mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-09-19 16:20:50 +00:00
Merge pull request #69326 from rschu1ze/overlay-followup
Minor follow-up for #66933
This commit is contained in:
Robert Schulze
GitHub
commit
26f21a0d94
@ -20,10 +20,10 @@ overlay(s, replace, offset[, length])
|
||||
|
||||
**Parameters**
|
||||
|
||||
- `input`: A string type [String](../data-types/string.md).
|
||||
- `s`: A string type [String](../data-types/string.md).
|
||||
- `replace`: A string type [String](../data-types/string.md).
|
||||
- `offset`: An integer type [Int](../data-types/int-uint.md). If `offset` is negative, it is counted from the end of the `input` string.
|
||||
- `length`: Optional. An integer type [Int](../data-types/int-uint.md). `length` specifies the length of the snippet within input to be replaced. If `length` is not specified, the number of bytes removed from `input` equals the length of `replace`; otherwise `length` bytes are removed.
|
||||
- `offset`: An integer type [Int](../data-types/int-uint.md) (1-based). If `offset` is negative, it is counted from the end of the string `s`.
|
||||
- `length`: Optional. An integer type [Int](../data-types/int-uint.md). `length` specifies the length of the snippet within the input string `s` to be replaced. If `length` is not specified, the number of bytes removed from `s` equals the length of `replace`; otherwise `length` bytes are removed.
|
||||
|
||||
**Returned value**
|
||||
|
||||
@ -32,22 +32,35 @@ overlay(s, replace, offset[, length])
|
||||
**Example**
|
||||
|
||||
```sql
|
||||
SELECT overlay('ClickHouse SQL', 'CORE', 12) AS res;
|
||||
SELECT overlay('My father is from Mexico.', 'mother', 4) AS res;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```text
|
||||
┌─res─────────────┐
|
||||
│ ClickHouse CORE │
|
||||
└─────────────────┘
|
||||
┌─res──────────────────────┐
|
||||
│ My mother is from Mexico.│
|
||||
└──────────────────────────┘
|
||||
```
|
||||
|
||||
```sql
|
||||
SELECT overlay('My father is from Mexico.', 'dad', 4, 6) AS res;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```text
|
||||
┌─res───────────────────┐
|
||||
│ My dad is from Mexico.│
|
||||
└───────────────────────┘
|
||||
```
|
||||
|
||||
## overlayUTF8
|
||||
|
||||
Replace part of the string `input` with another string `replace`, starting at the 1-based index `offset`.
|
||||
|
||||
Assumes that the string contains valid UTF-8 encoded text. If this assumption is violated, no exception is thrown and the result is undefined.
|
||||
Assumes that the string contains valid UTF-8 encoded text.
|
||||
If this assumption is violated, no exception is thrown and the result is undefined.
|
||||
|
||||
**Syntax**
|
||||
|
||||
@ -59,8 +72,8 @@ overlayUTF8(s, replace, offset[, length])
|
||||
|
||||
- `s`: A string type [String](../data-types/string.md).
|
||||
- `replace`: A string type [String](../data-types/string.md).
|
||||
- `offset`: An integer type [Int](../data-types/int-uint.md). If `offset` is negative, it is counted from the end of the `input` string.
|
||||
- `length`: Optional. An integer type [Int](../data-types/int-uint.md). `length` specifies the length of the snippet within input to be replaced. If `length` is not specified, the number of characters removed from `input` equals the length of `replace`; otherwise `length` characters are removed.
|
||||
- `offset`: An integer type [Int](../data-types/int-uint.md) (1-based). If `offset` is negative, it is counted from the end of the input string `s`.
|
||||
- `length`: Optional. An integer type [Int](../data-types/int-uint.md). `length` specifies the length of the snippet within the input string `s` to be replaced. If `length` is not specified, the number of characters removed from `s` equals the length of `replace`; otherwise `length` characters are removed.
|
||||
|
||||
**Returned value**
|
||||
|
||||
@ -69,15 +82,15 @@ overlayUTF8(s, replace, offset[, length])
|
||||
**Example**
|
||||
|
||||
```sql
|
||||
SELECT overlayUTF8('ClickHouse是一款OLAP数据库', '开源', 12, 2) AS res;
|
||||
SELECT overlay('Mein Vater ist aus Österreich.', 'der Türkei', 20) AS res;
|
||||
```
|
||||
|
||||
Result:
|
||||
|
||||
```text
|
||||
┌─res────────────────────────┐
|
||||
│ ClickHouse是开源OLAP数据库 │
|
||||
└────────────────────────────┘
|
||||
┌─res───────────────────────────┐
|
||||
│ Mein Vater ist aus der Türkei.│
|
||||
└───────────────────────────────┘
|
||||
```
|
||||
|
||||
## replaceOne
|
||||
|
||||
@ -1,12 +1,12 @@
|
||||
#include <Columns/ColumnConst.h>
|
||||
#include <Columns/ColumnString.h>
|
||||
#include <Common/StringUtils.h>
|
||||
#include <Common/UTF8Helpers.h>
|
||||
#include <DataTypes/DataTypeString.h>
|
||||
#include <Functions/FunctionFactory.h>
|
||||
#include <Functions/FunctionHelpers.h>
|
||||
#include <Functions/GatherUtils/Sources.h>
|
||||
#include <Functions/IFunction.h>
|
||||
#include <Common/StringUtils.h>
|
||||
#include <Common/UTF8Helpers.h>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
@ -16,8 +16,8 @@ namespace
|
||||
|
||||
/// If 'is_utf8' - measure offset and length in code points instead of bytes.
|
||||
/// Syntax:
|
||||
/// - overlay(input, replace, offset[, length])
|
||||
/// - overlayUTF8(input, replace, offset[, length]) - measure offset and length in code points instead of bytes
|
||||
/// - overlay(s, replace, offset[, length])
|
||||
/// - overlayUTF8(s, replace, offset[, length]) - measure offset and length in code points instead of bytes
|
||||
template <bool is_utf8>
|
||||
class FunctionOverlay : public IFunction
|
||||
{
|
||||
@ -34,7 +34,7 @@ public:
|
||||
DataTypePtr getReturnTypeImpl(const ColumnsWithTypeAndName & arguments) const override
|
||||
{
|
||||
FunctionArgumentDescriptors mandatory_args{
|
||||
{"input", static_cast<FunctionArgumentDescriptor::TypeValidator>(&isString), nullptr, "String"},
|
||||
{"s", static_cast<FunctionArgumentDescriptor::TypeValidator>(&isString), nullptr, "String"},
|
||||
{"replace", static_cast<FunctionArgumentDescriptor::TypeValidator>(&isString), nullptr, "String"},
|
||||
{"offset", static_cast<FunctionArgumentDescriptor::TypeValidator>(&isNativeInteger), nullptr, "(U)Int8/16/32/64"},
|
||||
};
|
||||
@ -100,7 +100,6 @@ public:
|
||||
res_data.reserve(col_input_string->getChars().size());
|
||||
}
|
||||
|
||||
|
||||
#define OVERLAY_EXECUTE_CASE(HAS_FOUR_ARGS, OFFSET_IS_CONST, LENGTH_IS_CONST) \
|
||||
if (input_is_const && replace_is_const) \
|
||||
constantConstant<HAS_FOUR_ARGS, OFFSET_IS_CONST, LENGTH_IS_CONST>( \
|
||||
@ -186,7 +185,6 @@ public:
|
||||
return res_col;
|
||||
}
|
||||
|
||||
|
||||
private:
|
||||
/// input offset is 1-based, maybe negative
|
||||
/// output result is 0-based valid offset, within [0, input_size]
|
||||
@ -229,6 +227,7 @@ private:
|
||||
ColumnString::Chars & res_data,
|
||||
ColumnString::Offsets & res_offsets) const
|
||||
{
|
||||
/// Free us from handling negative length in the code below
|
||||
if (has_four_args && length_is_const && const_length < 0)
|
||||
{
|
||||
constantConstant<true, offset_is_const, false>(
|
||||
@ -343,6 +342,7 @@ private:
|
||||
ColumnString::Chars & res_data,
|
||||
ColumnString::Offsets & res_offsets) const
|
||||
{
|
||||
/// Free us from handling negative length in the code below
|
||||
if (has_four_args && length_is_const && const_length < 0)
|
||||
{
|
||||
vectorConstant<true, offset_is_const, false>(
|
||||
@ -461,6 +461,7 @@ private:
|
||||
ColumnString::Chars & res_data,
|
||||
ColumnString::Offsets & res_offsets) const
|
||||
{
|
||||
/// Free us from handling negative length in the code below
|
||||
if (has_four_args && length_is_const && const_length < 0)
|
||||
{
|
||||
constantVector<true, offset_is_const, false>(
|
||||
@ -577,6 +578,7 @@ private:
|
||||
ColumnString::Chars & res_data,
|
||||
ColumnString::Offsets & res_offsets) const
|
||||
{
|
||||
/// Free us from handling negative length in the code below
|
||||
if (has_four_args && length_is_const && const_length < 0)
|
||||
{
|
||||
vectorVector<true, offset_is_const, false>(
|
||||
|
||||
@ -35,10 +35,10 @@ SELECT overlay('Spark SQL', materialize('ANSI '), materialize(7), materialize(0)
|
||||
SELECT overlay(materialize('Spark SQL'), materialize('ANSI '), materialize(7), materialize(0)), overlayUTF8(materialize('Spark SQL和CH'), materialize('ANSI '), materialize(7), materialize(0));
|
||||
|
||||
SELECT 'Test with special offset values';
|
||||
WITH number - 12 AS offset SELECT offset, overlay('Spark SQL', '__', offset), overlayUTF8('Spark SQL和CH', '之', offset) FROM numbers(26);
|
||||
WITH number - 12 AS offset SELECT offset, overlay('Spark SQL', '__', offset), overlayUTF8('Spark SQL和CH', '之', offset) FROM numbers(26) ORDER BY number;
|
||||
|
||||
SELECT 'Test with special length values';
|
||||
WITH number - 1 AS length SELECT length, overlay('Spark SQL', 'ANSI ', 7, length), overlayUTF8('Spark SQL和CH', 'ANSI ', 7, length) FROM numbers(8);
|
||||
WITH number - 1 AS length SELECT length, overlay('Spark SQL', 'ANSI ', 7, length), overlayUTF8('Spark SQL和CH', 'ANSI ', 7, length) FROM numbers(8) ORDER BY number;
|
||||
|
||||
SELECT 'Test with special input and replace values';
|
||||
SELECT overlay('', '_', 6), overlayUTF8('', '_', 6);
|
||||
|
||||
Loading…
Reference in New Issue
Block a user