd --host=unix:///var/run/docker.sock --host=tcp://0.0.0.0:2375 --default-address-pool base=172.17.0.0/12,size=24 &>/ClickHouse/tests/integration/dockerd.log &
# Binding to an IP address without --tlsverify is deprecated. Startup is intentionally being slowed
# unless --tls=false or --tlsverify=false is set
dockerd --host=unix:///var/run/docker.sock --tls=false --host=tcp://0.0.0.0:2375 --default-address-pool base=172.17.0.0/12,size=24 &>/ClickHouse/tests/integration/dockerd.log &
set +e
reties=0

45
docker/test/sqllogic/Dockerfile Normal file
View File

@ -0,0 +1,45 @@
# docker build -t clickhouse/sqllogic-test .
ARG FROM_TAG=latest
FROM clickhouse/test-base:$FROM_TAG
RUN apt-get update --yes \
&& env DEBIAN_FRONTEND=noninteractive \
apt-get install --yes --no-install-recommends \
wget \
git \
python3 \
python3-dev \
python3-pip \
sqlite3 \
unixodbc \
unixodbc-dev \
sudo \
&& apt-get clean
RUN pip3 install \
numpy \
pyodbc \
deepdiff
ARG odbc_repo="https://github.com/ClickHouse/clickhouse-odbc.git"
RUN git clone --recursive ${odbc_repo} \
&& mkdir -p /clickhouse-odbc/build \
&& cmake -S /clickhouse-odbc -B /clickhouse-odbc/build \
&& ls /clickhouse-odbc/build/driver \
&& make -j 10 -C /clickhouse-odbc/build \
&& ls /clickhouse-odbc/build/driver \
&& mkdir -p /usr/local/lib64/ && cp /clickhouse-odbc/build/driver/lib*.so /usr/local/lib64/ \
&& odbcinst -i -d -f /clickhouse-odbc/packaging/odbcinst.ini.sample \
&& odbcinst -i -s -l -f /clickhouse-odbc/packaging/odbc.ini.sample
ENV TZ=Europe/Amsterdam
ENV MAX_RUN_TIME=900
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
ARG sqllogic_test_repo="https://github.com/gregrahn/sqllogictest.git"
RUN git clone --recursive ${sqllogic_test_repo}
COPY run.sh /
CMD ["/bin/bash", "/run.sh"]

100
docker/test/sqllogic/run.sh Executable file
View File

@ -0,0 +1,100 @@
#!/bin/bash
set -exu
trap "exit" INT TERM
echo "ENV"
env
# fail on errors, verbose and export all env variables
set -e -x -a
echo "Current directory"
pwd
echo "Files in current directory"
ls -la ./
echo "Files in root directory"
ls -la /
echo "Files in /clickhouse-tests directory"
ls -la /clickhouse-tests
echo "Files in /clickhouse-tests/sqllogic directory"
ls -la /clickhouse-tests/sqllogic
echo "Files in /package_folder directory"
ls -la /package_folder
echo "Files in /test_output"
ls -la /test_output
echo "File in /sqllogictest"
ls -la /sqllogictest
dpkg -i package_folder/clickhouse-common-static_*.deb
dpkg -i package_folder/clickhouse-common-static-dbg_*.deb
dpkg -i package_folder/clickhouse-server_*.deb
dpkg -i package_folder/clickhouse-client_*.deb
# install test configs
# /clickhouse-tests/config/install.sh
sudo clickhouse start
sleep 5
for _ in $(seq 1 60); do if [[ $(wget --timeout=1 -q 'localhost:8123' -O-) == 'Ok.' ]]; then break ; else sleep 1; fi ; done
function run_tests()
{
set -x
cd /test_output
/clickhouse-tests/sqllogic/runner.py --help 2>&1 \
| ts '%Y-%m-%d %H:%M:%S'
mkdir -p /test_output/self-test
/clickhouse-tests/sqllogic/runner.py --log-file /test_output/runner-self-test.log \
self-test \
--self-test-dir /clickhouse-tests/sqllogic/self-test \
--out-dir /test_output/self-test \
2>&1 \
| ts '%Y-%m-%d %H:%M:%S'
cat /test_output/self-test/check_status.tsv >> /test_output/check_status.tsv
cat /test_output/self-test/test_results.tsv >> /test_output/test_results.tsv ||:
tar -zcvf self-test.tar.gz self-test 1>/dev/null
if [ -d /sqllogictest ]
then
mkdir -p /test_output/statements-test
/clickhouse-tests/sqllogic/runner.py \
--log-file /test_output/runner-statements-test.log \
--log-level info \
statements-test \
--input-dir /sqllogictest \
--out-dir /test_output/statements-test \
2>&1 \
| ts '%Y-%m-%d %H:%M:%S'
cat /test_output/statements-test/check_status.tsv >> /test_output/check_status.tsv
cat /test_output/statements-test/test_results.tsv >> /test_output/test_results.tsv
tar -zcvf statements-check.tar.gz statements-test 1>/dev/null
fi
}
export -f run_tests
timeout "${MAX_RUN_TIME:-9000}" bash -c run_tests || echo "timeout reached" >&2
#/process_functional_tests_result.py || echo -e "failure\tCannot parse results" > /test_output/check_status.tsv
clickhouse-client -q "system flush logs" ||:
# Stop server so we can safely read data with clickhouse-local.
# Why do we read data with clickhouse-local?
# Because it's the simplest way to read it when server has crashed.
sudo clickhouse stop ||:
for _ in $(seq 1 60); do if [[ $(wget --timeout=1 -q 'localhost:8123' -O-) == 'Ok.' ]]; then sleep 1 ; else break; fi ; done
grep -Fa "Fatal" /var/log/clickhouse-server/clickhouse-server.log ||:
pigz < /var/log/clickhouse-server/clickhouse-server.log > /test_output/clickhouse-server.log.gz &
# Compressed (FIXME: remove once only github actions will be left)
rm /var/log/clickhouse-server/clickhouse-server.log
mv /var/log/clickhouse-server/stderr.log /test_output/ ||:

2
docs/en/engines/table-engines/integrations/materialized-postgresql.md
View File

@ -33,6 +33,8 @@ PRIMARY KEY key;
3. Only database [Atomic](https://en.wikipedia.org/wiki/Atomicity_(database_systems)) is allowed.
4. The `MaterializedPostgreSQL` table engine only works for PostgreSQL versions >= 11 as the implementation requires the [pg_replication_slot_advance](https://pgpedia.info/p/pg_replication_slot_advance.html) PostgreSQL function.
## Virtual columns {#virtual-columns}
- `_version` — Transaction counter. Type: [UInt64](../../../sql-reference/data-types/int-uint.md).

37
docs/en/engines/table-engines/integrations/postgresql.md
View File

@ -111,7 +111,7 @@ In the example below replica `example01-1` has the highest priority:
## Usage Example {#usage-example}
Table in PostgreSQL:
### Table in PostgreSQL
``` text
postgres=# CREATE TABLE "public"."test" (
@ -134,7 +134,9 @@ postgresql> SELECT * FROM test;
(1 row)
```
Table in ClickHouse, retrieving data from the PostgreSQL table created above:
### Creating Table in ClickHouse, and connecting to PostgreSQL table created above
This example uses the [PostgreSQL table engine](/docs/en/engines/table-engines/integrations/postgresql.md) to connect the ClickHouse table to the PostgreSQL table:
``` sql
CREATE TABLE default.postgresql_table
@ -146,6 +148,35 @@ CREATE TABLE default.postgresql_table
ENGINE = PostgreSQL('localhost:5432', 'public', 'test', 'postges_user', 'postgres_password');
```
### Inserting initial data from PostgreSQL table into ClickHouse table, using a SELECT query
The [postgresql table function](/docs/en/sql-reference/table-functions/postgresql.md) copies the data from PostgreSQL to ClickHouse, which is often used for improving the query performance of the data by querying or performing analytics in ClickHouse rather than in PostgreSQL, or can also be used for migrating data from PostgreSQL to ClickHouse:
``` sql
INSERT INTO default.postgresql_table
SELECT * FROM postgresql('localhost:5432', 'public', 'test', 'postges_user', 'postgres_password');
```
### Inserting incremental data from PostgreSQL table into ClickHouse table
If then performing ongoing synchronization between the PostgreSQL table and ClickHouse table after the initial insert, you can use a WHERE clause in ClickHouse to insert only data added to PostgreSQL based on a timestamp or unique sequence ID.
This would require keeping track of the max ID or timestamp previously added, such as the following:
``` sql
SELECT max(`int_id`) AS maxIntID FROM default.postgresql_table;
```
Then inserting values from PostgreSQL table greater than the max
``` sql
INSERT INTO default.postgresql_table
SELECT * FROM postgresql('localhost:5432', 'public', 'test', 'postges_user', 'postgres_password');
WHERE int_id > maxIntID;
```
### Selecting data from the resulting ClickHouse table
``` sql
SELECT * FROM postgresql_table WHERE str IN ('test');
```
@ -156,7 +187,7 @@ SELECT * FROM postgresql_table WHERE str IN ('test');
└────────────────┴──────┴────────┘
```
Using Non-default Schema:
### Using Non-default Schema
```text
postgres=# CREATE SCHEMA "nice.schema";

2
docs/en/engines/table-engines/mergetree-family/mergetree.md
View File

@ -483,7 +483,7 @@ Indexes of type `set` can be utilized by all functions. The other index types ar
Functions with a constant argument that is less than ngram size can’t be used by `ngrambf_v1` for query optimization.
(*) For `hasTokenCaseInsensitve` and `hasTokenCaseInsensitive` to be effective, the `tokenbf_v1` index must be created on lowercased data, for example `INDEX idx (lower(str_col)) TYPE tokenbf_v1(512, 3, 0)`.
(*) For `hasTokenCaseInsensitive` and `hasTokenCaseInsensitiveOrNull` to be effective, the `tokenbf_v1` index must be created on lowercased data, for example `INDEX idx (lower(str_col)) TYPE tokenbf_v1(512, 3, 0)`.
:::note
Bloom filters can have false positive matches, so the `ngrambf_v1`, `tokenbf_v1`, and `bloom_filter` indexes can not be used for optimizing queries where the result of a function is expected to be false.

16
docs/en/engines/table-engines/mergetree-family/replacingmergetree.md
View File

@ -20,12 +20,12 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
name1 [type1] [DEFAULT|MATERIALIZED|ALIAS expr1],
name2 [type2] [DEFAULT|MATERIALIZED|ALIAS expr2],
...
) ENGINE = ReplacingMergeTree([ver])
) ENGINE = ReplacingMergeTree([ver [, is_deleted]])
[PARTITION BY expr]
[ORDER BY expr]
[PRIMARY KEY expr]
[SAMPLE BY expr]
[SETTINGS name=value, ...]
[SETTINGS name=value, clean_deleted_rows=value, ...]
```
For a description of request parameters, see [statement description](../../../sql-reference/statements/create/table.md).
@ -88,6 +88,18 @@ SELECT * FROM mySecondReplacingMT FINAL;
└─────┴─────────┴─────────────────────┘
```
### is_deleted
`is_deleted` — Name of the column with the type of row: `1` is a “deleted“ row, `0` is a “state“ row.
Column data type — `Int8`.
Can only be enabled when `ver` is used.
The row is deleted when use the `OPTIMIZE ... FINAL CLEANUP`, or `OPTIMIZE ... FINAL` if the engine settings `clean_deleted_rows` has been set to `Always`.
No matter the operation on the data, the version must be increased. If two inserted rows have the same version number, the last inserted one is the one kept.
## Query clauses
When creating a `ReplacingMergeTree` table the same [clauses](../../../engines/table-engines/mergetree-family/mergetree.md) are required, as when creating a `MergeTree` table.

10
docs/en/getting-started/example-datasets/youtube-dislikes.md
View File

@ -122,17 +122,10 @@ SELECT
super_titles,
ifNull(uploader_badges, '') AS uploader_badges,
ifNull(video_badges, '') AS video_badges
FROM s3Cluster(
'default',
FROM s3(
'https://clickhouse-public-datasets.s3.amazonaws.com/youtube/original/files/*.zst',
'JSONLines'
)
SETTINGS
max_download_threads = 24,
max_insert_threads = 64,
max_insert_block_size = 100000000,
min_insert_block_size_rows = 100000000,
min_insert_block_size_bytes = 500000000;
```
Some comments about our `INSERT` command:
@ -140,7 +133,6 @@ Some comments about our `INSERT` command:
- The `parseDateTimeBestEffortUSOrZero` function is handy when the incoming date fields may not be in the proper format. If `fetch_date` does not get parsed properly, it will be set to `0`
- The `upload_date` column contains valid dates, but it also contains strings like "4 hours ago" - which is certainly not a valid date. We decided to store the original value in `upload_date_str` and attempt to parse it with `toDate(parseDateTimeBestEffortUSOrZero(upload_date::String))`. If the parsing fails we just get `0`
- We used `ifNull` to avoid getting `NULL` values in our table. If an incoming value is `NULL`, the `ifNull` function is setting the value to an empty string
- It takes a long time to download the data, so we added a `SETTINGS` clause to spread out the work over more threads while making sure the block sizes stayed fairly large
4. Open a new tab in the SQL Console of ClickHouse Cloud (or a new `clickhouse-client` window) and watch the count increase. It will take a while to insert 4.56B rows, depending on your server resources. (Without any tweaking of settings, it takes about 4.5 hours.)

35
docs/en/interfaces/formats.md
View File

@ -38,6 +38,7 @@ The supported formats are:
| [JSONCompactStrings](#jsoncompactstrings) | ✗ | ✔ |
| [JSONCompactColumns](#jsoncompactcolumns) | ✔ | ✔ |
| [JSONEachRow](#jsoneachrow) | ✔ | ✔ |
| [PrettyJSONEachRow](#prettyjsoneachrow) | ✗ | ✔ |
| [JSONEachRowWithProgress](#jsoneachrowwithprogress) | ✗ | ✔ |
| [JSONStringsEachRow](#jsonstringseachrow) | ✔ | ✔ |
| [JSONStringsEachRowWithProgress](#jsonstringseachrowwithprogress) | ✗ | ✔ |
@ -917,6 +918,40 @@ Example:
While importing data columns with unknown names will be skipped if setting [input_format_skip_unknown_fields](/docs/en/operations/settings/settings-formats.md/#input_format_skip_unknown_fields) is set to 1.
## PrettyJSONEachRow {#prettyjsoneachrow}
Differs from JSONEachRow only in that JSON is pretty formatted with new line delimiters and 4 space indents. Suitable only for output.
Example
```json
{
"num": "42",
"str": "hello",
"arr": [
"0",
"1"
],
"tuple": {
"num": 42,
"str": "world"
}
}
{
"num": "43",
"str": "hello",
"arr": [
"0",
"1",
"2"
],
"tuple": {
"num": 43,
"str": "world"
}
}
```
## JSONStringsEachRow {#jsonstringseachrow}
Differs from JSONEachRow only in that data fields are output in strings, not in typed JSON values.

29
docs/en/interfaces/http.md
View File

@ -333,6 +333,35 @@ You can create a query with parameters and pass values for them from the corresp
$ curl -sS "<address>?param_id=2&param_phrase=test" -d "SELECT * FROM table WHERE int_column = {id:UInt8} and string_column = {phrase:String}"
```
### Tabs in URL Parameters
Query parameters are parsed from the "escaped" format. This has some benefits, such as the possibility to unambiguously parse nulls as `\N`. This means the tab character should be encoded as `\t` (or `\` and a tab). For example, the following contains an actual tab between `abc` and `123` and the input string is split into two values:
```bash
curl -sS "http://localhost:8123" -d "SELECT splitByChar('\t', 'abc 123')"
```
```response
['abc','123']
```
However, if you try to encode an actual tab using `%09` in a URL parameter, it won't get parsed properly:
```bash
curl -sS "http://localhost:8123?param_arg1=abc%09123" -d "SELECT splitByChar('\t', {arg1:String})"
Code: 457. DB::Exception: Value abc 123 cannot be parsed as String for query parameter 'arg1' because it isn't parsed completely: only 3 of 7 bytes was parsed: abc. (BAD_QUERY_PARAMETER) (version 23.4.1.869 (official build))
```
If you are using URL parameters, you will need to encode the `\t` as `%5C%09`. For example:
```bash
curl -sS "http://localhost:8123?param_arg1=abc%5C%09123" -d "SELECT splitByChar('\t', {arg1:String})"
```
```response
['abc','123']
```
## Predefined HTTP Interface {#predefined_http_interface}
ClickHouse supports specific queries through the HTTP interface. For example, you can write data to a table as follows:

2
docs/en/operations/server-configuration-parameters/settings.md
View File

@ -1397,7 +1397,7 @@ Data for the query cache is allocated in DRAM. If memory is scarce, make sure to
```xml
<query_cache>
<size>1073741824</size>
<max_size>1073741824</max_size>
<max_entries>1024</max_entries>
<max_entry_size>1048576</max_entry_size>
<max_entry_rows>30000000</max_entry_rows>

10
docs/en/operations/settings/merge-tree-settings.md
View File

@ -831,3 +831,13 @@ You can see which parts of `s` were stored using the sparse serialization:
│ s │ Sparse │
└────────┴────────────────────┘
```
## clean_deleted_rows
Enable/disable automatic deletion of rows flagged as `is_deleted` when perform `OPTIMIZE ... FINAL` on a table using the ReplacingMergeTree engine. When disabled, the `CLEANUP` keyword has to be added to the `OPTIMIZE ... FINAL` to have the same behaviour.
Possible values:
- `Always` or `Never`.
Default value: `Never`

2
docs/en/operations/settings/query-complexity.md
View File

@ -26,7 +26,7 @@ It can take one of two values: `throw` or `break`. Restrictions on aggregation (
The maximum amount of RAM to use for running a query on a single server.
In the default configuration file, the maximum is 10 GB.
The default setting is unlimited (set to `0`).
The setting does not consider the volume of available memory or the total volume of memory on the machine.
The restriction applies to a single query within a single server.

15
docs/en/operations/system-tables/mutations.md
View File

@ -5,7 +5,7 @@ slug: /en/operations/system-tables/mutations
The table contains information about [mutations](/docs/en/sql-reference/statements/alter/index.md#mutations) of [MergeTree](/docs/en/engines/table-engines/mergetree-family/mergetree.md) tables and their progress. Each mutation command is represented by a single row.
Columns:
## Columns:
- `database` ([String](/docs/en/sql-reference/data-types/string.md)) — The name of the database to which the mutation was applied.
@ -43,6 +43,19 @@ If there were problems with mutating some data parts, the following columns cont
- `latest_fail_reason` ([String](/docs/en/sql-reference/data-types/string.md)) — The exception message that caused the most recent part mutation failure.
## Monitoring Mutations
To track the progress on the system.mutations table, use a query like the following - this requires read permissions on the system.* tables:
``` sql
SELECT * FROM clusterAllReplicas('cluster_name', 'db', system.mutations)
WHERE is_done=0 AND table='tmp';
```
:::tip
replace `tmp` in `table='tmp'` with the name of the table that you are checking mutations on.
:::
**See Also**
- [Mutations](/docs/en/sql-reference/statements/alter/index.md#mutations)

277
docs/en/sql-reference/functions/arithmetic-functions.md
View File

@ -1,12 +1,14 @@
---
slug: /en/sql-reference/functions/arithmetic-functions
sidebar_position: 34
sidebar_position: 5
sidebar_label: Arithmetic
---
# Arithmetic Functions
For all arithmetic functions, the result type is calculated as the smallest number type that the result fits in, if there is such a type. The minimum is taken simultaneously based on the number of bits, whether it is signed, and whether it floats. If there are not enough bits, the highest bit type is taken.
The result type of all arithmetic functions is the smallest type which can represent all possible results. Size promotion happens for integers up to 32 bit, e.g. `UInt8 + UInt16 = UInt32`. If one of the inters has 64 or more bits, the result is of the same type as the bigger of the input integers, e.g. `UInt16 + UInt128 = UInt128`. While this introduces a risk of overflows around the value range boundary, it ensures that calculations are performed quickly using the maximum native integer width of 64 bit.
The result of addition or multiplication of two integers is unsigned unless one of the integers is signed.
Example:
@ -20,39 +22,78 @@ SELECT toTypeName(0), toTypeName(0 + 0), toTypeName(0 + 0 + 0), toTypeName(0 + 0
└───────────────┴────────────────────────┴─────────────────────────────────┴──────────────────────────────────────────┘
```
Arithmetic functions work for any pair of types from UInt8, UInt16, UInt32, UInt64, Int8, Int16, Int32, Int64, Float32, or Float64.
Arithmetic functions work for any pair of `UInt8`, `UInt16`, `UInt32`, `UInt64`, `Int8`, `Int16`, `Int32`, `Int64`, `Float32`, or `Float64` values.
Overflow is produced the same way as in C++.
Overflows are produced the same way as in C++.
## plus(a, b), a + b operator
## plus
Calculates the sum of the numbers.
You can also add integer numbers with a date or date and time. In the case of a date, adding an integer means adding the corresponding number of days. For a date with time, it means adding the corresponding number of seconds.
Calculates the sum of two values `a` and `b`.
## minus(a, b), a - b operator
**Syntax**
Calculates the difference. The result is always signed.
```sql
plus(a, b)
```
You can also calculate integer numbers from a date or date with time. The idea is the same – see above for ‘plus’.
It is possible to add an integer and a date or date with time. The former operation increments the number of days in the date, the latter operation increments the number of seconds in the date with time.
## multiply(a, b), a \* b operator
Alias: `a + b` (operator)
Calculates the product of the numbers.
## minus
## divide(a, b), a / b operator
Calculates the difference of two values `a` and `b`. The result is always signed.
Calculates the quotient of the numbers. The result type is always a floating-point type.
It is not integer division. For integer division, use the ‘intDiv’ function.
When dividing by zero you get ‘inf’, ‘-inf’, or ‘nan’.
Similar to `plus`, it is possible to subtract an integer from a date or date with time.
## intDiv(a, b)
**Syntax**
Calculates the quotient of the numbers. Divides into integers, rounding down (by the absolute value).
```sql
minus(a, b)
```
Returns an integer of the type of the dividend (the first parameter).
Alias: `a - b` (operator)
## multiply
Calculates the product of two values `a` and `b`.
**Syntax**
```sql
multiply(a, b)
```
Alias: `a \* b` (operator)
## divide
Calculates the quotient of two values `a` and `b`. The result is always a floating-point value. If you need integer division, you can use the `intDiv` function.
Division by 0 returns `inf`, `-inf`, or `nan`.
**Syntax**
```sql
divide(a, b)
```
Alias: `a / b` (operator)
## intDiv
Performs an integer division of two values `a` by `b`, i.e. computes the quotient rounded down to the next smallest integer.
The result has the same type as the dividend (the first parameter).
An exception is thrown when dividing by zero, when the quotient does not fit in the range of the dividend, or when dividing a minimal negative number by minus one.
**Syntax**
```sql
intDiv(a, b)
```
**Example**
Query:
@ -62,6 +103,7 @@ SELECT
intDiv(toFloat64(1), 0.001) AS res,
toTypeName(res)
```
```response
┌──res─┬─toTypeName(intDiv(toFloat64(1), 0.001))─┐
│ 1000 │ Int64 │
@ -73,30 +115,65 @@ SELECT
intDiv(1, 0.001) AS res,
toTypeName(res)
```
```response
Received exception from server (version 23.2.1):
Code: 153. DB::Exception: Received from localhost:9000. DB::Exception: Cannot perform integer division, because it will produce infinite or too large number: While processing intDiv(1, 0.001) AS res, toTypeName(res). (ILLEGAL_DIVISION)
```
## intDivOrZero(a, b)
## intDivOrZero
Differs from ‘intDiv’ in that it returns zero when dividing by zero or when dividing a minimal negative number by minus one.
Same as `intDiv` but returns zero when dividing by zero or when dividing a minimal negative number by minus one.
## modulo(a, b), a % b operator
**Syntax**
```sql
intDivOrZero(a, b)
```
## modulo
Calculates the remainder of the division of two values `a` by `b`.
Calculates the remainder when dividing `a` by `b`.
The result type is an integer if both inputs are integers. If one of the inputs is a floating-point number, the result is a floating-point number.
The remainder is computed like in C++. Truncated division is used for negative numbers.
An exception is thrown when dividing by zero or when dividing a minimal negative number by minus one.
## moduloOrZero(a, b)
**Syntax**
Differs from [modulo](#modulo) in that it returns zero when the divisor is zero.
```sql
modulo(a, b)
```
## positiveModulo(a, b), positive_modulo(a, b), pmod(a, b)
Calculates the remainder when dividing `a` by `b`. Similar to the function `modulo` except that `positive_modulo` always returns a non-negative number.
Alias: `a % b` (operator)
Notice that `positive_modulo` is 4-5 times slower than `modulo`. You should not use `positive_modulo` unless you want to get a positive result and don't care about performance too much.
## moduloOrZero
Like [modulo](#modulo) but returns zero when the divisor is zero.
**Syntax**
```sql
moduloOrZero(a, b)
```
## positiveModulo(a, b)
Like [modulo](#modulo) but always returns a non-negative number.
This function is 4-5 times slower than `modulo`.
**Syntax**
```sql
positiveModulo(a, b)
```
Alias:
- `positive_modulo(a, b)`
- `pmod(a, b)`
**Example**
@ -108,51 +185,67 @@ SELECT positiveModulo(-1, 10)
Result:
```text
```result
┌─positiveModulo(-1, 10)─┐
│ 9 │
└────────────────────────┘
```
## negate(a), -a operator
## negate
Calculates a number with the reverse sign. The result is always signed.
## abs(a)
Calculates the absolute value of the number (a). That is, if a \< 0, it returns -a. For unsigned types it does not do anything. For signed integer types, it returns an unsigned number.
## gcd(a, b)
Returns the greatest common divisor of the numbers.
An exception is thrown when dividing by zero or when dividing a minimal negative number by minus one.
## lcm(a, b)
Returns the least common multiple of the numbers.
An exception is thrown when dividing by zero or when dividing a minimal negative number by minus one.
## max2
Compares two values and returns the maximum. The returned value is converted to [Float64](../../sql-reference/data-types/float.md).
Negates a value `a`. The result is always signed.
**Syntax**
```sql
max2(value1, value2)
negate(a)
```
**Arguments**
Alias: `-a`
- `value1` — First value. [Int/UInt](../../sql-reference/data-types/int-uint.md) or [Float](../../sql-reference/data-types/float.md).
- `value2` — Second value. [Int/UInt](../../sql-reference/data-types/int-uint.md) or [Float](../../sql-reference/data-types/float.md).
## abs
**Returned value**
Calculates the absolute value of `a`. Has no effect if `a` is of an unsigned type. If `a` is of a signed type, it returns an unsigned number.
- The maximum of two values.
**Syntax**
Type: [Float](../../sql-reference/data-types/float.md).
```sql
abs(a)
```
## gcd
Returns the greatest common divisor of two values `a` and `b`.
An exception is thrown when dividing by zero or when dividing a minimal negative number by minus one.
**Syntax**
```sql
gcd(a, b)
```
## lcm(a, b)
Returns the least common multiple of two values `a` and `b`.
An exception is thrown when dividing by zero or when dividing a minimal negative number by minus one.
**Syntax**
```sql
lcm(a, b)
```
## max2
Returns the bigger of two values `a` and `b`. The returned value is of type [Float64](../../sql-reference/data-types/float.md).
**Syntax**
```sql
max2(a, b)
```
**Example**
@ -164,7 +257,7 @@ SELECT max2(-1, 2);
Result:
```text
```result
┌─max2(-1, 2)─┐
│ 2 │
└─────────────┘
@ -172,25 +265,14 @@ Result:
## min2
Compares two values and returns the minimum. The returned value is converted to [Float64](../../sql-reference/data-types/float.md).
Returns the smaller of two values `a` and `b`. The returned value is of type [Float64](../../sql-reference/data-types/float.md).
**Syntax**
```sql
min2(value1, value2)
min2(a, b)
```
**Arguments**
- `value1` — First value. [Int/UInt](../../sql-reference/data-types/int-uint.md) or [Float](../../sql-reference/data-types/float.md).
- `value2` — Second value. [Int/UInt](../../sql-reference/data-types/int-uint.md) or [Float](../../sql-reference/data-types/float.md).
**Returned value**
- The minimum of two values.
Type: [Float](../../sql-reference/data-types/float.md).
**Example**
Query:
@ -201,21 +283,19 @@ SELECT min2(-1, 2);
Result:
```text
```result
┌─min2(-1, 2)─┐
│ -1 │
└─────────────┘
```
## multiplyDecimal(a, b[, result_scale])
## multiplyDecimal
Performs multiplication on two decimals. Result value will be of type [Decimal256](../../sql-reference/data-types/decimal.md).
Result scale can be explicitly specified by `result_scale` argument (const Integer in range `[0, 76]`). If not specified, the result scale is the max scale of given arguments.
Multiplies two decimals `a` and `b`. The result value will be of type [Decimal256](../../sql-reference/data-types/decimal.md).
:::note
These functions work significantly slower than usual `multiply`.
In case you don't really need controlled precision and/or need fast computation, consider using [multiply](#multiply)
:::
The scale of the result can be explicitly specified by `result_scale`. If `result_scale` is not specified, it is assumed to be the maximum scale of the input values.
This function work significantly slower than usual `multiply`. In case no control over the result precision is needed and/or fast computation is desired, consider using `multiply`.
**Syntax**
@ -237,19 +317,22 @@ Type: [Decimal256](../../sql-reference/data-types/decimal.md).
**Example**
```text
```result
┌─multiplyDecimal(toDecimal256(-12, 0), toDecimal32(-2.1, 1), 1)─┐
│ 25.2 │
└────────────────────────────────────────────────────────────────┘
```
**Difference from regular multiplication:**
**Differences compared to regular multiplication:**
```sql
SELECT toDecimal64(-12.647, 3) * toDecimal32(2.1239, 4);
SELECT toDecimal64(-12.647, 3) as a, toDecimal32(2.1239, 4) as b, multiplyDecimal(a, b);
```
```text
Result:
```result
┌─multiply(toDecimal64(-12.647, 3), toDecimal32(2.1239, 4))─┐
│ -26.8609633 │
└───────────────────────────────────────────────────────────┘
@ -270,7 +353,9 @@ SELECT
a * b;
```
```text
Result:
```result
┌─────────────a─┬─────────────b─┬─multiplyDecimal(toDecimal64(-12.647987876, 9), toDecimal64(123.967645643, 9))─┐
│ -12.647987876 │ 123.967645643 │ -1567.941279108 │
└───────────────┴───────────────┴───────────────────────────────────────────────────────────────────────────────┘
@ -279,15 +364,14 @@ Received exception from server (version 22.11.1):
Code: 407. DB::Exception: Received from localhost:9000. DB::Exception: Decimal math overflow: While processing toDecimal64(-12.647987876, 9) AS a, toDecimal64(123.967645643, 9) AS b, a * b. (DECIMAL_OVERFLOW)
```
## divideDecimal(a, b[, result_scale])
## divideDecimal
Performs division on two decimals. Result value will be of type [Decimal256](../../sql-reference/data-types/decimal.md).
Result scale can be explicitly specified by `result_scale` argument (const Integer in range `[0, 76]`). If not specified, the result scale is the max scale of given arguments.
:::note
These function work significantly slower than usual `divide`.
In case you don't really need controlled precision and/or need fast computation, consider using [divide](#divide).
:::
Divides two decimals `a` and `b`. The result value will be of type [Decimal256](../../sql-reference/data-types/decimal.md).
The scale of the result can be explicitly specified by `result_scale`. If `result_scale` is not specified, it is assumed to be the maximum scale of the input values.
This function work significantly slower than usual `divide`. In case no control over the result precision is needed and/or fast computation is desired, consider using `divide`.
**Syntax**
@ -309,19 +393,22 @@ Type: [Decimal256](../../sql-reference/data-types/decimal.md).
**Example**
```text
```result
┌─divideDecimal(toDecimal256(-12, 0), toDecimal32(2.1, 1), 10)─┐
│ -5.7142857142 │
└──────────────────────────────────────────────────────────────┘
```
**Difference from regular division:**
**Differences compared to regular division:**
```sql
SELECT toDecimal64(-12, 1) / toDecimal32(2.1, 1);
SELECT toDecimal64(-12, 1) as a, toDecimal32(2.1, 1) as b, divideDecimal(a, b, 1), divideDecimal(a, b, 5);
```
```text
Result:
```result
┌─divide(toDecimal64(-12, 1), toDecimal32(2.1, 1))─┐
│ -5.7 │
└──────────────────────────────────────────────────┘
@ -336,7 +423,9 @@ SELECT toDecimal64(-12, 0) / toDecimal32(2.1, 1);
SELECT toDecimal64(-12, 0) as a, toDecimal32(2.1, 1) as b, divideDecimal(a, b, 1), divideDecimal(a, b, 5);
```
```text
Result:
```result
DB::Exception: Decimal result's scale is less than argument's one: While processing toDecimal64(-12, 0) / toDecimal32(2.1, 1). (ARGUMENT_OUT_OF_BOUND)
┌───a─┬───b─┬─divideDecimal(toDecimal64(-12, 0), toDecimal32(2.1, 1), 1)─┬─divideDecimal(toDecimal64(-12, 0), toDecimal32(2.1, 1), 5)─┐

53
docs/en/sql-reference/functions/array-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/array-functions
sidebar_position: 35
sidebar_position: 10
sidebar_label: Arrays
---
@ -118,12 +118,12 @@ Accepts zero arguments and returns an empty array of the appropriate type.
Accepts an empty array and returns a one-element array that is equal to the default value.
## range(end), range(\[start, \] end \[, step\])
Returns an array of numbers from `start` to `end - 1` by `step`. The supported types are [UInt8, UInt16, UInt32, UInt64, Int8, Int16, Int32, Int64](../data-types/int-uint.md).
**Syntax**
``` sql
range([start, ] end [, step])
```
@ -146,10 +146,13 @@ range([start, ] end [, step])
**Examples**
Query:
``` sql
SELECT range(5), range(1, 5), range(1, 5, 2), range(-1, 5, 2);
```
Result:
```txt
┌─range(5)────┬─range(1, 5)─┬─range(1, 5, 2)─┬─range(-1, 5, 2)─┐
│ [0,1,2,3,4] │ [1,2,3,4] │ [1,3] │ [-1,1,3] │
@ -301,6 +304,7 @@ In other words, the functions will check whether all the elements of `array2` ar
the `hasAll` function. In addition, it will check that the elements are observed in the same order in both `array1` and `array2`.
For Example:
- `hasSubstr([1,2,3,4], [2,3])` returns 1. However, `hasSubstr([1,2,3,4], [3,2])` will return `0`.
- `hasSubstr([1,2,3,4], [1,2,3])` returns 1. However, `hasSubstr([1,2,3,4], [1,2,4])` will return `0`.
@ -857,7 +861,7 @@ A special function. See the section [“ArrayJoin function”](../../sql-referen
## arrayDifference
Calculates the difference between adjacent array elements. Returns an array where the first element will be 0, the second is the difference between `a[1] - a[0]`, etc. The type of elements in the resulting array is determined by the type inference rules for subtraction (e.g. `UInt8` - `UInt8` = `Int16`).
Calculates an array of differences between adjacent array elements. The first element of the result array will be 0, the second `a[1] - a[0]`, the third `a[2] - a[1]`, etc. The type of elements in the result array is determined by the type inference rules for subtraction (e.g. `UInt8` - `UInt8` = `Int16`).
**Syntax**
@ -871,7 +875,7 @@ arrayDifference(array)
**Returned values**
Returns an array of differences between adjacent elements.
Returns an array of differences between adjacent array elements.
Type: [UInt\*](https://clickhouse.com/docs/en/data_types/int_uint/#uint-ranges), [Int\*](https://clickhouse.com/docs/en/data_types/int_uint/#int-ranges), [Float\*](https://clickhouse.com/docs/en/data_types/float/).
@ -1213,7 +1217,7 @@ Result:
## arrayAUC
Calculate AUC (Area Under the Curve, which is a concept in machine learning, see more details: https://en.wikipedia.org/wiki/Receiver_operating_characteristic#Area_under_the_curve).
Calculate AUC (Area Under the Curve, which is a concept in machine learning, see more details: <https://en.wikipedia.org/wiki/Receiver_operating_characteristic#Area_under_the_curve>).
**Syntax**
@ -1639,7 +1643,23 @@ Result:
## arrayCumSum(\[func,\] arr1, …)
Returns an array of partial sums of elements in the source array (a running sum). If the `func` function is specified, then the values of the array elements are converted by `func(arr1[i], …, arrN[i])` before summing.
Returns an array of the partial (running) sums of the elements in the source array `arr1`. If `func` is specified, then the sum is computed from applying `func` to `arr1`, `arr2`, ..., `arrN`, i.e. `func(arr1[i], …, arrN[i])`.
**Syntax**
``` sql
arrayCumSum(arr)
```
**Arguments**
- `arr` — [Array](../../sql-reference/data-types/array.md) of numeric values.
**Returned value**
- Returns an array of the partial sums of the elements in the source array.
Type: [UInt\*](https://clickhouse.com/docs/en/data_types/int_uint/#uint-ranges), [Int\*](https://clickhouse.com/docs/en/data_types/int_uint/#int-ranges), [Float\*](https://clickhouse.com/docs/en/data_types/float/).
Example:
@ -1655,9 +1675,25 @@ SELECT arrayCumSum([1, 1, 1, 1]) AS res
Note that the `arrayCumSum` is a [higher-order function](../../sql-reference/functions/index.md#higher-order-functions). You can pass a lambda function to it as the first argument.
## arrayCumSumNonNegative(arr)
## arrayCumSumNonNegative(\[func,\] arr1, …)
Same as `arrayCumSum`, returns an array of partial sums of elements in the source array (a running sum). Different `arrayCumSum`, when then returned value contains a value less than zero, the value is replace with zero and the subsequent calculation is performed with zero parameters. For example:
Same as `arrayCumSum`, returns an array of the partial (running) sums of the elements in the source array. If `func` is specified, then the sum is computed from applying `func` to `arr1`, `arr2`, ..., `arrN`, i.e. `func(arr1[i], …, arrN[i])`. Unlike `arrayCumSum`, if the current running sum is smaller than `0`, it is replaced by `0`.
**Syntax**
``` sql
arrayCumSumNonNegative(arr)
```
**Arguments**
- `arr` — [Array](../../sql-reference/data-types/array.md) of numeric values.
**Returned value**
- Returns an array of non-negative partial sums of elements in the source array.
Type: [UInt\*](https://clickhouse.com/docs/en/data_types/int_uint/#uint-ranges), [Int\*](https://clickhouse.com/docs/en/data_types/int_uint/#int-ranges), [Float\*](https://clickhouse.com/docs/en/data_types/float/).
``` sql
SELECT arrayCumSumNonNegative([1, 1, -4, 1]) AS res
@ -1668,6 +1704,7 @@ SELECT arrayCumSumNonNegative([1, 1, -4, 1]) AS res
│ [1,2,0,1] │
└───────────┘
```
Note that the `arraySumNonNegative` is a [higher-order function](../../sql-reference/functions/index.md#higher-order-functions). You can pass a lambda function to it as the first argument.
## arrayProduct

2
docs/en/sql-reference/functions/array-join.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/array-join
sidebar_position: 61
sidebar_position: 15
sidebar_label: arrayJoin
---

2
docs/en/sql-reference/functions/bit-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/bit-functions
sidebar_position: 48
sidebar_position: 20
sidebar_label: Bit
---

180
docs/en/sql-reference/functions/bitmap-functions.md
View File

@ -1,22 +1,18 @@
---
slug: /en/sql-reference/functions/bitmap-functions
sidebar_position: 49
sidebar_position: 25
sidebar_label: Bitmap
---
# Bitmap Functions
Bitmap functions work for two bitmaps Object value calculation, it is to return new bitmap or cardinality while using formula calculation, such as and, or, xor, and not, etc.
There are 2 kinds of construction methods for Bitmap Object. One is to be constructed by aggregation function groupBitmap with -State, the other is to be constructed by Array Object. It is also to convert Bitmap Object to Array Object.
RoaringBitmap is wrapped into a data structure while actual storage of Bitmap objects. When the cardinality is less than or equal to 32, it uses Set objet. When the cardinality is greater than 32, it uses RoaringBitmap object. That is why storage of low cardinality set is faster.
For more information on RoaringBitmap, see: [CRoaring](https://github.com/RoaringBitmap/CRoaring).
Bitmaps can be constructed in two ways. The first way is constructed by aggregation function groupBitmap with `-State`, the other way is to constructed a bitmap from an Array object.
## bitmapBuild
Build a bitmap from unsigned integer array.
Builds a bitmap from an unsigned integer array.
**Syntax**
``` sql
bitmapBuild(array)
@ -40,7 +36,9 @@ SELECT bitmapBuild([1, 2, 3, 4, 5]) AS res, toTypeName(res);
## bitmapToArray
Convert bitmap to integer array.
Converts bitmap to an integer array.
**Syntax**
``` sql
bitmapToArray(bitmap)
@ -56,6 +54,8 @@ bitmapToArray(bitmap)
SELECT bitmapToArray(bitmapBuild([1, 2, 3, 4, 5])) AS res;
```
Result:
``` text
┌─res─────────┐
│ [1,2,3,4,5] │
@ -64,7 +64,9 @@ SELECT bitmapToArray(bitmapBuild([1, 2, 3, 4, 5])) AS res;
## bitmapSubsetInRange
Return subset in specified range (not include the range_end).
Returns the subset of a bitmap with bits within a value interval.
**Syntax**
``` sql
bitmapSubsetInRange(bitmap, range_start, range_end)
@ -73,8 +75,8 @@ bitmapSubsetInRange(bitmap, range_start, range_end)
**Arguments**
- `bitmap` – [Bitmap object](#bitmap_functions-bitmapbuild).
- `range_start` – Range start point. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `range_end` – Range end point (excluded). Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `range_start` – Start of the range (inclusive). Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `range_end` – End of the range (exclusive). Type: [UInt32](../../sql-reference/data-types/int-uint.md).
**Example**
@ -82,6 +84,8 @@ bitmapSubsetInRange(bitmap, range_start, range_end)
SELECT bitmapToArray(bitmapSubsetInRange(bitmapBuild([0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,100,200,500]), toUInt32(30), toUInt32(200))) AS res;
```
Result:
``` text
┌─res───────────────┐
│ [30,31,32,33,100] │
@ -90,7 +94,7 @@ SELECT bitmapToArray(bitmapSubsetInRange(bitmapBuild([0,1,2,3,4,5,6,7,8,9,10,11,
## bitmapSubsetLimit
Creates a subset of bitmap with n elements taken between `range_start` and `cardinality_limit`.
Returns a subset of a bitmap with smallest bit value `range_start` and at most `cardinality_limit` elements.
**Syntax**
@ -101,19 +105,11 @@ bitmapSubsetLimit(bitmap, range_start, cardinality_limit)
**Arguments**
- `bitmap` – [Bitmap object](#bitmap_functions-bitmapbuild).
- `range_start` – The subset starting point. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `cardinality_limit` – The subset cardinality upper limit. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
**Returned value**
The subset.
Type: [Bitmap object](#bitmap_functions-bitmapbuild).
- `range_start` – Start of the range (inclusive). Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `cardinality_limit` – Maximum cardinality of the subset. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT bitmapToArray(bitmapSubsetLimit(bitmapBuild([0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,100,200,500]), toUInt32(30), toUInt32(200))) AS res;
```
@ -128,7 +124,7 @@ Result:
## subBitmap
Returns the bitmap elements, starting from the `offset` position. The number of returned elements is limited by the `cardinality_limit` parameter. Analog of the [substring](string-functions.md#substring)) string function, but for bitmap.
Returns a subset of the bitmap, starting from position `offset`. The maximum cardinality of the returned bitmap is `cardinality_limit`.
**Syntax**
@ -142,16 +138,8 @@ subBitmap(bitmap, offset, cardinality_limit)
- `offset` – The position of the first element of the subset. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `cardinality_limit` – The maximum number of elements in the subset. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
**Returned value**
The subset.
Type: [Bitmap object](#bitmap_functions-bitmapbuild).
**Example**
Query:
``` sql
SELECT bitmapToArray(subBitmap(bitmapBuild([0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,100,200,500]), toUInt32(10), toUInt32(10))) AS res;
```
@ -169,18 +157,18 @@ Result:
Checks whether the bitmap contains an element.
``` sql
bitmapContains(haystack, needle)
bitmapContains(bitmap, needle)
```
**Arguments**
- `haystack` – [Bitmap object](#bitmap_functions-bitmapbuild), where the function searches.
- `needle` – Value that the function searches. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
- `bitmap` – [Bitmap object](#bitmap_functions-bitmapbuild).
- `needle` – Searched bit value. Type: [UInt32](../../sql-reference/data-types/int-uint.md).
**Returned values**
- 0 — If `haystack` does not contain `needle`.
- 1 — If `haystack` contains `needle`.
- 0 — If `bitmap` does not contain `needle`.
- 1 — If `bitmap` contains `needle`.
Type: `UInt8`.
@ -190,6 +178,8 @@ Type: `UInt8`.
SELECT bitmapContains(bitmapBuild([1,5,7,9]), toUInt32(9)) AS res;
```
Result:
``` text
┌─res─┐
│ 1 │
@ -198,21 +188,24 @@ SELECT bitmapContains(bitmapBuild([1,5,7,9]), toUInt32(9)) AS res;
## bitmapHasAny
Checks whether two bitmaps have intersection by some elements.
Checks whether two bitmaps intersect.
If `bitmap2` contains exactly one element, consider using [bitmapContains](#bitmap_functions-bitmapcontains) instead as it works more efficiently.
**Syntax**
``` sql
bitmapHasAny(bitmap1, bitmap2)
```
If you are sure that `bitmap2` contains strictly one element, consider using the [bitmapContains](#bitmap_functions-bitmapcontains) function. It works more efficiently.
**Arguments**
- `bitmap*` – Bitmap object.
- `bitmap1` – Bitmap object 1.
- `bitmap2` – Bitmap object 2.
**Return values**
- `1`, if `bitmap1` and `bitmap2` have one similar element at least.
- `1`, if `bitmap1` and `bitmap2` have at least one shared element.
- `0`, otherwise.
**Example**
@ -221,6 +214,8 @@ If you are sure that `bitmap2` contains strictly one element, consider using the
SELECT bitmapHasAny(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 1 │
@ -229,16 +224,21 @@ SELECT bitmapHasAny(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
## bitmapHasAll
Analogous to `hasAll(array, array)` returns 1 if the first bitmap contains all the elements of the second one, 0 otherwise.
If the second argument is an empty bitmap then returns 1.
Returns 1 if the first bitmap contains all elements of the second bitmap, otherwise 0.
If the second bitmap is empty, returns 1.
Also see `hasAll(array, array)`.
**Syntax**
``` sql
bitmapHasAll(bitmap,bitmap)
bitmapHasAll(bitmap1, bitmap2)
```
**Arguments**
- `bitmap` – Bitmap object.
- `bitmap1` – Bitmap object 1.
- `bitmap2` – Bitmap object 2.
**Example**
@ -246,6 +246,8 @@ bitmapHasAll(bitmap,bitmap)
SELECT bitmapHasAll(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 0 │
@ -254,7 +256,9 @@ SELECT bitmapHasAll(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
## bitmapCardinality
Retrun bitmap cardinality of type UInt64.
Rerturn the cardinality of a bitmap.
**Syntax**
``` sql
bitmapCardinality(bitmap)
@ -270,6 +274,8 @@ bitmapCardinality(bitmap)
SELECT bitmapCardinality(bitmapBuild([1, 2, 3, 4, 5])) AS res;
```
Result:
``` text
┌─res─┐
│ 5 │
@ -278,9 +284,13 @@ SELECT bitmapCardinality(bitmapBuild([1, 2, 3, 4, 5])) AS res;
## bitmapMin
Retrun the smallest value of type UInt64 in the set, UINT32_MAX if the set is empty.
Computes the smallest bit set in a bitmap, or UINT32_MAX if the bitmap is empty.
bitmapMin(bitmap)
**Syntax**
```sql
bitmapMin(bitmap)
```
**Arguments**
@ -292,6 +302,8 @@ Retrun the smallest value of type UInt64 in the set, UINT32_MAX if the set is em
SELECT bitmapMin(bitmapBuild([1, 2, 3, 4, 5])) AS res;
```
Result:
``` text
┌─res─┐
│ 1 │
@ -300,9 +312,13 @@ SELECT bitmapMin(bitmapBuild([1, 2, 3, 4, 5])) AS res;
## bitmapMax
Retrun the greatest value of type UInt64 in the set, 0 if the set is empty.
Computes the greatest bit set in a bitmap, or 0 if the bitmap is empty.
bitmapMax(bitmap)
**Syntax**
```sql
bitmapMax(bitmap)
```
**Arguments**
@ -314,6 +330,8 @@ Retrun the greatest value of type UInt64 in the set, 0 if the set is empty.
SELECT bitmapMax(bitmapBuild([1, 2, 3, 4, 5])) AS res;
```
Result:
``` text
┌─res─┐
│ 5 │
@ -322,15 +340,21 @@ SELECT bitmapMax(bitmapBuild([1, 2, 3, 4, 5])) AS res;
## bitmapTransform
Transform an array of values in a bitmap to another array of values, the result is a new bitmap.
Replaces at most N bits in a bitmap. The old and new value of the i-th replaced bit is given by `from_array[i]` and `to_array[i]`.
bitmapTransform(bitmap, from_array, to_array)
The result depends on the array ordering if `from_array` and `to_array`.
**Syntax**
``` sql
bitmapTransform(bitmap, from_array, to_array)
```
**Arguments**
- `bitmap` – Bitmap object.
- `from_array` – UInt32 array. For idx in range \[0, from_array.size()), if bitmap contains from_array\[idx\], then replace it with to_array\[idx\]. Note that the result depends on array ordering if there are common elements between from_array and to_array.
- `to_array` – UInt32 array, its size shall be the same to from_array.
- `from_array` – UInt32 array. For idx in range \[0, from_array.size()), if bitmap contains from_array\[idx\], then replace it with to_array\[idx\].
- `to_array` – UInt32 array with the same size as `from_array`.
**Example**
@ -338,6 +362,8 @@ Transform an array of values in a bitmap to another array of values, the result
SELECT bitmapToArray(bitmapTransform(bitmapBuild([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]), cast([5,999,2] as Array(UInt32)), cast([2,888,20] as Array(UInt32)))) AS res;
```
Result:
``` text
┌─res───────────────────┐
│ [1,3,4,6,7,8,9,10,20] │
@ -346,7 +372,9 @@ SELECT bitmapToArray(bitmapTransform(bitmapBuild([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
## bitmapAnd
Two bitmap and calculation, the result is a new bitmap.
Computes the logical conjunction of two two bitmaps.
**Syntax**
``` sql
bitmapAnd(bitmap,bitmap)
@ -362,6 +390,8 @@ bitmapAnd(bitmap,bitmap)
SELECT bitmapToArray(bitmapAnd(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS res;
```
Result:
``` text
┌─res─┐
│ [3] │
@ -370,7 +400,9 @@ SELECT bitmapToArray(bitmapAnd(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS re
## bitmapOr
Two bitmap or calculation, the result is a new bitmap.
Computes the logical disjunction of two bitmaps.
**Syntax**
``` sql
bitmapOr(bitmap,bitmap)
@ -386,6 +418,8 @@ bitmapOr(bitmap,bitmap)
SELECT bitmapToArray(bitmapOr(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS res;
```
Result:
``` text
┌─res─────────┐
│ [1,2,3,4,5] │
@ -394,7 +428,9 @@ SELECT bitmapToArray(bitmapOr(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS res
## bitmapXor
Two bitmap xor calculation, the result is a new bitmap.
Xor-s two bitmaps.
**Syntax**
``` sql
bitmapXor(bitmap,bitmap)
@ -410,6 +446,8 @@ bitmapXor(bitmap,bitmap)
SELECT bitmapToArray(bitmapXor(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS res;
```
Result:
``` text
┌─res───────┐
│ [1,2,4,5] │
@ -418,7 +456,9 @@ SELECT bitmapToArray(bitmapXor(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS re
## bitmapAndnot
Two bitmap andnot calculation, the result is a new bitmap.
Computes the logical conjunction of two bitmaps and negates the result.
**Syntax**
``` sql
bitmapAndnot(bitmap,bitmap)
@ -434,6 +474,8 @@ bitmapAndnot(bitmap,bitmap)
SELECT bitmapToArray(bitmapAndnot(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS res;
```
Result:
``` text
┌─res───┐
│ [1,2] │
@ -442,7 +484,9 @@ SELECT bitmapToArray(bitmapAndnot(bitmapBuild([1,2,3]),bitmapBuild([3,4,5]))) AS
## bitmapAndCardinality
Two bitmap and calculation, return cardinality of type UInt64.
Returns the cardinality of the logical conjunction of two bitmaps.
**Syntax**
``` sql
bitmapAndCardinality(bitmap,bitmap)
@ -458,6 +502,8 @@ bitmapAndCardinality(bitmap,bitmap)
SELECT bitmapAndCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 1 │
@ -466,7 +512,7 @@ SELECT bitmapAndCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
## bitmapOrCardinality
Two bitmap or calculation, return cardinality of type UInt64.
Returns the cardinality of the logical disjunction of two bitmaps.
``` sql
bitmapOrCardinality(bitmap,bitmap)
@ -482,6 +528,8 @@ bitmapOrCardinality(bitmap,bitmap)
SELECT bitmapOrCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 5 │
@ -490,7 +538,7 @@ SELECT bitmapOrCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
## bitmapXorCardinality
Two bitmap xor calculation, return cardinality of type UInt64.
Returns the cardinality of the XOR of two bitmaps.
``` sql
bitmapXorCardinality(bitmap,bitmap)
@ -506,6 +554,8 @@ bitmapXorCardinality(bitmap,bitmap)
SELECT bitmapXorCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 4 │
@ -514,7 +564,7 @@ SELECT bitmapXorCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
## bitmapAndnotCardinality
Two bitmap andnot calculation, return cardinality of type UInt64.
Returns the cardinality of the AND-NOT operation of two bitmaps.
``` sql
bitmapAndnotCardinality(bitmap,bitmap)
@ -530,6 +580,8 @@ bitmapAndnotCardinality(bitmap,bitmap)
SELECT bitmapAndnotCardinality(bitmapBuild([1,2,3]),bitmapBuild([3,4,5])) AS res;
```
Result:
``` text
┌─res─┐
│ 2 │

79
docs/en/sql-reference/functions/comparison-functions.md
View File

@ -1,34 +1,89 @@
---
slug: /en/sql-reference/functions/comparison-functions
sidebar_position: 36
sidebar_position: 35
sidebar_label: Comparison
---
# Comparison Functions
Comparison functions always return 0 or 1 (Uint8).
Below comparison functions return 0 or 1 as Uint8.
The following types can be compared:
- numbers
- strings and fixed strings
- dates
- dates with times
within each group, but not between different groups.
Only values within the same group can be compared (e.g. UInt16 and UInt64) but not accross groups (e.g. UInt16 and DateTime).
For example, you can’t compare a date with a string. You have to use a function to convert the string to a date, or vice versa.
Strings are compared byte-by-byte. Note that this may lead to unexpected results if one of the strings contains UTF-8 encoded multi-byte characters.
Strings are compared by bytes. A shorter string is smaller than all strings that start with it and that contain at least one more character.
A string S1 which has another string S2 as prefix is considered longer than S2.
### equals, a `=` b and a `==` b operator
## equals
### notEquals, a `!=` b and a `<>` b operator
**Syntax**
### less, `<` operator
```sql
equals(a, b)
```
### greater, `>` operator
Alias:
- `a = b` (operator)
- `a == b` (operator)
### lessOrEquals, `<=` operator
## notEquals
### greaterOrEquals, `>=` operator
**Syntax**
```sql
notEquals(a, b)
```
Alias:
- `a != b` (operator)
- `a <> b` (operator)
## less
**Syntax**
```sql
less(a, b)
```
Alias:
- `a < b` (operator)
## greater
**Syntax**
```sql
greater(a, b)
```
Alias:
- `a > b` (operator)
## lessOrEquals
**Syntax**
```sql
lessOrEquals(a, b)
```
Alias:
- `a <= b` (operator)
## greaterOrEquals, `>=` operator
**Syntax**
```sql
greaterOrEquals(a, b)
```
Alias:
- `a >= b` (operator)

130
docs/en/sql-reference/functions/conditional-functions.md
View File

@ -1,39 +1,40 @@
---
slug: /en/sql-reference/functions/conditional-functions
sidebar_position: 43
sidebar_label: 'Conditional '
sidebar_position: 40
sidebar_label: Conditional
---
# Conditional Functions
## if
Controls conditional branching. Unlike most systems, ClickHouse always evaluate both expressions `then` and `else`.
Performs conditional branching.
If the condition `cond` evaluates to a non-zero value, the function returns the result of the expression `then`. If `cond` evaluates to zero or `NULL`, then the result of the `else` expression is returned.
Setting [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) controls whether short-circuit evaluation is used. If enabled, the `then` expression is evaluated only on rows where `cond` is `true` and the `else` expression where `cond` is `false`. For example, with short-circuit evaluation, no division-by-zero exception is thrown when executing the query `SELECT if(number = 0, 0, intDiv(42, number)) FROM numbers(10)`.
`then` and `else` must be of a similar type.
**Syntax**
``` sql
if(cond, then, else)
```
If the condition `cond` evaluates to a non-zero value, returns the result of the expression `then`, and the result of the expression `else`, if present, is skipped. If the `cond` is zero or `NULL`, then the result of the `then` expression is skipped and the result of the `else` expression, if present, is returned.
You can use the [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) setting to calculate the `if` function according to a short scheme. If this setting is enabled, `then` expression is evaluated only on rows where `cond` is true, `else` expression – where `cond` is false. For example, an exception about division by zero is not thrown when executing the query `SELECT if(number = 0, 0, intDiv(42, number)) FROM numbers(10)`, because `intDiv(42, number)` will be evaluated only for numbers that doesn't satisfy condition `number = 0`.
Alias: `cond ? then : else` (ternary operator)
**Arguments**
- `cond` – The condition for evaluation that can be zero or not. The type is UInt8, Nullable(UInt8) or NULL.
- `then` – The expression to return if condition is met.
- `else` – The expression to return if condition is not met.
- `cond` – The evaluated condition. UInt8, Nullable(UInt8) or NULL.
- `then` – The expression returned if `condition` is true.
- `else` – The expression returned if `condition` is `false` or NULL.
**Returned values**
The function executes `then` and `else` expressions and returns its result, depending on whether the condition `cond` ended up being zero or not.
The result of either the `then` and `else` expressions, depending on condition `cond`.
**Example**
Query:
``` sql
SELECT if(1, plus(2, 2), plus(2, 6));
```
@ -46,30 +47,34 @@ Result:
└────────────┘
```
Query:
## multiIf
Allows to write the [CASE](../../sql-reference/operators/index.md#operator_case) operator more compactly in the query.
**Syntax**
``` sql
SELECT if(0, plus(2, 2), plus(2, 6));
multiIf(cond_1, then_1, cond_2, then_2, ..., else)
```
Result:
Setting [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) controls whether short-circuit evaluation is used. If enabled, the `then_i` expression is evaluated only on rows where `((NOT cond_1) AND (NOT cond_2) AND ... AND (NOT cond_{i-1}) AND cond_i)` is `true`, `cond_i` will be evaluated only on rows where `((NOT cond_1) AND (NOT cond_2) AND ... AND (NOT cond_{i-1}))` is `true`. For example, with short-circuit evaluation, no division-by-zero exception is thrown when executing the query `SELECT multiIf(number = 2, intDiv(1, number), number = 5) FROM numbers(10)`.
**Arguments**
The function accepts `2N+1` parameters:
- `cond_N` — The N-th evaluated condition which controls if `then_N` is returned.
- `then_N` — The result of the function when `cond_N` is true.
- `else` — The result of the function if none of conditions is true.
**Returned values**
The result of either any of the `then_N` or `else` expressions, depending on the conditions `cond_N`.
**Example**
Assuming this table:
``` text
┌─plus(2, 6)─┐
│ 8 │
└────────────┘
```
- `then` and `else` must have the lowest common type.
**Example:**
Take this `LEFT_RIGHT` table:
``` sql
SELECT *
FROM LEFT_RIGHT
┌─left─┬─right─┐
│ ᴺᵁᴸᴸ │ 4 │
│ 1 │ 3 │
@ -79,69 +84,6 @@ FROM LEFT_RIGHT
└──────┴───────┘
```
The following query compares `left` and `right` values:
``` sql
SELECT
left,
right,
if(left < right, 'left is smaller than right', 'right is greater or equal than left') AS is_smaller
FROM LEFT_RIGHT
WHERE isNotNull(left) AND isNotNull(right)
┌─left─┬─right─┬─is_smaller──────────────────────────┐
│ 1 │ 3 │ left is smaller than right │
│ 2 │ 2 │ right is greater or equal than left │
│ 3 │ 1 │ right is greater or equal than left │
└──────┴───────┴─────────────────────────────────────┘
```
Note: `NULL` values are not used in this example, check [NULL values in conditionals](#null-values-in-conditionals) section.
## Ternary Operator
It works same as `if` function.
Syntax: `cond ? then : else`
Returns `then` if the `cond` evaluates to be true (greater than zero), otherwise returns `else`.
- `cond` must be of type of `UInt8`, and `then` and `else` must have the lowest common type.
- `then` and `else` can be `NULL`
**See also**
- [ifNotFinite](../../sql-reference/functions/other-functions.md#ifnotfinite).
## multiIf
Allows you to write the [CASE](../../sql-reference/operators/index.md#operator_case) operator more compactly in the query.
**Syntax**
``` sql
multiIf(cond_1, then_1, cond_2, then_2, ..., else)
```
You can use the [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) setting to calculate the `multiIf` function according to a short scheme. If this setting is enabled, `then_i` expression is evaluated only on rows where `((NOT cond_1) AND (NOT cond_2) AND ... AND (NOT cond_{i-1}) AND cond_i)` is true, `cond_i` will be evaluated only on rows where `((NOT cond_1) AND (NOT cond_2) AND ... AND (NOT cond_{i-1}))` is true. For example, an exception about division by zero is not thrown when executing the query `SELECT multiIf(number = 2, intDiv(1, number), number = 5) FROM numbers(10)`.
**Arguments**
- `cond_N` — The condition for the function to return `then_N`.
- `then_N` — The result of the function when executed.
- `else` — The result of the function if none of the conditions is met.
The function accepts `2N+1` parameters.
**Returned values**
The function returns one of the values `then_N` or `else`, depending on the conditions `cond_N`.
**Example**
Again using `LEFT_RIGHT` table.
``` sql
SELECT
left,

88
docs/en/sql-reference/functions/date-time-functions.md
View File

@ -1,14 +1,14 @@
---
slug: /en/sql-reference/functions/date-time-functions
sidebar_position: 39
sidebar_position: 45
sidebar_label: Dates and Times
---
# Functions for Working with Dates and Times
Support for time zones.
Most functions in this section accept an optional time zone argument, e.g. `Europe/Amsterdam`. In this case, the time zone is the specified one instead of the local (default) one.
All functions for working with the date and time that have a logical use for the time zone can accept a second optional time zone argument. Example: Asia/Yekaterinburg. In this case, they use the specified time zone instead of the local (default) one.
**Example**
``` sql
SELECT
@ -27,11 +27,11 @@ SELECT
## timeZone
Returns the default timezone of the server for current session. This can be modified using `SET session_timezone = 'New/Value'`
If it is executed in the context of a distributed table, then it generates a normal column with values relevant to each shard. Otherwise it produces a constant value.
If the function is executed in the context of a distributed table, then it generates a normal column with values relevant to each shard, otherwise it produces a constant value.
**Syntax**
``` sql
```sql
timeZone()
```
@ -64,7 +64,7 @@ Type: [String](../../sql-reference/data-types/string.md).
## toTimeZone
Converts time or date and time to the specified time zone. The time zone is an attribute of the `Date` and `DateTime` data types. The internal value (number of seconds) of the table field or of the resultset's column does not change, the column's type changes and its string representation changes accordingly.
Converts a date or date with time to the specified time zone. Does not change the internal value (number of unix seconds) of the data, only the value's time zone attribute and the value's string representation changes.
**Syntax**
@ -87,8 +87,6 @@ Type: [DateTime](../../sql-reference/data-types/datetime.md).
**Example**
Query:
```sql
SELECT toDateTime('2019-01-01 00:00:00', 'UTC') AS time_utc,
toTypeName(time_utc) AS type_utc,
@ -118,8 +116,6 @@ type_samoa: DateTime('US/Samoa')
int32samoa: 1546300800
```
`toTimeZone(time_utc, 'Asia/Yekaterinburg')` changes the `DateTime('UTC')` type to `DateTime('Asia/Yekaterinburg')`. The value (Unixtimestamp) 1546300800 stays the same, but the string representation (the result of the toString() function) changes from `time_utc: 2019-01-01 00:00:00` to `time_yekat: 2019-01-01 05:00:00`.
## timeZoneOf
Returns the timezone name of [DateTime](../../sql-reference/data-types/datetime.md) or [DateTime64](../../sql-reference/data-types/datetime64.md) data types.
@ -144,7 +140,6 @@ Type: [String](../../sql-reference/data-types/string.md).
**Example**
Query:
``` sql
SELECT timezoneOf(now());
```
@ -158,8 +153,9 @@ Result:
## timeZoneOffset
Returns a timezone offset in seconds from [UTC](https://en.wikipedia.org/wiki/Coordinated_Universal_Time). The function takes into account [daylight saving time](https://en.wikipedia.org/wiki/Daylight_saving_time) and historical timezone changes at the specified date and time.
[IANA timezone database](https://www.iana.org/time-zones) is used to calculate the offset.
Returns the timezone offset in seconds from [UTC](https://en.wikipedia.org/wiki/Coordinated_Universal_Time).
The function [daylight saving time](https://en.wikipedia.org/wiki/Daylight_saving_time) and historical timezone changes at the specified date and time into account.
The [IANA timezone database](https://www.iana.org/time-zones) is used to calculate the offset.
**Syntax**
@ -181,8 +177,6 @@ Type: [Int32](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT toDateTime('2021-04-21 10:20:30', 'America/New_York') AS Time, toTypeName(Time) AS Type,
timeZoneOffset(Time) AS Offset_in_seconds, (Offset_in_seconds / 3600) AS Offset_in_hours;
@ -198,37 +192,37 @@ Result:
## toYear
Converts a date or date with time to a UInt16 number containing the year number (AD).
Converts a date or date with time to the year number (AD) as UInt16 value.
Alias: `YEAR`.
## toQuarter
Converts a date or date with time to a UInt8 number containing the quarter number.
Converts a date or date with time to the quarter number as UInt8 value.
Alias: `QUARTER`.
## toMonth
Converts a date or date with time to a UInt8 number containing the month number (1-12).
Converts a date or date with time to the month number (1-12) as UInt8 value.
Alias: `MONTH`.
## toDayOfYear
Converts a date or date with time to a UInt16 number containing the number of the day of the year (1-366).
Converts a date or date with time to the number of the day of the year (1-366) as UInt16 value.
Alias: `DAYOFYEAR`.
## toDayOfMonth
Converts a date or date with time to a UInt8 number containing the number of the day of the month (1-31).
Converts a date or date with time to the number of the day in the month (1-31) as UInt8 value.
Aliases: `DAYOFMONTH`, `DAY`.
## toDayOfWeek
Converts a date or date with time to a UInt8 number containing the number of the day of the week.
Converts a date or date with time to the number of the day in the week as UInt8 value.
The two-argument form of `toDayOfWeek()` enables you to specify whether the week starts on Monday or Sunday, and whether the return value should be in the range from 0 to 6 or 1 to 7. If the mode argument is ommited, the default mode is 0. The time zone of the date can be specified as the third argument.
@ -249,27 +243,28 @@ toDayOfWeek(t[, mode[, timezone]])
## toHour
Converts a date with time to a UInt8 number containing the number of the hour in 24-hour time (0-23).
This function assumes that if clocks are moved ahead, it is by one hour and occurs at 2 a.m., and if clocks are moved back, it is by one hour and occurs at 3 a.m. (which is not always true – even in Moscow the clocks were twice changed at a different time).
Converts a date with time the number of the hour in 24-hour time (0-23) as UInt8 value.
Assumes that if clocks are moved ahead, it is by one hour and occurs at 2 a.m., and if clocks are moved back, it is by one hour and occurs at 3 a.m. (which is not always true – even in Moscow the clocks were twice changed at a different time).
Alias: `HOUR`.
## toMinute
Converts a date with time to a UInt8 number containing the number of the minute of the hour (0-59).
Converts a date with time to the number of the minute of the hour (0-59) as UInt8 value.
Alias: `MINUTE`.
## toSecond
Converts a date with time to a UInt8 number containing the number of the second in the minute (0-59).
Leap seconds are not accounted for.
Converts a date with time to the second in the minute (0-59) as UInt8 value. Leap seconds are not considered.
Alias: `SECOND`.
## toUnixTimestamp
For DateTime argument: converts value to the number with type UInt32 -- Unix Timestamp (https://en.wikipedia.org/wiki/Unix_time).
For DateTime arguments: converts the value to the number with type UInt32 -- Unix Timestamp (https://en.wikipedia.org/wiki/Unix_time).
For String argument: converts the input string to the datetime according to the timezone (optional second argument, server timezone is used by default) and returns the corresponding unix timestamp.
**Syntax**
@ -287,8 +282,6 @@ Type: `UInt32`.
**Example**
Query:
``` sql
SELECT toUnixTimestamp('2017-11-05 08:07:47', 'Asia/Tokyo') AS unix_timestamp
```
@ -615,8 +608,6 @@ Type: [Int](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT age('hour', toDateTime('2018-01-01 22:30:00'), toDateTime('2018-01-02 23:00:00'));
```
@ -629,8 +620,6 @@ Result:
└───────────────────────────────────────────────────────────────────────────────────┘
```
Query:
``` sql
SELECT
toDate('2022-01-01') AS e,
@ -694,8 +683,6 @@ Type: [Int](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT dateDiff('hour', toDateTime('2018-01-01 22:00:00'), toDateTime('2018-01-02 23:00:00'));
```
@ -708,8 +695,6 @@ Result:
└────────────────────────────────────────────────────────────────────────────────────────┘
```
Query:
``` sql
SELECT
toDate('2022-01-01') AS e,
@ -833,8 +818,6 @@ Type: [Date](../../sql-reference/data-types/date.md) or [DateTime](../../sql-ref
**Example**
Query:
```sql
SELECT date_add(YEAR, 3, toDate('2018-01-01'));
```
@ -885,8 +868,6 @@ Type: [Date](../../sql-reference/data-types/date.md) or [DateTime](../../sql-ref
**Example**
Query:
``` sql
SELECT date_sub(YEAR, 3, toDate('2018-01-01'));
```
@ -935,8 +916,6 @@ Type: [Date](../../sql-reference/data-types/date.md) or [DateTime](../../sql-ref
**Example**
Query:
```sql
select timestamp_add(toDate('2018-01-01'), INTERVAL 3 MONTH);
```
@ -986,8 +965,6 @@ Type: [Date](../../sql-reference/data-types/date.md) or [DateTime](../../sql-ref
**Example**
Query:
```sql
select timestamp_sub(MONTH, 5, toDateTime('2018-12-18 01:02:03'));
```
@ -1320,8 +1297,6 @@ Note 2: In ClickHouse versions earlier than v23.4, `%M` prints the minute (00-59
**Example**
Query:
``` sql
SELECT formatDateTime(toDate('2010-01-04'), '%g')
```
@ -1334,8 +1309,6 @@ Result:
└────────────────────────────────────────────┘
```
Query:
``` sql
SELECT formatDateTime(toDateTime64('2010-01-04 12:34:56.123456', 7), '%f')
```
@ -1392,8 +1365,6 @@ Using replacement fields, you can define a pattern for the resulting string.
**Example**
Query:
``` sql
SELECT formatDateTimeInJodaSyntax(toDateTime('2010-01-04 12:34:56'), 'yyyy-MM-dd HH:mm:ss')
```
@ -1431,8 +1402,6 @@ Type: [String](../../sql-reference/data-types/string.md#string)
**Example**
Query:
```sql
WITH toDateTime('2021-04-14 11:22:33') AS date_value
SELECT
@ -1471,8 +1440,6 @@ Type: [String](../../sql-reference/data-types/string.md#string)
**Example**
Query:
```sql
WITH toDateTime('2021-04-14 11:22:33') AS date_value
SELECT monthName(date_value);
@ -1496,8 +1463,6 @@ Alias: `FROM_UNIXTIME`.
**Example:**
Query:
```sql
SELECT fromUnixTimestamp(423543535);
```
@ -1534,7 +1499,6 @@ Similar to fromUnixTimestamp, except that it formats time in Joda style instead
**Example:**
Query:
``` sql
SELECT fromUnixTimestampInJodaSyntax(1669804872, 'yyyy-MM-dd HH:mm:ss', 'UTC');
```
@ -1568,8 +1532,6 @@ Type: [Int32](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT toModifiedJulianDay('2020-01-01');
```
@ -1604,8 +1566,6 @@ Type: [Nullable(Int32)](../../sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT toModifiedJulianDayOrNull('2020-01-01');
```
@ -1640,8 +1600,6 @@ Type: [String](../../sql-reference/data-types/string.md)
**Example**
Query:
``` sql
SELECT fromModifiedJulianDay(58849);
```
@ -1676,8 +1634,6 @@ Type: [Nullable(String)](../../sql-reference/data-types/string.md)
**Example**
Query:
``` sql
SELECT fromModifiedJulianDayOrNull(58849);
```

5
docs/en/sql-reference/functions/distance-functions.md
View File

@ -1,7 +1,10 @@
---
slug: /en/sql-reference/functions/distance-functions
sidebar_position: 55
sidebar_label: Distance
---
# Distance functions
# Distance Functions
## L1Norm

2
docs/en/sql-reference/functions/encoding-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/encoding-functions
sidebar_position: 52
sidebar_position: 65
sidebar_label: Encoding
---

3
docs/en/sql-reference/functions/encryption-functions.md
View File

@ -1,8 +1,7 @@
---
slug: /en/sql-reference/functions/encryption-functions
sidebar_position: 67
sidebar_position: 70
sidebar_label: Encryption
title: "Encryption functions"
---
These functions implement encryption and decryption of data with AES (Advanced Encryption Standard) algorithm.

2
docs/en/sql-reference/functions/ext-dict-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/ext-dict-functions
sidebar_position: 58
sidebar_position: 50
sidebar_label: Dictionaries
---

16
docs/en/sql-reference/functions/files.md
View File

@ -1,13 +1,14 @@
---
slug: /en/sql-reference/functions/files
sidebar_position: 43
sidebar_position: 75
sidebar_label: Files
title: "Functions for Working with Files"
---
## file
Reads file as a String. The file content is not parsed, so any information is read as one string and placed into the specified column.
Reads file as string and loads the data into the specified column. The actual file content is not interpreted.
Also see table function [file](../table-functions/file.md).
**Syntax**
@ -17,8 +18,8 @@ file(path[, default])
**Arguments**
- `path` — The relative path to the file from [user_files_path](../../operations/server-configuration-parameters/settings.md#server_configuration_parameters-user_files_path). Path to file support following wildcards: `*`, `?`, `{abc,def}` and `{N..M}` where `N`, `M` — numbers, `'abc', 'def'` — strings.
- `default` — The value that will be returned in the case when a file does not exist or cannot be accessed. Data types supported: [String](../../sql-reference/data-types/string.md) and [NULL](../../sql-reference/syntax.md#null-literal).
- `path` — The path of the file relative to [user_files_path](../../operations/server-configuration-parameters/settings.md#server_configuration_parameters-user_files_path). Supports the following wildcards: `*`, `?`, `{abc,def}` and `{N..M}` where `N`, `M` are numbers and `'abc', 'def'` are strings.
- `default` — The value that will be returned in the case the file does not exist or cannot be accessed. Supported data types: [String](../../sql-reference/data-types/string.md) and [NULL](../../sql-reference/syntax.md#null-literal).
**Example**
@ -29,8 +30,3 @@ Query:
``` sql
INSERT INTO table SELECT file('a.txt'), file('b.txt');
```
**See Also**
- [user_files_path](../../operations/server-configuration-parameters/settings.md#server_configuration_parameters-user_files_path)
- [file](../table-functions/file.md)

2
docs/en/sql-reference/functions/functions-for-nulls.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/functions-for-nulls
sidebar_position: 63
sidebar_position: 135
sidebar_label: Nullable
---

2
docs/en/sql-reference/functions/hash-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/hash-functions
sidebar_position: 50
sidebar_position: 85
sidebar_label: Hash
---

2
docs/en/sql-reference/functions/in-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/in-functions
sidebar_position: 60
sidebar_position: 90
sidebar_label: IN Operator
---

6
docs/en/sql-reference/functions/index.md
View File

@ -1,10 +1,10 @@
---
slug: /en/sql-reference/functions/
sidebar_position: 32
sidebar_label: Functions
sidebar_position: 1
sidebar_label: Overview
---
# Functions
# Regular Functions
There are at least\* two types of functions - regular functions (they are just called “functions”) and aggregate functions. These are completely different concepts. Regular functions work as if they are applied to each row separately (for each row, the result of the function does not depend on the other rows). Aggregate functions accumulate a set of values from various rows (i.e. they depend on the entire set of rows).

2
docs/en/sql-reference/functions/introspection.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/introspection
sidebar_position: 65
sidebar_position: 100
sidebar_label: Introspection
---

2
docs/en/sql-reference/functions/ip-address-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/ip-address-functions
sidebar_position: 55
sidebar_position: 95
sidebar_label: IP Addresses
---

3
docs/en/sql-reference/functions/json-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/json-functions
sidebar_position: 56
sidebar_position: 105
sidebar_label: JSON
---
@ -206,6 +206,7 @@ Examples:
``` sql
SELECT JSONExtract('{"a": "hello", "b": [-100, 200.0, 300]}', 'Tuple(String, Array(Float64))') = ('hello',[-100,200,300])
SELECT JSONExtract('{"a": "hello", "b": [-100, 200.0, 300]}', 'Tuple(b Array(Float64), a String)') = ([-100,200,300],'hello')
SELECT JSONExtract('{"a": "hello", "b": "world"}', 'Map(String, String)') = map('a', 'hello', 'b', 'world');
SELECT JSONExtract('{"a": "hello", "b": [-100, 200.0, 300]}', 'b', 'Array(Nullable(Int8))') = [-100, NULL, NULL]
SELECT JSONExtract('{"a": "hello", "b": [-100, 200.0, 300]}', 'b', 4, 'Nullable(Int64)') = NULL
SELECT JSONExtract('{"passed": true}', 'passed', 'UInt8') = 1

54
docs/en/sql-reference/functions/logical-functions.md
View File

@ -1,18 +1,18 @@
---
slug: /en/sql-reference/functions/logical-functions
sidebar_position: 37
sidebar_position: 110
sidebar_label: Logical
---
# Logical Functions
Performs logical operations on arguments of any numeric types, but returns a [UInt8](../../sql-reference/data-types/int-uint.md) number equal to 0, 1 or `NULL` in some cases.
Below functions perform logical operations on arguments of arbitrary numeric types. They return either 0 or 1 as [UInt8](../../sql-reference/data-types/int-uint.md) or in some cases `NULL`.
Zero as an argument is considered `false`, while any non-zero value is considered `true`.
Zero as an argument is considered `false`, non-zero values are considered `true`.
## and
Calculates the result of the logical conjunction between two or more values. Corresponds to [Logical AND Operator](../../sql-reference/operators/index.md#logical-and-operator).
Calculates the logical conjunction between two or more values.
**Syntax**
@ -20,7 +20,9 @@ Calculates the result of the logical conjunction between two or more values. Cor
and(val1, val2...)
```
You can use the [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) setting to calculate the `and` function according to a short scheme. If this setting is enabled, `vali` is evaluated only on rows where `(val1 AND val2 AND ... AND val{i-1})` is true. For example, an exception about division by zero is not thrown when executing the query `SELECT and(number = 2, intDiv(1, number)) FROM numbers(10)`.
Setting [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) controls whether short-circuit evaluation is used. If enabled, `val_i` is evaluated only if `(val_1 AND val_2 AND ... AND val_{i-1})` is `true`. For example, with short-circuit evaluation, no division-by-zero exception is thrown when executing the query `SELECT and(number = 2, intDiv(1, number)) FROM numbers(5)`.
Alias: The [AND Operator](../../sql-reference/operators/index.md#logical-and-operator).
**Arguments**
@ -28,16 +30,14 @@ You can use the [short_circuit_function_evaluation](../../operations/settings/se
**Returned value**
- `0`, if there is at least one zero value argument.
- `NULL`, if there are no zero values arguments and there is at least one `NULL` argument.
- `0`, if there at least one argument evaluates to `false`,
- `NULL`, if no argumetn evaluates to `false` and at least one argument is `NULL`,
- `1`, otherwise.
Type: [UInt8](../../sql-reference/data-types/int-uint.md) or [Nullable](../../sql-reference/data-types/nullable.md)([UInt8](../../sql-reference/data-types/int-uint.md)).
**Example**
Query:
``` sql
SELECT and(0, 1, -2);
```
@ -66,7 +66,7 @@ Result:
## or
Calculates the result of the logical disjunction between two or more values. Corresponds to [Logical OR Operator](../../sql-reference/operators/index.md#logical-or-operator).
Calculates the logical disjunction between two or more values.
**Syntax**
@ -74,7 +74,9 @@ Calculates the result of the logical disjunction between two or more values. Cor
or(val1, val2...)
```
You can use the [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) setting to calculate the `or` function according to a short scheme. If this setting is enabled, `vali` is evaluated only on rows where `((NOT val1) AND (NOT val2) AND ... AND (NOT val{i-1}))` is true. For example, an exception about division by zero is not thrown when executing the query `SELECT or(number = 0, intDiv(1, number) != 0) FROM numbers(10)`.
Setting [short_circuit_function_evaluation](../../operations/settings/settings.md#short-circuit-function-evaluation) controls whether short-circuit evaluation is used. If enabled, `val_i` is evaluated only if `((NOT val_1) AND (NOT val_2) AND ... AND (NOT val_{i-1}))` is `true`. For example, with short-circuit evaluation, no division-by-zero exception is thrown when executing the query `SELECT or(number = 0, intDiv(1, number) != 0) FROM numbers(5)`.
Alias: The [OR Operator](../../sql-reference/operators/index.md#logical-or-operator).
**Arguments**
@ -82,16 +84,14 @@ You can use the [short_circuit_function_evaluation](../../operations/settings/se
**Returned value**
- `1`, if there is at least one non-zero value.
- `0`, if there are only zero values.
- `NULL`, if there are only zero values and `NULL`.
- `1`, if at least one argument evalutes to `true`,
- `0`, if all arguments evaluate to `false`,
- `NULL`, if all arguments evaluate to `false` and at least one argument is `NULL`.
Type: [UInt8](../../sql-reference/data-types/int-uint.md) or [Nullable](../../sql-reference/data-types/nullable.md)([UInt8](../../sql-reference/data-types/int-uint.md)).
**Example**
Query:
``` sql
SELECT or(1, 0, 0, 2, NULL);
```
@ -120,7 +120,7 @@ Result:
## not
Calculates the result of the logical negation of the value. Corresponds to [Logical Negation Operator](../../sql-reference/operators/index.md#logical-negation-operator).
Calculates logical negation of a value.
**Syntax**
@ -128,22 +128,22 @@ Calculates the result of the logical negation of the value. Corresponds to [Logi
not(val);
```
Alias: The [Negation Operator](../../sql-reference/operators/index.md#logical-negation-operator).
**Arguments**
- `val` — The value. [Int](../../sql-reference/data-types/int-uint.md), [UInt](../../sql-reference/data-types/int-uint.md), [Float](../../sql-reference/data-types/float.md) or [Nullable](../../sql-reference/data-types/nullable.md).
**Returned value**
- `1`, if the `val` is `0`.
- `0`, if the `val` is a non-zero value.
- `NULL`, if the `val` is a `NULL` value.
- `1`, if `val` evaluates to `false`,
- `0`, if `val` evaluates to `true`,
- `NULL`, if `val` is `NULL`.
Type: [UInt8](../../sql-reference/data-types/int-uint.md) or [Nullable](../../sql-reference/data-types/nullable.md)([UInt8](../../sql-reference/data-types/int-uint.md)).
**Example**
Query:
``` sql
SELECT NOT(1);
```
@ -158,7 +158,7 @@ Result:
## xor
Calculates the result of the logical exclusive disjunction between two or more values. For more than two values the function works as if it calculates `XOR` of the first two values and then uses the result with the next value to calculate `XOR` and so on.
Calculates the logical exclusive disjunction between two or more values. For more than two values the function first xor-s the first two values, then xor-s the result with the third value etc.
**Syntax**
@ -172,16 +172,14 @@ xor(val1, val2...)
**Returned value**
- `1`, for two values: if one of the values is zero and other is not.
- `0`, for two values: if both values are zero or non-zero at the same time.
- `NULL`, if there is at least one `NULL` value.
- `1`, for two values: if one of the values evaluates to `false` and other does not,
- `0`, for two values: if both values evalute to `false` or to both `true`,
- `NULL`, if at least one of the inputs is `NULL`
Type: [UInt8](../../sql-reference/data-types/int-uint.md) or [Nullable](../../sql-reference/data-types/nullable.md)([UInt8](../../sql-reference/data-types/int-uint.md)).
**Example**
Query:
``` sql
SELECT xor(0, 1, 1);
```

2
docs/en/sql-reference/functions/machine-learning-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/machine-learning-functions
sidebar_position: 64
sidebar_position: 115
sidebar_label: Machine Learning
---

358
docs/en/sql-reference/functions/math-functions.md
View File

@ -1,120 +1,260 @@
---
slug: /en/sql-reference/functions/math-functions
sidebar_position: 44
sidebar_position: 125
sidebar_label: Mathematical
---
# Mathematical Functions
All the functions return a Float64 number. The accuracy of the result is close to the maximum precision possible, but the result might not coincide with the machine representable number nearest to the corresponding real number.
All the functions return a Float64 number. Results are generally as close to the actual result as possible, but in some cases less precise than the machine-representable number.
## e()
## e
Returns a Float64 number that is close to the number e.
Returns e.
## pi()
**Syntax**
Returns a Float64 number that is close to the number π.
```sql
e()
```
## exp(x)
## pi
Accepts a numeric argument and returns a Float64 number close to the exponent of the argument.
Returns π.
## log(x), ln(x)
**Syntax**
Accepts a numeric argument and returns a Float64 number close to the natural logarithm of the argument.
```sql
pi()
```
## exp2(x)
## exp
Accepts a numeric argument and returns a Float64 number close to 2 to the power of x.
Returns e to the power of the given argument.
## log2(x)
**Syntax**
Accepts a numeric argument and returns a Float64 number close to the binary logarithm of the argument.
```sql
exp(x)
```
## exp10(x)
## log
Accepts a numeric argument and returns a Float64 number close to 10 to the power of x.
Returns the natural logarithm of the argument.
## log10(x)
**Syntax**
Accepts a numeric argument and returns a Float64 number close to the decimal logarithm of the argument.
```sql
log(x)
```
## sqrt(x)
Alias: `ln(x)`
Accepts a numeric argument and returns a Float64 number close to the square root of the argument.
## exp2
## cbrt(x)
Returns 2 to the power of the given argumetn
Accepts a numeric argument and returns a Float64 number close to the cubic root of the argument.
**Syntax**
## erf(x)
```sql
exp2(x)
```
If ‘x’ is non-negative, then `erf(x / σ√2)` is the probability that a random variable having a normal distribution with standard deviation ‘σ’ takes the value that is separated from the expected value by more than ‘x’.
## intExp2
Example (three sigma rule):
Like `exp` but returns a UInt64.
**Syntax**
```sql
intExp2(x)
```
## log2
Returns the binary logarithm of the argument.
**Syntax**
```sql
log2(x)
```
## exp10
Returns 10 to the power of the given argumetn
**Syntax**
```sql
exp10(x)
```
## intExp10
Like `exp10` but returns a UInt64.
**Syntax**
```sql
intExp10(x)
```
## log10
Returns the decimal logarithm of the argument.
**Syntax**
```sql
log10(x)
```
## sqrt
Returns the square root of the argument.
```sql
sqrt(x)
```
## cbrt
Returns the cubic root of the argument.
```sql
cbrt(x)
```
## erf
If `x` is non-negative, then `erf(x / σ√2)` is the probability that a random variable having a normal distribution with standard deviation `σ` takes the value that is separated from the expected value by more than `x`.
**Syntax**
```sql
erf(x)
```
**Example**
(three sigma rule)
``` sql
SELECT erf(3 / sqrt(2));
```
``` text
```result
┌─erf(divide(3, sqrt(2)))─┐
│ 0.9973002039367398 │
└─────────────────────────┘
```
## erfc(x)
## erfc
Accepts a numeric argument and returns a Float64 number close to 1 - erf(x), but without loss of precision for large ‘x’ values.
Returns a number close to `1 - erf(x)` without loss of precision for large ‘x’ values.
## lgamma(x)
**Syntax**
The logarithm of the gamma function.
```sql
erfc(x)
```
## tgamma(x)
## lgamma
Gamma function.
Returns the logarithm of the gamma function.
## sin(x)
**Syntax**
The sine.
```sql
lgamma(x)
```
## cos(x)
## tgamma
The cosine.
Returns the gamma function.
## tan(x)
**Syntax**
The tangent.
```sql
gamma(x)
```
## asin(x)
## sin
The arc sine.
Returns the sine of the argument
## acos(x)
**Syntax**
The arc cosine.
```sql
sin(x)
```
## atan(x)
## cos
The arc tangent.
Returns the cosine of the argument.
## pow(x, y), power(x, y)
**Syntax**
Takes two numeric arguments x and y. Returns a Float64 number close to x to the power of y.
```sql
cos(x)
```
## intExp2
## tan
Accepts a numeric argument and returns a UInt64 number close to 2 to the power of x.
Returns the tangent of the argument.
## intExp10
**Syntax**
Accepts a numeric argument and returns a UInt64 number close to 10 to the power of x.
```sql
tan(x)
```
## cosh(x)
## asin
[Hyperbolic cosine](https://in.mathworks.com/help/matlab/ref/cosh.html).
Returns the arc sine of the argument.
**Syntax**
```sql
asin(x)
```
## acos
Returns the arc cosine of the argument.
**Syntax**
```sql
acos(x)
```
## atan
Returns the arc tangent of the argument.
**Syntax**
```sql
atan(x)
```
## pow
Returns `x` to the power of `y`.
**Syntax**
```sql
pow(x, y)
```
Alias: `power(x, y)`
## cosh
Returns the [hyperbolic cosine](https://in.mathworks.com/help/matlab/ref/cosh.html) of the argument.
**Syntax**
@ -134,23 +274,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT cosh(0);
```
Result:
``` text
```result
┌─cosh(0)──┐
│ 1 │
└──────────┘
```
## acosh(x)
## acosh
[Inverse hyperbolic cosine](https://www.mathworks.com/help/matlab/ref/acosh.html).
Returns the [inverse hyperbolic cosine](https://www.mathworks.com/help/matlab/ref/acosh.html).
**Syntax**
@ -170,27 +308,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT acosh(1);
```
Result:
``` text
```result
┌─acosh(1)─┐
│ 0 │
└──────────┘
```
**See Also**
## sinh
- [cosh(x)](../../sql-reference/functions/math-functions.md#coshx)
## sinh(x)
[Hyperbolic sine](https://www.mathworks.com/help/matlab/ref/sinh.html).
Returns the [hyperbolic sine](https://www.mathworks.com/help/matlab/ref/sinh.html).
**Syntax**
@ -210,23 +342,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT sinh(0);
```
Result:
``` text
```result
┌─sinh(0)──┐
│ 0 │
└──────────┘
```
## asinh(x)
## asinh
[Inverse hyperbolic sine](https://www.mathworks.com/help/matlab/ref/asinh.html).
Returns the [inverse hyperbolic sine](https://www.mathworks.com/help/matlab/ref/asinh.html).
**Syntax**
@ -246,27 +376,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT asinh(0);
```
Result:
``` text
```result
┌─asinh(0)─┐
│ 0 │
└──────────┘
```
**See Also**
## atanh
- [sinh(x)](../../sql-reference/functions/math-functions.md#sinhx)
## atanh(x)
[Inverse hyperbolic tangent](https://www.mathworks.com/help/matlab/ref/atanh.html).
Returns the [inverse hyperbolic tangent](https://www.mathworks.com/help/matlab/ref/atanh.html).
**Syntax**
@ -286,23 +410,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT atanh(0);
```
Result:
``` text
```result
┌─atanh(0)─┐
│ 0 │
└──────────┘
```
## atan2(y, x)
## atan2
The [function](https://en.wikipedia.org/wiki/Atan2) calculates the angle in the Euclidean plane, given in radians, between the positive x axis and the ray to the point `(x, y) ≠ (0, 0)`.
Returns the [atan2](https://en.wikipedia.org/wiki/Atan2) as the angle in the Euclidean plane, given in radians, between the positive x axis and the ray to the point `(x, y) ≠ (0, 0)`.
**Syntax**
@ -323,23 +445,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT atan2(1, 1);
```
Result:
``` text
```result
┌────────atan2(1, 1)─┐
│ 0.7853981633974483 │
└────────────────────┘
```
## hypot(x, y)
## hypot
Calculates the length of the hypotenuse of a right-angle triangle. The [function](https://en.wikipedia.org/wiki/Hypot) avoids problems that occur when squaring very large or very small numbers.
Returns the length of the hypotenuse of a right-angle triangle. [Hypot](https://en.wikipedia.org/wiki/Hypot) avoids problems that occur when squaring very large or very small numbers.
**Syntax**
@ -360,23 +480,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT hypot(1, 1);
```
Result:
``` text
```result
┌────────hypot(1, 1)─┐
│ 1.4142135623730951 │
└────────────────────┘
```
## log1p(x)
## log1p
Calculates `log(1+x)`. The [function](https://en.wikipedia.org/wiki/Natural_logarithm#lnp1) `log1p(x)` is more accurate than `log(1+x)` for small values of x.
Calculates `log(1+x)`. The [calculation](https://en.wikipedia.org/wiki/Natural_logarithm#lnp1) `log1p(x)` is more accurate than `log(1+x)` for small values of x.
**Syntax**
@ -396,25 +514,19 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT log1p(0);
```
Result:
``` text
```result
┌─log1p(0)─┐
│ 0 │
└──────────┘
```
**See Also**
- [log(x)](../../sql-reference/functions/math-functions.md#logx-lnx)
## sign(x)
## sign
Returns the sign of a real number.
@ -444,7 +556,7 @@ SELECT sign(0);
Result:
``` text
```result
┌─sign(0)─┐
│ 0 │
└─────────┘
@ -458,7 +570,7 @@ SELECT sign(1);
Result:
``` text
```result
┌─sign(1)─┐
│ 1 │
└─────────┘
@ -472,15 +584,15 @@ SELECT sign(-1);
Result:
``` text
```result
┌─sign(-1)─┐
│ -1 │
└──────────┘
```
## degrees(x)
## degrees
Converts the input value in radians to degrees.
Converts radians to degrees.
**Syntax**
@ -500,23 +612,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT degrees(3.141592653589793);
```
Result:
``` text
```result
┌─degrees(3.141592653589793)─┐
│ 180 │
└────────────────────────────┘
```
## radians(x)
## radians
Converts the input value in degrees to radians.
Converts degrees to radians.
**Syntax**
@ -536,24 +646,21 @@ Type: [Float64](../../sql-reference/data-types/float.md#float32-float64).
**Example**
Query:
``` sql
SELECT radians(180);
```
Result:
``` text
```result
┌──────radians(180)─┐
│ 3.141592653589793 │
└───────────────────┘
```
## factorial
## factorial(n)
Computes the factorial of an integer value. It works with any native integer type including UInt(8|16|32|64) and Int(8|16|32|64). The return type is UInt64.
Computes the factorial of an integer value. Works with any native integer type including UInt(8|16|32|64) and Int(8|16|32|64). The return type is UInt64.
The factorial of 0 is 1. Likewise, the factorial() function returns 1 for any negative value. The maximum positive value for the input argument is 20, a value of 21 or greater will cause exception throw.
@ -566,21 +673,19 @@ factorial(n)
**Example**
Query:
``` sql
SELECT factorial(10);
```
Result:
``` text
```result
┌─factorial(10)─┐
│ 3628800 │
└───────────────┘
```
## width_bucket(operand, low, high, count)
## width_bucket
Returns the number of the bucket in which `operand` falls in a histogram having `count` equal-width buckets spanning the range `low` to `high`. Returns `0` if `operand < low`, and returns `count+1` if `operand >= high`.
@ -591,20 +696,17 @@ Returns the number of the bucket in which `operand` falls in a histogram having
```sql
widthBucket(operand, low, high, count)
```
There is also a case insensitive alias called `WIDTH_BUCKET` to provide compatibility with other databases.
Alias: `WIDTH_BUCKET`
**Example**
Query:
``` sql
SELECT widthBucket(10.15, -8.6, 23, 18);
```
Result:
``` text
```result
┌─widthBucket(10.15, -8.6, 23, 18)─┐
│ 11 │
└──────────────────────────────────┘

5
docs/en/sql-reference/functions/nlp-functions.md
View File

@ -1,8 +1,7 @@
---
slug: /en/sql-reference/functions/nlp-functions
sidebar_position: 67
sidebar_label: NLP
title: "[experimental] Natural Language Processing functions"
sidebar_position: 130
sidebar_label: NLP (experimental)
---
:::note

2
docs/en/sql-reference/functions/other-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/other-functions
sidebar_position: 67
sidebar_position: 140
sidebar_label: Other
---

233
docs/en/sql-reference/functions/random-functions.md
View File

@ -1,58 +1,44 @@
---
slug: /en/sql-reference/functions/random-functions
sidebar_position: 51
sidebar_label: Pseudo-Random Numbers
sidebar_position: 145
sidebar_label: Random Numbers
---
# Functions for Generating Pseudo-Random Numbers
# Functions for Generating Random Numbers
All the functions accept zero arguments or one argument. If an argument is passed, it can be any type, and its value is not used for anything. The only purpose of this argument is to prevent common subexpression elimination, so that two different instances of the same function return different columns with different random numbers.
All functions in this section accept zero or one arguments. The only use of the argument (if provided) is to prevent prevent [common subexpression
elimination](../../sql-reference/functions/index.md#common-subexpression-elimination) such that two different execution of the same random
function in a query return different random values.
Related content
- Blog: [Generating random data in ClickHouse](https://clickhouse.com/blog/generating-random-test-distribution-data-for-clickhouse)
:::note
Non-cryptographic generators of pseudo-random numbers are used.
The random numbers are generated by non-cryptographic algorithms.
:::
## rand, rand32
Returns a pseudo-random UInt32 number, evenly distributed among all UInt32-type numbers.
Returns a random UInt32 number, evenly distributed accross the range of all possible UInt32 numbers.
Uses a linear congruential generator.
## rand64
Returns a pseudo-random UInt64 number, evenly distributed among all UInt64-type numbers.
Returns a random UInt64 number, evenly distributed accross the range of all possible UInt64 numbers.
Uses a linear congruential generator.
## randCanonical
The function generates pseudo random results with independent and identically distributed uniformly distributed values in [0, 1).
Non-deterministic. Return type is Float64.
Returns a Float64 value, evenly distributed in [0, 1).
## randConstant
Produces a constant column with a random value.
**Syntax**
``` sql
randConstant([x])
```
**Arguments**
- `x` — [Expression](../../sql-reference/syntax.md#syntax-expressions) resulting in any of the [supported data types](../../sql-reference/data-types/index.md#data_types). The resulting value is discarded, but the expression itself if used for bypassing [common subexpression elimination](../../sql-reference/functions/index.md#common-subexpression-elimination) if the function is called multiple times in one query. Optional parameter.
**Returned value**
- Pseudo-random number.
Type: [UInt32](../../sql-reference/data-types/int-uint.md).
Like `rand` but produces a constant column with a random value.
**Example**
Query:
``` sql
SELECT rand(), rand(1), rand(number), randConstant(), randConstant(1), randConstant(number)
FROM numbers(3)
@ -60,7 +46,7 @@ FROM numbers(3)
Result:
``` text
``` result
┌─────rand()─┬────rand(1)─┬─rand(number)─┬─randConstant()─┬─randConstant(1)─┬─randConstant(number)─┐
│ 3047369878 │ 4132449925 │ 4044508545 │ 2740811946 │ 4229401477 │ 1924032898 │
│ 2938880146 │ 1267722397 │ 4154983056 │ 2740811946 │ 4229401477 │ 1924032898 │
@ -68,17 +54,11 @@ Result:
└────────────┴────────────┴──────────────┴────────────────┴─────────────────┴──────────────────────┘
```
# Functions for Generating Random Numbers based on Distributions
:::note
These functions are available starting from 22.10.
:::
# Functions for Generating Random Numbers based on a Distribution
## randUniform
Return random number based on [continuous uniform distribution](https://en.wikipedia.org/wiki/Continuous_uniform_distribution) in a specified range from `min` to `max`.
Returns a Float64 drawn uniformly from the interval between `min` and `max` ([continuous uniform distribution](https://en.wikipedia.org/wiki/Continuous_uniform_distribution)).
**Syntax**
@ -93,21 +73,19 @@ randUniform(min, max)
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randUniform(5.5, 10) FROM numbers(5)
```
Result:
``` text
``` result
┌─randUniform(5.5, 10)─┐
│ 8.094978491443102 │
│ 7.3181248914450885 │
@ -117,40 +95,36 @@ Result:
└──────────────────────┘
```
## randNormal
Return random number based on [normal distribution](https://en.wikipedia.org/wiki/Normal_distribution).
Returns a Float64 drawn from a [normal distribution](https://en.wikipedia.org/wiki/Normal_distribution).
**Syntax**
``` sql
randNormal(meam, variance)
randNormal(mean, variance)
```
**Arguments**
- `meam` - `Float64` mean value of distribution,
- `mean` - `Float64` - mean value of distribution,
- `variance` - `Float64` - [variance](https://en.wikipedia.org/wiki/Variance).
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randNormal(10, 2) FROM numbers(5)
```
Result:
``` text
``` result
┌──randNormal(10, 2)─┐
│ 13.389228911709653 │
│ 8.622949707401295 │
@ -160,40 +134,36 @@ Result:
└────────────────────┘
```
## randLogNormal
Return random number based on [log-normal distribution](https://en.wikipedia.org/wiki/Log-normal_distribution).
Returns a Float64 drawn from a [log-normal distribution](https://en.wikipedia.org/wiki/Log-normal_distribution).
**Syntax**
``` sql
randLogNormal(meam, variance)
randLogNormal(mean, variance)
```
**Arguments**
- `meam` - `Float64` mean value of distribution,
- `mean` - `Float64` - mean value of distribution,
- `variance` - `Float64` - [variance](https://en.wikipedia.org/wiki/Variance).
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randLogNormal(100, 5) FROM numbers(5)
```
Result:
``` text
``` result
┌─randLogNormal(100, 5)─┐
│ 1.295699673937363e48 │
│ 9.719869109186684e39 │
@ -203,11 +173,9 @@ Result:
└───────────────────────┘
```
## randBinomial
Return random number based on [binomial distribution](https://en.wikipedia.org/wiki/Binomial_distribution).
Returns a UInt64 drawn from a [binomial distribution](https://en.wikipedia.org/wiki/Binomial_distribution).
**Syntax**
@ -217,26 +185,24 @@ randBinomial(experiments, probability)
**Arguments**
- `experiments` - `UInt64` number of experiments,
- `experiments` - `UInt64` - number of experiments,
- `probability` - `Float64` - probability of success in each experiment (values in `0...1` range only).
**Returned value**
- Pseudo-random number.
- Random number.
Type: [UInt64](/docs/en/sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT randBinomial(100, .75) FROM numbers(5)
```
Result:
``` text
``` result
┌─randBinomial(100, 0.75)─┐
│ 74 │
│ 78 │
@ -246,11 +212,9 @@ Result:
└─────────────────────────┘
```
## randNegativeBinomial
Return random number based on [negative binomial distribution](https://en.wikipedia.org/wiki/Negative_binomial_distribution).
Returns a UInt64 drawn from a [negative binomial distribution](https://en.wikipedia.org/wiki/Negative_binomial_distribution).
**Syntax**
@ -260,26 +224,24 @@ randNegativeBinomial(experiments, probability)
**Arguments**
- `experiments` - `UInt64` number of experiments,
- `experiments` - `UInt64` - number of experiments,
- `probability` - `Float64` - probability of failure in each experiment (values in `0...1` range only).
**Returned value**
- Pseudo-random number.
- Random number.
Type: [UInt64](/docs/en/sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT randNegativeBinomial(100, .75) FROM numbers(5)
```
Result:
``` text
``` result
┌─randNegativeBinomial(100, 0.75)─┐
│ 33 │
│ 32 │
@ -289,11 +251,9 @@ Result:
└─────────────────────────────────┘
```
## randPoisson
Return random number based on [Poisson distribution](https://en.wikipedia.org/wiki/Poisson_distribution).
Returns a UInt64 drawn from a [Poisson distribution](https://en.wikipedia.org/wiki/Poisson_distribution).
**Syntax**
@ -303,25 +263,23 @@ randPoisson(n)
**Arguments**
- `n` - `UInt64` mean number of occurrences.
- `n` - `UInt64` - mean number of occurrences.
**Returned value**
- Pseudo-random number.
- Random number.
Type: [UInt64](/docs/en/sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT randPoisson(10) FROM numbers(5)
```
Result:
``` text
``` result
┌─randPoisson(10)─┐
│ 8 │
│ 8 │
@ -331,11 +289,9 @@ Result:
└─────────────────┘
```
## randBernoulli
Return random number based on [Bernoulli distribution](https://en.wikipedia.org/wiki/Bernoulli_distribution).
Returns a UInt64 drawn from a [Bernoulli distribution](https://en.wikipedia.org/wiki/Bernoulli_distribution).
**Syntax**
@ -349,21 +305,19 @@ randBernoulli(probability)
**Returned value**
- Pseudo-random number.
- Random number.
Type: [UInt64](/docs/en/sql-reference/data-types/int-uint.md).
**Example**
Query:
``` sql
SELECT randBernoulli(.75) FROM numbers(5)
```
Result:
``` text
``` result
┌─randBernoulli(0.75)─┐
│ 1 │
│ 1 │
@ -373,11 +327,9 @@ Result:
└─────────────────────┘
```
## randExponential
Return random number based on [exponential distribution](https://en.wikipedia.org/wiki/Exponential_distribution).
Returns a Float64 drawn from a [exponential distribution](https://en.wikipedia.org/wiki/Exponential_distribution).
**Syntax**
@ -387,25 +339,23 @@ randExponential(lambda)
**Arguments**
- `lambda` - `Float64` lambda value.
- `lambda` - `Float64` - lambda value.
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randExponential(1/10) FROM numbers(5)
```
Result:
``` text
``` result
┌─randExponential(divide(1, 10))─┐
│ 44.71628934340778 │
│ 4.211013337903262 │
@ -415,11 +365,9 @@ Result:
└────────────────────────────────┘
```
## randChiSquared
Return random number based on [Chi-square distribution](https://en.wikipedia.org/wiki/Chi-squared_distribution) - a distribution of a sum of the squares of k independent standard normal random variables.
Returns a Float64 drawn from a [Chi-square distribution](https://en.wikipedia.org/wiki/Chi-squared_distribution) - a distribution of a sum of the squares of k independent standard normal random variables.
**Syntax**
@ -429,25 +377,23 @@ randChiSquared(degree_of_freedom)
**Arguments**
- `degree_of_freedom` - `Float64` degree of freedom.
- `degree_of_freedom` - `Float64` - degree of freedom.
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randChiSquared(10) FROM numbers(5)
```
Result:
``` text
``` result
┌─randChiSquared(10)─┐
│ 10.015463656521543 │
│ 9.621799919882768 │
@ -457,11 +403,9 @@ Result:
└────────────────────┘
```
## randStudentT
Return random number based on [Student's t-distribution](https://en.wikipedia.org/wiki/Student%27s_t-distribution).
Returns a Float64 drawn from a [Student's t-distribution](https://en.wikipedia.org/wiki/Student%27s_t-distribution).
**Syntax**
@ -471,25 +415,23 @@ randStudentT(degree_of_freedom)
**Arguments**
- `degree_of_freedom` - `Float64` degree of freedom.
- `degree_of_freedom` - `Float64` - degree of freedom.
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randStudentT(10) FROM numbers(5)
```
Result:
``` text
``` result
┌─────randStudentT(10)─┐
│ 1.2217309938538725 │
│ 1.7941971681200541 │
@ -499,11 +441,9 @@ Result:
└──────────────────────┘
```
## randFisherF
Return random number based on [F-distribution](https://en.wikipedia.org/wiki/F-distribution).
Returns a Float64 drawn from a [F-distribution](https://en.wikipedia.org/wiki/F-distribution).
**Syntax**
@ -513,26 +453,24 @@ randFisherF(d1, d2)
**Arguments**
- `d1` - `Float64` d1 degree of freedom in `X = (S1 / d1) / (S2 / d2)`,
- `d2` - `Float64` d2 degree of freedom in `X = (S1 / d1) / (S2 / d2)`,
- `d1` - `Float64` - d1 degree of freedom in `X = (S1 / d1) / (S2 / d2)`,
- `d2` - `Float64` - d2 degree of freedom in `X = (S1 / d1) / (S2 / d2)`,
**Returned value**
- Pseudo-random number.
- Random number.
Type: [Float64](/docs/en/sql-reference/data-types/float.md).
**Example**
Query:
``` sql
SELECT randFisherF(10, 3) FROM numbers(5)
```
Result:
``` text
``` result
┌──randFisherF(10, 3)─┐
│ 7.286287504216609 │
│ 0.26590779413050386 │
@ -542,35 +480,61 @@ Result:
└─────────────────────┘
```
# Random Functions for Working with Strings
# Functions for Generating Random Strings
## randomString
Returns a random String of specified `length`. Not all characters may be printable.
**Syntax**
```sql
randomString(length)
```
## randomFixedString
Like `randomString` but returns a FixedString.
## randomPrintableASCII
Returns a random String of specified `length`. All characters are printable.
**Syntax**
```sql
randomPrintableASCII(length)
```
## randomStringUTF8
Returns a random String containing `length` many UTF8 codepoints. Not all characters may be printable
**Syntax**
```sql
randomStringUTF8(length)
```
## fuzzBits
**Syntax**
``` sql
fuzzBits([s], [prob])
```
Inverts the bits of String or FixedString `s`, each with probability `prob`.
Inverts bits of `s`, each with probability `prob`.
**Syntax**
``` sql
fuzzBits(s, prob)
```
**Arguments**
- `s` - `String` or `FixedString`
- `prob` - constant `Float32/64`
**Returned value**
Fuzzed string with same as s type.
Fuzzed string with same type as `s`.
**Example**
@ -581,13 +545,10 @@ FROM numbers(3)
Result:
``` text
``` result
┌─fuzzBits(materialize('abacaba'), 0.1)─┐
│ abaaaja │
│ a*cjab+ │
│ aeca2A │
└───────────────────────────────────────┘
```
## Related content
- Blog: [Generating random data in ClickHouse](https://clickhouse.com/blog/generating-random-test-distribution-data-for-clickhouse)

2
docs/en/sql-reference/functions/rounding-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/rounding-functions
sidebar_position: 45
sidebar_position: 155
sidebar_label: Rounding
---

56
docs/en/sql-reference/functions/splitting-merging-functions.md
View File

@ -1,14 +1,14 @@
---
slug: /en/sql-reference/functions/splitting-merging-functions
sidebar_position: 47
sidebar_label: Splitting and Merging Strings and Arrays
sidebar_position: 165
sidebar_label: Splitting Strings
---
# Functions for Splitting and Merging Strings and Arrays
# Functions for Splitting Strings
## splitByChar(separator, s[, max_substrings])
## splitByChar
Splits a string into substrings separated by a specified character. It uses a constant string `separator` which consists of exactly one character.
Splits a string into substrings separated by a specified character. Uses a constant string `separator` which consists of exactly one character.
Returns an array of selected substrings. Empty substrings may be selected if the separator occurs at the beginning or end of the string, or if there are multiple consecutive separators.
**Syntax**
@ -39,13 +39,15 @@ Type: [Array](../../sql-reference/data-types/array.md)([String](../../sql-refere
SELECT splitByChar(',', '1,2,3,abcde');
```
Result:
``` text
┌─splitByChar(',', '1,2,3,abcde')─┐
│ ['1','2','3','abcde'] │
└─────────────────────────────────┘
```
## splitByString(separator, s[, max_substrings])
## splitByString
Splits a string into substrings separated by a string. It uses a constant string `separator` of multiple characters as the separator. If the string `separator` is empty, it will split the string `s` into an array of single characters.
@ -78,6 +80,8 @@ Type: [Array](../../sql-reference/data-types/array.md)([String](../../sql-refere
SELECT splitByString(', ', '1, 2 3, 4,5, abcde');
```
Result:
``` text
┌─splitByString(', ', '1, 2 3, 4,5, abcde')─┐
│ ['1','2 3','4,5','abcde'] │
@ -88,13 +92,15 @@ SELECT splitByString(', ', '1, 2 3, 4,5, abcde');
SELECT splitByString('', 'abcde');
```
Result:
``` text
┌─splitByString('', 'abcde')─┐
│ ['a','b','c','d','e'] │
└────────────────────────────┘
```
## splitByRegexp(regexp, s[, max_substrings])
## splitByRegexp
Splits a string into substrings separated by a regular expression. It uses a regular expression string `regexp` as the separator. If the `regexp` is empty, it will split the string `s` into an array of single characters. If no match is found for this regular expression, the string `s` won't be split.
@ -123,8 +129,6 @@ Type: [Array](../../sql-reference/data-types/array.md)([String](../../sql-refere
**Example**
Query:
``` sql
SELECT splitByRegexp('\\d+', 'a12bc23de345f');
```
@ -137,8 +141,6 @@ Result:
└────────────────────────────────────────┘
```
Query:
``` sql
SELECT splitByRegexp('', 'abcde');
```
@ -151,7 +153,7 @@ Result:
└────────────────────────────┘
```
## splitByWhitespace(s[, max_substrings])
## splitByWhitespace
Splits a string into substrings separated by whitespace characters.
Returns an array of selected substrings.
@ -180,13 +182,15 @@ Type: [Array](../../sql-reference/data-types/array.md)([String](../../sql-refere
SELECT splitByWhitespace(' 1! a, b. ');
```
Result:
``` text
┌─splitByWhitespace(' 1! a, b. ')─┐
│ ['1!','a,','b.'] │
└─────────────────────────────────────┘
```
## splitByNonAlpha(s[, max_substrings])
## splitByNonAlpha
Splits a string into substrings separated by whitespace and punctuation characters.
Returns an array of selected substrings.
@ -221,23 +225,32 @@ SELECT splitByNonAlpha(' 1! a, b. ');
└───────────────────────────────────┘
```
## arrayStringConcat(arr\[, separator\])
## arrayStringConcat
Concatenates string representations of values listed in the array with the separator. `separator` is an optional parameter: a constant string, set to an empty string by default.
Returns the string.
**Syntax**
```sql
arrayStringConcat(arr\[, separator\])
```
**Example**
``` sql
SELECT arrayStringConcat(['12/05/2021', '12:50:00'], ' ') AS DateString;
```
Result:
```text
┌─DateString──────────┐
│ 12/05/2021 12:50:00 │
└─────────────────────┘
```
## alphaTokens(s[, max_substrings]), splitByAlpha(s[, max_substrings])
## alphaTokens
Selects substrings of consecutive bytes from the ranges a-z and A-Z.Returns an array of substrings.
@ -245,9 +258,10 @@ Selects substrings of consecutive bytes from the ranges a-z and A-Z.Returns an a
``` sql
alphaTokens(s[, max_substrings]))
splitByAlpha(s[, max_substrings])
```
Alias: `splitByAlpha`
**Arguments**
- `s` — The string to split. [String](../../sql-reference/data-types/string.md).
@ -271,7 +285,7 @@ SELECT alphaTokens('abca1abc');
└─────────────────────────┘
```
## extractAllGroups(text, regexp)
## extractAllGroups
Extracts all groups from non-overlapping substrings matched by a regular expression.
@ -296,8 +310,6 @@ Type: [Array](../data-types/array.md).
**Example**
Query:
``` sql
SELECT extractAllGroups('abc=123, 8="hkl"', '("[^"]+"|\\w+)=("[^"]+"|\\w+)');
```
@ -312,7 +324,7 @@ Result:
## ngrams
Splits the UTF-8 string into n-grams of `ngramsize` symbols.
Splits a UTF-8 string into n-grams of `ngramsize` symbols.
**Syntax**
@ -333,8 +345,6 @@ Type: [Array](../../sql-reference/data-types/array.md)([String](../../sql-refere
**Example**
Query:
``` sql
SELECT ngrams('ClickHouse', 3);
```
@ -363,8 +373,6 @@ Type: [Array](../data-types/array.md).
**Example**
Query:
``` sql
SELECT tokens('test1,;\\ test2,;\\ test3,;\\ test4') AS tokens;
```

535
docs/en/sql-reference/functions/string-functions.md
View File

File diff suppressed because it is too large Load Diff

115
docs/en/sql-reference/functions/string-replace-functions.md
View File

@ -1,35 +1,54 @@
---
slug: /en/sql-reference/functions/string-replace-functions
sidebar_position: 42
sidebar_position: 150
sidebar_label: Replacing in Strings
---
# Functions for Searching and Replacing in Strings
# Functions for Replacing in Strings
:::note
Functions for [searching](../../sql-reference/functions/string-search-functions.md) and [other manipulations with strings](../../sql-reference/functions/string-functions.md) are described separately.
:::
[General strings functions](string-functions.md) and [functions for searchin in strings](string-search-functions.md) are described separately.
## replaceOne(haystack, pattern, replacement)
## replaceOne
Replaces the first occurrence of the substring ‘pattern’ (if it exists) in ‘haystack’ by the ‘replacement’ string.
Replaces the first occurrence of the substring `pattern` in `haystack` by the `replacement` string.
## replaceAll(haystack, pattern, replacement), replace(haystack, pattern, replacement)
**Syntax**
Replaces all occurrences of the substring ‘pattern’ in ‘haystack’ by the ‘replacement’ string.
```sql
replaceOne(haystack, pattern, replacement)
```
## replaceAll
Replaces all occurrences of the substring `pattern` in `haystack` by the `replacement` string.
**Syntax**
```sql
replaceAll(haystack, pattern, replacement)
```
Alias: `replace`.
## replaceRegexpOne(haystack, pattern, replacement)
## replaceRegexpOne
Replaces the first occurrence of the substring matching the regular expression ‘pattern’ in ‘haystack‘ by the ‘replacement‘ string.
‘pattern‘ must be a [re2 regular expression](https://github.com/google/re2/wiki/Syntax).
‘replacement’ must be a plain string or a string containing substitutions `\0-\9`.
Replaces the first occurrence of the substring matching the regular expression `pattern` (in [re2 syntax](https://github.com/google/re2/wiki/Syntax)) in `haystack` by the `replacement` string.
`replacement` can containing substitutions `\0-\9`.
Substitutions `\1-\9` correspond to the 1st to 9th capturing group (submatch), substitution `\0` corresponds to the entire match.
To use a verbatim `\` character in the ‘pattern‘ or ‘replacement‘ string, escape it using `\`.
Also keep in mind that string literals require an extra escaping.
Example 1. Converting ISO dates to American format:
To use a verbatim `\` character in the `pattern` or `replacement` strings, escape it using `\`.
Also keep in mind that string literals require extra escaping.
**Syntax**
```sql
replaceRegexpOne(haystack, pattern, replacement)
```
**Example**
Converting ISO dates to American format:
``` sql
SELECT DISTINCT
@ -40,6 +59,8 @@ LIMIT 7
FORMAT TabSeparated
```
Result:
``` text
2014-03-17 03/17/2014
2014-03-18 03/18/2014
@ -50,83 +71,91 @@ FORMAT TabSeparated
2014-03-23 03/23/2014
```
Example 2. Copying a string ten times:
Copying a string ten times:
``` sql
SELECT replaceRegexpOne('Hello, World!', '.*', '\\0\\0\\0\\0\\0\\0\\0\\0\\0\\0') AS res
```
Result:
``` text
┌─res────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ Hello, World!Hello, World!Hello, World!Hello, World!Hello, World!Hello, World!Hello, World!Hello, World!Hello, World!Hello, World! │
└────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
```
## replaceRegexpAll(haystack, pattern, replacement)
## replaceRegexpAll
Like ‘replaceRegexpOne‘, but replaces all occurrences of the pattern. Example:
Like `replaceRegexpOne` but replaces all occurrences of the pattern.
Alias: `REGEXP_REPLACE`.
**Example**
``` sql
SELECT replaceRegexpAll('Hello, World!', '.', '\\0\\0') AS res
```
Result:
``` text
┌─res────────────────────────┐
│ HHeelllloo,, WWoorrlldd!! │
└────────────────────────────┘
```
As an exception, if a regular expression worked on an empty substring, the replacement is not made more than once.
Example:
As an exception, if a regular expression worked on an empty substring, the replacement is not made more than once, e.g.:
``` sql
SELECT replaceRegexpAll('Hello, World!', '^', 'here: ') AS res
```
Result:
``` text
┌─res─────────────────┐
│ here: Hello, World! │
└─────────────────────┘
```
Alias: `REGEXP_REPLACE`.
## regexpQuoteMeta
## regexpQuoteMeta(s)
Adds a backslash before these characters with special meaning in regular expressions: `\0`, `\\`, `|`, `(`, `)`, `^`, `$`, `.`, `[`, `]`, `?`, `*`, `+`, `{`, `:`, `-`.
The function adds a backslash before some predefined characters in the string.
Predefined characters: `\0`, `\\`, `|`, `(`, `)`, `^`, `$`, `.`, `[`, `]`, `?`, `*`, `+`, `{`, `:`, `-`.
This implementation slightly differs from re2::RE2::QuoteMeta. It escapes zero byte as `\0` instead of `\x00` and it escapes only required characters.
For more information, see the link: [RE2](https://github.com/google/re2/blob/master/re2/re2.cc#L473)
For more information, see [RE2](https://github.com/google/re2/blob/master/re2/re2.cc#L473)
**Syntax**
## translate(s, from, to)
```sql
regexpQuoteMeta(s)
```
The function replaces characters in the string ‘s’ in accordance with one-to-one character mapping defined by ‘from’ and ‘to’ strings. ‘from’ and ‘to’ must be constant ASCII strings of the same size. Non-ASCII characters in the original string are not modified.
## translate
Example:
Replaces characters in the string `s` using a one-to-one character mapping defined by `from` and `to` strings. `from` and `to` must be constant ASCII strings of the same size. Non-ASCII characters in the original string are not modified.
**Syntax**
```sql
translate(s, from, to)
```
**Example**
``` sql
SELECT translate('Hello, World!', 'delor', 'DELOR') AS res
```
Result:
``` text
┌─res───────────┐
│ HELLO, WORLD! │
└───────────────┘
```
## translateUTF8(string, from, to)
## translateUTF8
Similar to previous function, but works with UTF-8 arguments. ‘from’ and ‘to’ must be valid constant UTF-8 strings of the same size.
Example:
``` sql
SELECT translateUTF8('Hélló, Wórld¡', 'óé¡', 'oe!') AS res
```
``` text
┌─res───────────┐
│ Hello, World! │
└───────────────┘
```
Like [translate](#translate) but assumes `s`, `from` and `to` are UTF-8 encoded strings.

732
docs/en/sql-reference/functions/string-search-functions.md
View File

@ -1,24 +1,26 @@
---
slug: /en/sql-reference/functions/string-search-functions
sidebar_position: 41
sidebar_position: 160
sidebar_label: Searching in Strings
---
# Functions for Searching in Strings
The search is case-sensitive by default in all these functions. There are separate variants for case insensitive search.
All functions in this section search by default case-sensitively. Case-insensitive search is usually provided by separate function variants.
Note that case-insensitive search follows the lowercase-uppercase rules of the English language. E.g. Uppercased `i` in English language is
`I` whereas in Turkish language it is `İ` - results for languages other than English may be unexpected.
:::note
Functions for [replacing](../../sql-reference/functions/string-replace-functions.md) and [other manipulations with strings](../../sql-reference/functions/string-functions.md) are described separately.
:::
Functions in this section also assume that the searched string and the search string are single-byte encoded text. If this assumption is
violated, no exception is thrown and results are undefined. Search with UTF-8 encoded strings is usually provided by separate function
variants. Likewise, if a UTF-8 function variant is used and the input strings are not UTF-8 encoded text, no exception is thrown and the
results are undefined. Note that no automatic Unicode normalization is performed, you can use the
[normalizeUTF8*()](https://clickhouse.com/docs/en/sql-reference/functions/string-functions/) functions for that.
## position(haystack, needle), locate(haystack, needle)
[General strings functions](string-functions.md) and [functions for replacing in strings](string-replace-functions.md) are described separately.
Searches for the substring `needle` in the string `haystack`.
## position
Returns the position (in bytes) of the found substring in the string, starting from 1.
For a case-insensitive search, use the function [positionCaseInsensitive](#positioncaseinsensitive).
Returns the position (in bytes, starting at 1) of a substring `needle` in a string `haystack`.
**Syntax**
@ -26,35 +28,33 @@ For a case-insensitive search, use the function [positionCaseInsensitive](#posit
position(haystack, needle[, start_pos])
```
``` sql
position(needle IN haystack)
```
Alias: `locate(haystack, needle[, start_pos])`.
:::note
Syntax of `position(needle IN haystack)` provides SQL-compatibility, the function works the same way as to `position(haystack, needle)`.
:::
Alias:
- `position(needle IN haystack)`
- `locate(haystack, needle[, start_pos])`.
**Arguments**
- `haystack` — String, in which substring will to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `haystack` — String in which the search is performed. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` – Position of the first character in the string to start search. [UInt](../../sql-reference/data-types/int-uint.md). Optional.
- `start_pos` – Position (1-based) in `haystack` at which the search starts. [UInt](../../sql-reference/data-types/int-uint.md). Optional.
**Returned values**
- Starting position in bytes (counting from 1), if substring was found.
- Starting position in bytes and counting from 1, if the substring was found.
- 0, if the substring was not found.
If substring `needle` is empty, these rules apply:
- if no `start_pos` was specified: return `1`
- if `start_pos = 0`: return `1`
- if `start_pos >= 1` and `start_pos <= length(haystack) + 1`: return `start_pos`
- otherwise: return `0`
The same rules also apply to functions `positionCaseInsensitive`, `positionUTF8` and `positionCaseInsensitiveUTF8`
Type: `Integer`.
**Examples**
The phrase “Hello, world!” contains a set of bytes representing a single-byte encoded text. The function returns some expected result:
Query:
``` sql
SELECT position('Hello, world!', '!');
```
@ -67,6 +67,8 @@ Result:
└────────────────────────────────┘
```
Example with `start_pos` argument:
``` sql
SELECT
position('Hello, world!', 'o', 1),
@ -79,29 +81,21 @@ SELECT
└───────────────────────────────────┴───────────────────────────────────┘
```
The same phrase in Russian contains characters which can’t be represented using a single byte. The function returns some unexpected result (use [positionUTF8](#positionutf8) function for multi-byte encoded text):
Example for `needle IN haystack` syntax:
Query:
``` sql
SELECT position('Привет, мир!', '!');
```sql
SELECT 6 = position('/' IN s) FROM (SELECT 'Hello/World' AS s);
```
Result:
``` text
┌─position('Привет, мир!', '!')─┐
│ 21 │
└───────────────────────────────┘
```text
┌─equals(6, position(s, '/'))─┐
│ 1 │
└─────────────────────────────┘
```
If argument `needle` is empty the following rules apply:
- if no `start_pos` was specified: return `1`
- if `start_pos = 0`: return `1`
- if `start_pos >= 1` and `start_pos <= length(haystack) + 1`: return `start_pos`
- otherwise: return `0`
The same rules also apply to functions `positionCaseInsensitive`, `positionUTF8` and `positionCaseInsensitiveUTF8`
Examples with empty `needle` substring:
``` sql
SELECT
@ -120,223 +114,59 @@ SELECT
└─────────────────────┴────────────────────────┴────────────────────────┴────────────────────────┴────────────────────────┴────────────────────────┴────────────────────────┘
```
**Examples for POSITION(needle IN haystack) syntax**
Query:
```sql
SELECT 3 = position('c' IN 'abc');
```
Result:
```text
┌─equals(3, position('abc', 'c'))─┐
│ 1 │
└─────────────────────────────────┘
```
Query:
```sql
SELECT 6 = position('/' IN s) FROM (SELECT 'Hello/World' AS s);
```
Result:
```text
┌─equals(6, position(s, '/'))─┐
│ 1 │
└─────────────────────────────┘
```
## positionCaseInsensitive
The same as [position](#position) returns the position (in bytes) of the found substring in the string, starting from 1. Use the function for a case-insensitive search.
Works under the assumption that the string contains a set of bytes representing a single-byte encoded text. If this assumption is not met and a character can’t be represented using a single byte, the function does not throw an exception and returns some unexpected result. If character can be represented using two bytes, it will use two bytes and so on.
**Syntax**
``` sql
positionCaseInsensitive(haystack, needle[, start_pos])
```
**Arguments**
- `haystack` — String, in which substring will to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` — Optional parameter, position of the first character in the string to start search. [UInt](../../sql-reference/data-types/int-uint.md).
**Returned values**
- Starting position in bytes (counting from 1), if substring was found.
- 0, if the substring was not found.
Type: `Integer`.
**Example**
Query:
``` sql
SELECT positionCaseInsensitive('Hello, world!', 'hello');
```
Result:
``` text
┌─positionCaseInsensitive('Hello, world!', 'hello')─┐
│ 1 │
└───────────────────────────────────────────────────┘
```
Like [position](#position) but searches case-insensitively.
## positionUTF8
Returns the position (in Unicode points) of the found substring in the string, starting from 1.
Works under the assumption that the string contains a set of bytes representing a UTF-8 encoded text. If this assumption is not met, the function does not throw an exception and returns some unexpected result. If character can be represented using two Unicode points, it will use two and so on.
For a case-insensitive search, use the function [positionCaseInsensitiveUTF8](#positioncaseinsensitiveutf8).
**Syntax**
``` sql
positionUTF8(haystack, needle[, start_pos])
```
**Arguments**
- `haystack` — String, in which substring will to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` — Optional parameter, position of the first character in the string to start search. [UInt](../../sql-reference/data-types/int-uint.md)
**Returned values**
- Starting position in Unicode points (counting from 1), if substring was found.
- 0, if the substring was not found.
Type: `Integer`.
Like [position](#position) but assumes `haystack` and `needle` are UTF-8 encoded strings.
**Examples**
The phrase “Hello, world!” in Russian contains a set of Unicode points representing a single-point encoded text. The function returns some expected result:
Query:
Function `positionUTF8` correctly counts character `ö` (represented by two points) as a single Unicode codepoint:
``` sql
SELECT positionUTF8('Привет, мир!', '!');
SELECT positionUTF8('Motörhead', 'r');
```
Result:
``` text
┌─positionUTF8('Привет, мир!', '!')─┐
│ 12 │
└───────────────────────────────────┘
```
The phrase “Salut, étudiante!”, where character `é` can be represented using a one point (`U+00E9`) or two points (`U+0065U+0301`) the function can be returned some unexpected result:
Query for the letter `é`, which is represented one Unicode point `U+00E9`:
``` sql
SELECT positionUTF8('Salut, étudiante!', '!');
```
Result:
``` text
┌─positionUTF8('Salut, étudiante!', '!')─┐
│ 17 │
└────────────────────────────────────────┘
```
Query for the letter `é`, which is represented two Unicode points `U+0065U+0301`:
``` sql
SELECT positionUTF8('Salut, étudiante!', '!');
```
Result:
``` text
┌─positionUTF8('Salut, étudiante!', '!')─┐
│ 18 │
└────────────────────────────────────────┘
┌─position('Motörhead', 'r')─┐
│ 5 │
└────────────────────────────┘
```
## positionCaseInsensitiveUTF8
The same as [positionUTF8](#positionutf8), but is case-insensitive. Returns the position (in Unicode points) of the found substring in the string, starting from 1.
Works under the assumption that the string contains a set of bytes representing a UTF-8 encoded text. If this assumption is not met, the function does not throw an exception and returns some unexpected result. If character can be represented using two Unicode points, it will use two and so on.
**Syntax**
``` sql
positionCaseInsensitiveUTF8(haystack, needle[, start_pos])
```
**Arguments**
- `haystack` — String, in which substring will to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` — Optional parameter, position of the first character in the string to start search. [UInt](../../sql-reference/data-types/int-uint.md)
**Returned value**
- Starting position in Unicode points (counting from 1), if substring was found.
- 0, if the substring was not found.
Type: `Integer`.
**Example**
Query:
``` sql
SELECT positionCaseInsensitiveUTF8('Привет, мир!', 'Мир');
```
Result:
``` text
┌─positionCaseInsensitiveUTF8('Привет, мир!', 'Мир')─┐
│ 9 │
└────────────────────────────────────────────────────┘
```
Like [positionUTF8](#positionutf8) but searches case-insensitively.
## multiSearchAllPositions
The same as [position](../../sql-reference/functions/string-search-functions.md#position) but returns `Array` of positions (in bytes) of the found corresponding substrings in the string. Positions are indexed starting from 1.
Like [position](#position) but returns an array of positions (in bytes, starting at 1) for multiple `needle` substrings in a `haystack` string.
The search is performed on sequences of bytes without respect to string encoding and collation.
- For case-insensitive ASCII search, use the function `multiSearchAllPositionsCaseInsensitive`.
- For search in UTF-8, use the function [multiSearchAllPositionsUTF8](#multiSearchAllPositionsUTF8).
- For case-insensitive UTF-8 search, use the function multiSearchAllPositionsCaseInsensitiveUTF8.
:::note
All `multiSearch*()` functions only support up to 2<sup>8</sup> needles.
:::
**Syntax**
``` sql
multiSearchAllPositions(haystack, [needle1, needle2, ..., needlen])
multiSearchAllPositions(haystack, [needle1, needle2, ..., needleN])
```
**Arguments**
- `haystack` — String, in which substring will to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `haystack` — String in which the search is performed. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substrings to be searched. Array
**Returned values**
- Array of starting positions in bytes (counting from 1), if the corresponding substring was found and 0 if not found.
- Array of the starting position in bytes and counting from 1 (if the substring was found) or 0 (if the substring was not found)
**Example**
Query:
``` sql
SELECT multiSearchAllPositions('Hello, World!', ['hello', '!', 'world']);
```
@ -351,103 +181,172 @@ Result:
## multiSearchAllPositionsUTF8
See `multiSearchAllPositions`.
Like [multiSearchAllPositions](#multiSearchAllPositions) but assumes `haystack` and the `needle`-s are UTF-8 encoded strings.
## multiSearchFirstPosition(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
## multiSearchFirstPosition
The same as `position` but returns the leftmost offset of the string `haystack` that is matched to some of the needles.
Like `position` but returns the leftmost offset in a `haystack` string which matches any of multiple `needle` strings.
For a case-insensitive search or/and in UTF-8 format use functions `multiSearchFirstPositionCaseInsensitive, multiSearchFirstPositionUTF8, multiSearchFirstPositionCaseInsensitiveUTF8`.
Functions `multiSearchFirstPositionCaseInsensitive`, `multiSearchFirstPositionUTF8` and `multiSearchFirstPositionCaseInsensitiveUTF8` provide case-insensitive and/or UTF-8 variants of this function.
## multiSearchFirstIndex(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
**Syntax**
```sql
multiSearchFirstPosition(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
```
## multiSearchFirstIndex
Returns the index `i` (starting from 1) of the leftmost found needle<sub>i</sub> in the string `haystack` and 0 otherwise.
For a case-insensitive search or/and in UTF-8 format use functions `multiSearchFirstIndexCaseInsensitive, multiSearchFirstIndexUTF8, multiSearchFirstIndexCaseInsensitiveUTF8`.
Functions `multiSearchFirstIndexCaseInsensitive`, `multiSearchFirstIndexUTF8` and `multiSearchFirstIndexCaseInsensitiveUTF8` provide case-insensitive and/or UTF-8 variants of this function.
## multiSearchAny(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
**Syntax**
```sql
multiSearchFirstIndex(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
```
## multiSearchAny
Returns 1, if at least one string needle<sub>i</sub> matches the string `haystack` and 0 otherwise.
For a case-insensitive search or/and in UTF-8 format use functions `multiSearchAnyCaseInsensitive, multiSearchAnyUTF8, multiSearchAnyCaseInsensitiveUTF8`.
Functions `multiSearchAnyCaseInsensitive`, `multiSearchAnyUTF8` and `multiSearchAnyCaseInsensitiveUTF8` provide case-insensitive and/or UTF-8 variants of this function.
:::note
In all `multiSearch*` functions the number of needles should be less than 2<sup>8</sup> because of implementation specification.
:::
**Syntax**
## match(haystack, pattern), haystack REGEXP pattern operator
```sql
multiSearchAny(haystack, \[needle<sub>1</sub>, needle<sub>2</sub>, …, needle<sub>n</sub>\])
```
Checks whether string `haystack` matches the regular expression `pattern`. The pattern is an [re2 regular expression](https://github.com/google/re2/wiki/Syntax) which has a more limited syntax than Perl regular expressions.
## match
Returns 1 in case of a match, and 0 otherwise.
Returns whether string `haystack` matches the regular expression `pattern` in [re2 regular syntax](https://github.com/google/re2/wiki/Syntax).
Matching is based on UTF-8, e.g. `.` matches the Unicode code point `¥` which is represented in UTF-8 using two bytes. The regular expression must not contain null bytes.
If the haystack or the pattern are not valid UTF-8, then the behavior is undefined.
No automatic Unicode normalization is performed, you can use the [normalizeUTF8*()](https://clickhouse.com/docs/en/sql-reference/functions/string-functions/) functions for that.
Matching is based on UTF-8, e.g. `.` matches the Unicode code point `¥` which is represented in UTF-8 using two bytes. The regular
expression must not contain null bytes. If the haystack or the pattern are not valid UTF-8, then the behavior is undefined.
Unlike re2's default behavior, `.` matches line breaks. To disable this, prepend the pattern with `(?-s)`.
For patterns to search for substrings in a string, it is better to use functions [like](#like) or [position](#position) since they work much faster.
If you only want to search substrings in a string, you can use functions [like](#like) or [position](#position) instead - they work much faster than this function.
## multiMatchAny(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
**Syntax**
The same as `match`, but returns 0 if none of the regular expressions are matched and 1 if any of the patterns matches. For patterns to search substrings in a string, it is better to use `multiSearchAny` since it works much faster.
```sql
match(haystack, pattern)
```
Alias: `haystack REGEXP pattern operator`
## multiMatchAny
Like `match` but returns 1 if at least one of the patterns match and 0 otherwise.
:::note
Functions `multiMatchAny`, `multiMatchAnyIndex`, `multiMatchAllIndices` and their fuzzy equivalents (`multiFuzzyMatchAny`,
`multiFuzzyMatchAnyIndex`, `multiFuzzyMatchAllIndices`) use the (Vectorscan)[https://github.com/VectorCamp/vectorscan] library. As such,
they are only enabled if ClickHouse is compiled with support for vectorscan.
Functions in the `multi[Fuzzy]Match*()` family use the the (Vectorscan)[https://github.com/VectorCamp/vectorscan] library. As such, they are only enabled if ClickHouse is compiled with support for vectorscan.
To turn off all functions that use hyperscan, use setting `SET allow_hyperscan = 0;`.
Due to restrictions of vectorscan, the length of the `haystack` string must be less than 2<sup>32</sup> bytes.
Hyperscan is generally vulnerable to regular expression denial of service (ReDoS) attacks (e.g. see
(here)[https://www.usenix.org/conference/usenixsecurity22/presentation/turonova], (here)[https://doi.org/10.1007/s10664-021-10033-1] and
(here)[ https://doi.org/10.1145/3236024.3236027]. Users are adviced to check the provided patterns carefully.
(here)[https://doi.org/10.1145/3236024.3236027]. Users are adviced to check the provided patterns carefully.
:::
## multiMatchAnyIndex(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
If you only want to search multiple substrings in a string, you can use function [multiSearchAny](#multisearchany) instead - it works much faster than this function.
The same as `multiMatchAny`, but returns any index that matches the haystack.
**Syntax**
## multiMatchAllIndices(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```sql
multiMatchAny(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
The same as `multiMatchAny`, but returns the array of all indices that match the haystack in any order.
## multiMatchAnyIndex
## multiFuzzyMatchAny(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
Like `multiMatchAny` but returns any index that matches the haystack.
The same as `multiMatchAny`, but returns 1 if any pattern matches the haystack within a constant [edit distance](https://en.wikipedia.org/wiki/Edit_distance). This function relies on the experimental feature of [hyperscan](https://intel.github.io/hyperscan/dev-reference/compilation.html#approximate-matching) library, and can be slow for some corner cases. The performance depends on the edit distance value and patterns used, but it's always more expensive compared to a non-fuzzy variants.
**Syntax**
## multiFuzzyMatchAnyIndex(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```sql
multiMatchAnyIndex(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
The same as `multiFuzzyMatchAny`, but returns any index that matches the haystack within a constant edit distance.
## multiMatchAllIndices
## multiFuzzyMatchAllIndices(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
Like `multiMatchAny` but returns the array of all indices that match the haystack in any order.
The same as `multiFuzzyMatchAny`, but returns the array of all indices in any order that match the haystack within a constant edit distance.
**Syntax**
```sql
multiMatchAllIndices(haystack, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
## multiFuzzyMatchAny
Like `multiMatchAny` but returns 1 if any pattern matches the haystack within a constant [edit distance](https://en.wikipedia.org/wiki/Edit_distance). This function relies on the experimental feature of [hyperscan](https://intel.github.io/hyperscan/dev-reference/compilation.html#approximate-matching) library, and can be slow for some corner cases. The performance depends on the edit distance value and patterns used, but it's always more expensive compared to a non-fuzzy variants.
:::note
`multiFuzzyMatch*` functions do not support UTF-8 regular expressions, and such expressions are treated as bytes because of hyperscan restriction.
`multiFuzzyMatch*()` function family do not support UTF-8 regular expressions (it threats them as a sequence of bytes) due to restrictions of hyperscan.
:::
:::note
To turn off all functions that use hyperscan, use setting `SET allow_hyperscan = 0;`.
:::
**Syntax**
## extract(haystack, pattern)
```sql
multiFuzzyMatchAny(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
Extracts a fragment of a string using a regular expression. If ‘haystack’ does not match the ‘pattern’ regex, an empty string is returned. If the regex does not contain subpatterns, it takes the fragment that matches the entire regex. Otherwise, it takes the fragment that matches the first subpattern.
## multiFuzzyMatchAnyIndex
## extractAll(haystack, pattern)
Like `multiFuzzyMatchAny` but returns any index that matches the haystack within a constant edit distance.
Extracts all the fragments of a string using a regular expression. If ‘haystack’ does not match the ‘pattern’ regex, an empty string is returned. Returns an array of strings consisting of all matches to the regex. In general, the behavior is the same as the ‘extract’ function (it takes the first subpattern, or the entire expression if there isn’t a subpattern).
**Syntax**
```sql
multiFuzzyMatchAnyIndex(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
## multiFuzzyMatchAllIndices
Like `multiFuzzyMatchAny` but returns the array of all indices in any order that match the haystack within a constant edit distance.
**Syntax**
```sql
multiFuzzyMatchAllIndices(haystack, distance, \[pattern<sub>1</sub>, pattern<sub>2</sub>, …, pattern<sub>n</sub>\])
```
## extract
Extracts a fragment of a string using a regular expression. If `haystack` does not match the `pattern` regex, an empty string is returned.
For regex without subpatterns, the function uses the fragment that matches the entire regex. Otherwise, it uses the fragment that matches the first subpattern.
**Syntax**
```sql
extract(haystack, pattern)
```
## extractAll
Extracts all fragments of a string using a regular expression. If `haystack` does not match the `pattern` regex, an empty string is returned.
Returns an array of strings consisting of all matches of the regex.
The behavior with respect to subpatterns is the same as in function `extract`.
**Syntax**
```sql
extractAll(haystack, pattern)
```
## extractAllGroupsHorizontal
Matches all groups of the `haystack` string using the `pattern` regular expression. Returns an array of arrays, where the first array includes all fragments matching the first group, the second array - matching the second group, etc.
:::note
`extractAllGroupsHorizontal` function is slower than [extractAllGroupsVertical](#extractallgroups-vertical).
:::
This function is slower than [extractAllGroupsVertical](#extractallgroups-vertical).
**Syntax**
@ -468,8 +367,6 @@ If `haystack` does not match the `pattern` regex, an array of empty arrays is re
**Example**
Query:
``` sql
SELECT extractAllGroupsHorizontal('abc=111, def=222, ghi=333', '("[^"]+"|\\w+)=("[^"]+"|\\w+)');
```
@ -482,10 +379,6 @@ Result:
└──────────────────────────────────────────────────────────────────────────────────────────┘
```
**See Also**
- [extractAllGroupsVertical](#extractallgroups-vertical)
## extractAllGroupsVertical
Matches all groups of the `haystack` string using the `pattern` regular expression. Returns an array of arrays, where each array includes matching fragments from every group. Fragments are grouped in order of appearance in the `haystack`.
@ -509,8 +402,6 @@ If `haystack` does not match the `pattern` regex, an empty array is returned.
**Example**
Query:
``` sql
SELECT extractAllGroupsVertical('abc=111, def=222, ghi=333', '("[^"]+"|\\w+)=("[^"]+"|\\w+)');
```
@ -523,119 +414,88 @@ Result:
└────────────────────────────────────────────────────────────────────────────────────────┘
```
**See Also**
## like
- [extractAllGroupsHorizontal](#extractallgroups-horizontal)
Returns whether string `haystack` matches the LIKE expression `pattern`.
## like(haystack, pattern), haystack LIKE pattern operator
Checks whether a string matches a LIKE expression.
A LIKE expression contains a mix of normal characters and the following metasymbols:
A LIKE expression can contain normal characters and the following metasymbols:
- `%` indicates an arbitrary number of arbitrary characters (including zero characters).
- `_` indicates a single arbitrary character.
- `\` is for escaping literals `%`, `_` and `\`.
Matching is based on UTF-8, e.g. `_` matches the Unicode code point `¥` which is represented in UTF-8 using two bytes.
If the haystack or the pattern are not valid UTF-8, then the behavior is undefined.
If the haystack or the LIKE expression are not valid UTF-8, the behavior is undefined.
No automatic Unicode normalization is performed, you can use the [normalizeUTF8*()](https://clickhouse.com/docs/en/sql-reference/functions/string-functions/) functions for that.
To match against literals `%`, `_` and `/` (which are LIKE metacharacters), prepend them with a backslash, i.e. `\%`, `\_` and `\\`.
The backslash loses its special meaning, i.e. is interpreted literally, if it prepends a character different than `%`, `_` or `\`.
To match against literal `%`, `_` and `/` (which are LIKE metacharacters), prepend them with a backslash: `\%`, `\_` and `\\`.
The backslash loses its special meaning (i.e. is interpreted literally) if it prepends a character different than `%`, `_` or `\`.
Note that ClickHouse requires backslashes in strings [to be quoted as well](../syntax.md#string), so you would actually need to write `\\%`, `\\_` and `\\\\`.
For patterns of the form `%needle%`, the function is as fast as the `position` function.
Other LIKE expressions are internally converted to a regular expression and executed with a performance similar to function `match`.
## notLike(haystack, pattern), haystack NOT LIKE pattern operator
The same thing as `like`, but negative.
## ilike
Case insensitive variant of [like](https://clickhouse.com/docs/en/sql-reference/functions/string-search-functions/#function-like) function. You can use `ILIKE` operator instead of the `ilike` function.
The function ignores the language, e.g. for Turkish (i/İ), the result might be incorrect.
For LIKE expressions of the form `%needle%`, the function is as fast as the `position` function.
All other LIKE expressions are internally converted to a regular expression and executed with a performance similar to function `match`.
**Syntax**
``` sql
ilike(haystack, pattern)
```sql
like(haystack, pattern)
```
**Arguments**
Alias: `haystack LIKE pattern` (operator)
- `haystack` — Input string. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `pattern` — If `pattern` does not contain percent signs or underscores, then the `pattern` only represents the string itself. An underscore (`_`) in `pattern` stands for (matches) any single character. A percent sign (`%`) matches any sequence of zero or more characters.
## notLike
Some `pattern` examples:
Like `like` but negates the result.
``` text
'abc' ILIKE 'abc' true
'abc' ILIKE 'a%' true
'abc' ILIKE '_b_' true
'abc' ILIKE 'c' false
Alias: `haystack NOT LIKE pattern` (operator)
## ilike
Like `like` but searches case-insensitively.
Alias: `haystack ILIKE pattern` (operator)
## notILike
Like `ilike` but negates the result.
Alias: `haystack NOT ILIKE pattern` (operator)
## ngramDistance
Calculates the 4-gram distance between a `haystack` string and a `needle` string. For that, it counts the symmetric difference between two multisets of 4-grams and normalizes it by the sum of their cardinalities. Returns a Float32 between 0 and 1. The smaller the result is, the more strings are similar to each other. Throws an exception if constant `needle` or `haystack` arguments are more than 32Kb in size. If any of non-constant `haystack` or `needle` arguments is more than 32Kb in size, the distance is always 1.
Functions `ngramDistanceCaseInsensitive, ngramDistanceUTF8, ngramDistanceCaseInsensitiveUTF8` provide case-insensitive and/or UTF-8 variants of this function.
**Syntax**
```sql
ngramDistance(haystack, needle)
```
**Returned values**
## ngramSearch
- True, if the string matches `pattern`.
- False, if the string does not match `pattern`.
Like `ngramDistance` but calculates the non-symmetric difference between a `needle` string and a `haystack` string, i.e. the number of n-grams from needle minus the common number of n-grams normalized by the number of `needle` n-grams. Returns a Float32 between 0 and 1. The bigger the result is, the more likely `needle` is in the `haystack`. This function is useful for fuzzy string search. Also see function `soundex`.
**Example**
Input table:
``` text
┌─id─┬─name─────┬─days─┐
│ 1 │ January │ 31 │
│ 2 │ February │ 29 │
│ 3 │ March │ 31 │
│ 4 │ April │ 30 │
└────┴──────────┴──────┘
```
Query:
``` sql
SELECT * FROM Months WHERE ilike(name, '%j%');
```
Result:
``` text
┌─id─┬─name────┬─days─┐
│ 1 │ January │ 31 │
└────┴─────────┴──────┘
```
## notILike(haystack, pattern), haystack NOT ILIKE pattern operator
The same thing as `ilike`, but negative.
## ngramDistance(haystack, needle)
Calculates the 4-gram distance between `haystack` and `needle`: counts the symmetric difference between two multisets of 4-grams and normalizes it by the sum of their cardinalities. Returns float number from 0 to 1 – the closer to zero, the more strings are similar to each other. If the constant `needle` or `haystack` is more than 32Kb, throws an exception. If some of the non-constant `haystack` or `needle` strings are more than 32Kb, the distance is always one.
For case-insensitive search or/and in UTF-8 format use functions `ngramDistanceCaseInsensitive, ngramDistanceUTF8, ngramDistanceCaseInsensitiveUTF8`.
## ngramSearch(haystack, needle)
Same as `ngramDistance` but calculates the non-symmetric difference between `needle` and `haystack` – the number of n-grams from needle minus the common number of n-grams normalized by the number of `needle` n-grams. The closer to one, the more likely `needle` is in the `haystack`. Can be useful for fuzzy string search.
For case-insensitive search or/and in UTF-8 format use functions `ngramSearchCaseInsensitive, ngramSearchUTF8, ngramSearchCaseInsensitiveUTF8`.
Functions `ngramSearchCaseInsensitive, ngramSearchUTF8, ngramSearchCaseInsensitiveUTF8` provide case-insensitive and/or UTF-8 variants of this function.
:::note
For UTF-8 case we use 3-gram distance. All these are not perfectly fair n-gram distances. We use 2-byte hashes to hash n-grams and then calculate the (non-)symmetric difference between these hash tables – collisions may occur. With UTF-8 case-insensitive format we do not use fair `tolower` function – we zero the 5-th bit (starting from zero) of each codepoint byte and first bit of zeroth byte if bytes more than one – this works for Latin and mostly for all Cyrillic letters.
The UTF-8 variants use the 3-gram distance. These are not perfectly fair n-gram distances. We use 2-byte hashes to hash n-grams and then calculate the (non-)symmetric difference between these hash tables – collisions may occur. With UTF-8 case-insensitive format we do not use fair `tolower` function – we zero the 5-th bit (starting from zero) of each codepoint byte and first bit of zeroth byte if bytes more than one – this works for Latin and mostly for all Cyrillic letters.
:::
**Syntax**
```sql
ngramSearch(haystack, needle)
```
## countSubstrings
Returns the number of substring occurrences.
Returns how often substring `needle` occurs in string `haystack`.
For a case-insensitive search, use [countSubstringsCaseInsensitive](../../sql-reference/functions/string-search-functions.md#countSubstringsCaseInsensitive) or [countSubstringsCaseInsensitiveUTF8](../../sql-reference/functions/string-search-functions.md#countSubstringsCaseInsensitiveUTF8) functions.
Functions `countSubstringsCaseInsensitive` and `countSubstringsCaseInsensitiveUTF8` provide a case-insensitive and case-insensitive + UTF-8 variants of this function.
**Syntax**
@ -645,34 +505,18 @@ countSubstrings(haystack, needle[, start_pos])
**Arguments**
- `haystack` — The string to search in. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — The substring to search for. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` – Position of the first character in the string to start search. Optional. [UInt](../../sql-reference/data-types/int-uint.md).
- `haystack` — String in which the search is performed. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — Substring to be searched. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` – Position (1-based) in `haystack` at which the search starts. [UInt](../../sql-reference/data-types/int-uint.md). Optional.
**Returned values**
- Number of occurrences.
- The number of occurrences.
Type: [UInt64](../../sql-reference/data-types/int-uint.md).
**Examples**
Query:
``` sql
SELECT countSubstrings('foobar.com', '.');
```
Result:
``` text
┌─countSubstrings('foobar.com', '.')─┐
│ 1 │
└────────────────────────────────────┘
```
Query:
``` sql
SELECT countSubstrings('aaaa', 'aa');
```
@ -685,7 +529,7 @@ Result:
└───────────────────────────────┘
```
Query:
Example with `start_pos` argument:
```sql
SELECT countSubstrings('abc___abc', 'abc', 4);
@ -699,125 +543,7 @@ Result:
└────────────────────────────────────────┘
```
## countSubstringsCaseInsensitive
Returns the number of substring occurrences case-insensitive.
**Syntax**
``` sql
countSubstringsCaseInsensitive(haystack, needle[, start_pos])
```
**Arguments**
- `haystack` — The string to search in. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — The substring to search for. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` — Position of the first character in the string to start search. Optional. [UInt](../../sql-reference/data-types/int-uint.md).
**Returned values**
- Number of occurrences.
Type: [UInt64](../../sql-reference/data-types/int-uint.md).
**Examples**
Query:
``` sql
SELECT countSubstringsCaseInsensitive('aba', 'B');
```
Result:
``` text
┌─countSubstringsCaseInsensitive('aba', 'B')─┐
│ 1 │
└────────────────────────────────────────────┘
```
Query:
``` sql
SELECT countSubstringsCaseInsensitive('foobar.com', 'CoM');
```
Result:
``` text
┌─countSubstringsCaseInsensitive('foobar.com', 'CoM')─┐
│ 1 │
└─────────────────────────────────────────────────────┘
```
Query:
``` sql
SELECT countSubstringsCaseInsensitive('abC___abC', 'aBc', 2);
```
Result:
``` text
┌─countSubstringsCaseInsensitive('abC___abC', 'aBc', 2)─┐
│ 1 │
└───────────────────────────────────────────────────────┘
```
## countSubstringsCaseInsensitiveUTF8
Returns the number of substring occurrences in `UTF-8` case-insensitive.
**Syntax**
``` sql
SELECT countSubstringsCaseInsensitiveUTF8(haystack, needle[, start_pos])
```
**Arguments**
- `haystack` — The string to search in. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `needle` — The substring to search for. [String](../../sql-reference/syntax.md#syntax-string-literal).
- `start_pos` — Position of the first character in the string to start search. Optional. [UInt](../../sql-reference/data-types/int-uint.md).
**Returned values**
- Number of occurrences.
Type: [UInt64](../../sql-reference/data-types/int-uint.md).
**Examples**
Query:
``` sql
SELECT countSubstringsCaseInsensitiveUTF8('абв', 'A');
```
Result:
``` text
┌─countSubstringsCaseInsensitiveUTF8('абв', 'A')─┐
│ 1 │
└────────────────────────────────────────────────┘
```
Query:
```sql
SELECT countSubstringsCaseInsensitiveUTF8('аБв__АбВ__абв', 'Абв');
```
Result:
``` text
┌─countSubstringsCaseInsensitiveUTF8('аБв__АбВ__абв', 'Абв')─┐
│ 3 │
└────────────────────────────────────────────────────────────┘
```
## countMatches(haystack, pattern)
## countMatches
Returns the number of regular expression matches for a `pattern` in a `haystack`.
@ -840,8 +566,6 @@ Type: [UInt64](../../sql-reference/data-types/int-uint.md).
**Examples**
Query:
``` sql
SELECT countMatches('foobar.com', 'o+');
```
@ -854,8 +578,6 @@ Result:
└──────────────────────────────────┘
```
Query:
``` sql
SELECT countMatches('aaaa', 'aa');
```
@ -868,7 +590,7 @@ Result:
└───────────────────────────────┘
```
## regexpExtract(haystack, pattern[, index])
## regexpExtract
Extracts the first string in haystack that matches the regexp pattern and corresponds to the regex group index.
@ -899,8 +621,12 @@ SELECT
regexpExtract('100-200', '(\\d+)-(\\d+)', 1),
regexpExtract('100-200', '(\\d+)-(\\d+)', 2),
regexpExtract('100-200', '(\\d+)-(\\d+)', 0),
regexpExtract('100-200', '(\\d+)-(\\d+)')
regexpExtract('100-200', '(\\d+)-(\\d+)');
```
Result:
``` text
┌─regexpExtract('100-200', '(\\d+)-(\\d+)', 1)─┬─regexpExtract('100-200', '(\\d+)-(\\d+)', 2)─┬─regexpExtract('100-200', '(\\d+)-(\\d+)', 0)─┬─regexpExtract('100-200', '(\\d+)-(\\d+)')─┐
│ 100 │ 200 │ 100-200 │ 100 │
└──────────────────────────────────────────────┴──────────────────────────────────────────────┴──────────────────────────────────────────────┴───────────────────────────────────────────┘

2
docs/en/sql-reference/functions/time-window-functions.md
View File

@ -1,6 +1,6 @@
---
slug: /en/sql-reference/functions/time-window-functions
sidebar_position: 68
sidebar_position: 175
sidebar_label: Time Window
---

3
docs/en/sql-reference/functions/tuple-functions.md
View File

@ -1,8 +1,7 @@
---
slug: /en/sql-reference/functions/tuple-functions
sidebar_position: 66
sidebar_position: 180
sidebar_label: Tuples
title: "Functions for Working with Tuples"
---
## tuple

3
docs/en/sql-reference/functions/tuple-map-functions.md
View File

@ -1,8 +1,7 @@
---
slug: /en/sql-reference/functions/tuple-map-functions
sidebar_position: 46
sidebar_position: 120
sidebar_label: Maps
title: "Functions for Maps"
---
## map

Some files were not shown because too many files have changed in this diff Show More