2012-07-16 03:42:36 +00:00
|
|
|
|
#pragma once
|
|
|
|
|
|
|
|
|
|
#include <DB/DataTypes/DataTypeString.h>
|
|
|
|
|
#include <DB/Columns/ColumnString.h>
|
|
|
|
|
#include <DB/Columns/ColumnConst.h>
|
2016-12-15 12:05:05 +00:00
|
|
|
|
#include <DB/Common/StringUtils.h>
|
|
|
|
|
#include <DB/Common/StringView.h>
|
2012-07-16 03:42:36 +00:00
|
|
|
|
#include <DB/Functions/FunctionsString.h>
|
2013-03-05 13:30:23 +00:00
|
|
|
|
#include <DB/Functions/FunctionsStringArray.h>
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2016-10-26 22:27:38 +00:00
|
|
|
|
#ifdef __APPLE__
|
|
|
|
|
#include <common/apple_memrchr.h>
|
|
|
|
|
#endif
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2017-03-12 11:09:25 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
namespace DB
|
|
|
|
|
{
|
|
|
|
|
|
2016-10-20 05:21:49 +00:00
|
|
|
|
/** URL processing functions.
|
|
|
|
|
* All functions are not strictly follow RFC, instead they are maximally simplified for performance reasons.
|
2012-07-16 03:42:36 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Functions for extraction parts of URL.
|
|
|
|
|
* If URL has nothing like, then empty string is returned.
|
2014-06-26 00:58:14 +00:00
|
|
|
|
*
|
2012-07-16 03:42:36 +00:00
|
|
|
|
* domain
|
|
|
|
|
* domainWithoutWWW
|
|
|
|
|
* topLevelDomain
|
|
|
|
|
* protocol
|
|
|
|
|
* path
|
|
|
|
|
* queryString
|
|
|
|
|
* fragment
|
|
|
|
|
* queryStringAndFragment
|
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Functions, removing parts from URL.
|
|
|
|
|
* If URL has nothing like, then it is retured unchanged.
|
2014-06-26 00:58:14 +00:00
|
|
|
|
*
|
2012-07-16 03:42:36 +00:00
|
|
|
|
* cutWWW
|
|
|
|
|
* cutFragment
|
|
|
|
|
* cutQueryString
|
|
|
|
|
* cutQueryStringAndFragment
|
2012-07-21 03:45:48 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Extract value of parameter in query string or in fragment identifier. Return empty string, if URL has no such parameter.
|
|
|
|
|
* If there are many parameters with same name - return value of first one. Value is not %-decoded.
|
2012-07-21 03:45:48 +00:00
|
|
|
|
*
|
2013-03-18 10:27:45 +00:00
|
|
|
|
* extractURLParameter(URL, name)
|
2014-06-26 00:58:14 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Extract all parameters from URL in form of array of strings name=value.
|
2013-03-18 10:27:45 +00:00
|
|
|
|
* extractURLParameters(URL)
|
2013-08-02 13:55:43 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Extract names of all parameters from URL in form of array of strings.
|
2013-08-05 08:40:56 +00:00
|
|
|
|
* extractURLParameterNames(URL)
|
2014-06-26 00:58:14 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Remove specified parameter from URL.
|
2013-03-18 10:27:45 +00:00
|
|
|
|
* cutURLParameter(URL, name)
|
2014-06-26 00:58:14 +00:00
|
|
|
|
*
|
2016-10-20 05:21:49 +00:00
|
|
|
|
* Get array of URL 'hierarchy' as in Yandex.Metrica tree-like reports. See docs.
|
2013-03-18 10:27:45 +00:00
|
|
|
|
* URLHierarchy(URL)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
*/
|
|
|
|
|
|
2016-05-28 10:35:44 +00:00
|
|
|
|
using Pos = const char *;
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2016-12-15 12:05:05 +00:00
|
|
|
|
|
|
|
|
|
/// Extracts scheme from given url.
|
2016-12-15 12:33:50 +00:00
|
|
|
|
inline StringView getURLScheme(const StringView & url)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
2016-12-15 12:05:05 +00:00
|
|
|
|
// scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
|
2016-12-15 18:59:07 +00:00
|
|
|
|
const char * p = url.data();
|
|
|
|
|
const char * end = url.data() + url.size();
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-12-15 12:05:05 +00:00
|
|
|
|
if (isAlphaASCII(*p))
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
2016-12-15 12:05:05 +00:00
|
|
|
|
for (++p; p < end; ++p)
|
|
|
|
|
{
|
|
|
|
|
if (!(isAlphaNumericASCII(*p) || *p == '+' || *p == '-' || *p == '.'))
|
|
|
|
|
{
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return StringView(url.data(), p - url.data());
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-12-15 12:05:05 +00:00
|
|
|
|
return StringView();
|
|
|
|
|
}
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
|
|
|
|
|
2016-12-15 12:05:05 +00:00
|
|
|
|
/// Extracts host from given url.
|
2016-12-15 12:33:50 +00:00
|
|
|
|
inline StringView getURLHost(const StringView & url)
|
2016-12-15 12:05:05 +00:00
|
|
|
|
{
|
2016-12-15 12:33:50 +00:00
|
|
|
|
StringView scheme = getURLScheme(url);
|
2016-12-15 18:59:07 +00:00
|
|
|
|
const char * p = url.data() + scheme.size();
|
|
|
|
|
const char * end = url.data() + url.size();
|
2016-12-15 12:05:05 +00:00
|
|
|
|
|
|
|
|
|
// Colon must follows after scheme.
|
|
|
|
|
if (p == end || *p != ':')
|
|
|
|
|
return StringView();
|
|
|
|
|
// Authority component must starts with "//".
|
|
|
|
|
if (end - p < 2 || (p[1] != '/' || p[2] != '/'))
|
|
|
|
|
return StringView();
|
|
|
|
|
else
|
|
|
|
|
p += 3;
|
|
|
|
|
|
2016-12-15 18:59:07 +00:00
|
|
|
|
const char * st = p;
|
2016-12-15 12:05:05 +00:00
|
|
|
|
|
|
|
|
|
for (; p < end; ++p)
|
|
|
|
|
{
|
|
|
|
|
if (*p == '@')
|
|
|
|
|
{
|
|
|
|
|
st = p + 1;
|
|
|
|
|
}
|
|
|
|
|
else if (*p == ':' || *p == '/' || *p == '?' || *p == '#')
|
|
|
|
|
{
|
|
|
|
|
break;
|
|
|
|
|
}
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|
2016-12-15 12:05:05 +00:00
|
|
|
|
|
|
|
|
|
return (p == st) ? StringView() : StringView(st, p - st);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
struct ExtractProtocol
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement();
|
|
|
|
|
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size);
|
2012-07-16 03:42:36 +00:00
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
template <bool without_www>
|
|
|
|
|
struct ExtractDomain
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 15; }
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
2016-12-15 12:33:50 +00:00
|
|
|
|
StringView host = getURLHost(StringView(data, size));
|
2016-12-09 19:31:16 +00:00
|
|
|
|
|
|
|
|
|
if (host.empty())
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
if (without_www && host.size() > 4 && !strncmp(host.data(), "www.", 4))
|
|
|
|
|
host = host.substr(4);
|
|
|
|
|
|
|
|
|
|
res_data = host.data();
|
|
|
|
|
res_size = host.size();
|
|
|
|
|
}
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
2014-10-27 15:16:11 +00:00
|
|
|
|
struct ExtractFirstSignificantSubdomain
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 10; }
|
|
|
|
|
|
|
|
|
|
static void execute(const Pos data, const size_t size, Pos & res_data, size_t & res_size, Pos * out_domain_end = nullptr)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos tmp;
|
|
|
|
|
size_t domain_length;
|
|
|
|
|
ExtractDomain<true>::execute(data, size, tmp, domain_length);
|
|
|
|
|
|
|
|
|
|
if (domain_length == 0)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
if (out_domain_end)
|
|
|
|
|
*out_domain_end = tmp + domain_length;
|
|
|
|
|
|
|
|
|
|
/// cut useless dot
|
|
|
|
|
if (tmp[domain_length - 1] == '.')
|
|
|
|
|
--domain_length;
|
|
|
|
|
|
|
|
|
|
res_data = tmp;
|
|
|
|
|
res_size = domain_length;
|
|
|
|
|
|
|
|
|
|
auto begin = tmp;
|
|
|
|
|
auto end = begin + domain_length;
|
|
|
|
|
const char * last_3_periods[3]{};
|
|
|
|
|
auto pos = static_cast<const char *>(memchr(begin, '.', domain_length));
|
|
|
|
|
|
|
|
|
|
while (pos)
|
|
|
|
|
{
|
|
|
|
|
last_3_periods[2] = last_3_periods[1];
|
|
|
|
|
last_3_periods[1] = last_3_periods[0];
|
|
|
|
|
last_3_periods[0] = pos;
|
|
|
|
|
pos = static_cast<const char *>(memchr(pos + 1, '.', end - pos - 1));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!last_3_periods[0])
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
if (!last_3_periods[1])
|
|
|
|
|
{
|
|
|
|
|
res_size = last_3_periods[0] - begin;
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!last_3_periods[2])
|
|
|
|
|
last_3_periods[2] = begin - 1;
|
|
|
|
|
|
2016-10-20 05:21:49 +00:00
|
|
|
|
if (!strncmp(last_3_periods[1] + 1, "com.", 4) /// Note that in ColumnString every value has zero byte after it.
|
|
|
|
|
|| !strncmp(last_3_periods[1] + 1, "net.", 4)
|
|
|
|
|
|| !strncmp(last_3_periods[1] + 1, "org.", 4)
|
|
|
|
|
|| !strncmp(last_3_periods[1] + 1, "co.", 3))
|
2014-10-27 15:16:11 +00:00
|
|
|
|
{
|
|
|
|
|
res_data += last_3_periods[2] + 1 - begin;
|
|
|
|
|
res_size = last_3_periods[1] - last_3_periods[2] - 1;
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
res_data += last_3_periods[1] + 1 - begin;
|
|
|
|
|
res_size = last_3_periods[0] - last_3_periods[1] - 1;
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
struct CutToFirstSignificantSubdomain
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 15; }
|
|
|
|
|
|
|
|
|
|
static void execute(const Pos data, const size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos tmp_data;
|
|
|
|
|
size_t tmp_length;
|
|
|
|
|
Pos domain_end;
|
|
|
|
|
ExtractFirstSignificantSubdomain::execute(data, size, tmp_data, tmp_length, &domain_end);
|
|
|
|
|
|
|
|
|
|
if (tmp_length == 0)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
res_data = tmp_data;
|
|
|
|
|
res_size = domain_end - tmp_data;
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
struct ExtractTopLevelDomain
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 5; }
|
|
|
|
|
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
2016-12-15 12:33:50 +00:00
|
|
|
|
StringView host = getURLHost(StringView(data, size));
|
2016-12-09 22:49:21 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
2016-12-09 22:49:21 +00:00
|
|
|
|
if (!host.empty())
|
|
|
|
|
{
|
|
|
|
|
if (host.back() == '.')
|
|
|
|
|
host = StringView(host.data(), host.size() - 1);
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2016-12-09 22:49:21 +00:00
|
|
|
|
Pos last_dot = reinterpret_cast<Pos>(memrchr(host.data(), '.', host.size()));
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2016-12-09 22:49:21 +00:00
|
|
|
|
if (!last_dot)
|
|
|
|
|
return;
|
|
|
|
|
/// Для IPv4-адресов не выделяем ничего.
|
|
|
|
|
if (last_dot[1] <= '9')
|
|
|
|
|
return;
|
2012-07-16 03:42:36 +00:00
|
|
|
|
|
2016-12-09 22:49:21 +00:00
|
|
|
|
res_data = last_dot + 1;
|
|
|
|
|
res_size = (host.data() + host.size()) - res_data;
|
|
|
|
|
}
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
struct ExtractPath
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 25; }
|
|
|
|
|
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
2014-04-08 07:47:51 +00:00
|
|
|
|
if (nullptr != (pos = strchr(data, '/')) && pos[1] == '/' && nullptr != (pos = strchr(pos + 2, '/')))
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
Pos query_string_or_fragment = strpbrk(pos, "?#");
|
|
|
|
|
|
|
|
|
|
res_data = pos;
|
|
|
|
|
res_size = (query_string_or_fragment ? query_string_or_fragment : end) - res_data;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
2014-12-05 13:31:48 +00:00
|
|
|
|
struct ExtractPathFull
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 30; }
|
|
|
|
|
|
|
|
|
|
static void execute(const Pos data, const size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
|
|
|
|
if (nullptr != (pos = strchr(data, '/')) && pos[1] == '/' && nullptr != (pos = strchr(pos + 2, '/')))
|
|
|
|
|
{
|
2014-12-19 14:17:02 +00:00
|
|
|
|
res_data = pos;
|
2014-12-05 13:31:48 +00:00
|
|
|
|
res_size = end - res_data;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
template <bool without_leading_char>
|
|
|
|
|
struct ExtractQueryString
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 10; }
|
|
|
|
|
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
2014-04-08 07:47:51 +00:00
|
|
|
|
if (nullptr != (pos = strchr(data, '?')))
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
Pos fragment = strchr(pos, '#');
|
|
|
|
|
|
|
|
|
|
res_data = pos + (without_leading_char ? 1 : 0);
|
|
|
|
|
res_size = (fragment ? fragment : end) - res_data;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
template <bool without_leading_char>
|
|
|
|
|
struct ExtractFragment
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 10; }
|
|
|
|
|
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
2014-04-08 07:47:51 +00:00
|
|
|
|
if (nullptr != (pos = strchr(data, '#')))
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
res_data = pos + (without_leading_char ? 1 : 0);
|
|
|
|
|
res_size = end - res_data;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
template <bool without_leading_char>
|
|
|
|
|
struct ExtractQueryStringAndFragment
|
|
|
|
|
{
|
|
|
|
|
static size_t getReserveLengthForElement() { return 20; }
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
2014-04-08 07:47:51 +00:00
|
|
|
|
if (nullptr != (pos = strchr(data, '?')))
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
res_data = pos + (without_leading_char ? 1 : 0);
|
|
|
|
|
res_size = end - res_data;
|
|
|
|
|
}
|
2014-11-12 17:23:26 +00:00
|
|
|
|
else if (nullptr != (pos = strchr(data, '#')))
|
|
|
|
|
{
|
|
|
|
|
res_data = pos;
|
|
|
|
|
res_size = end - res_data;
|
|
|
|
|
}
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
/// С точкой на конце.
|
|
|
|
|
struct ExtractWWW
|
|
|
|
|
{
|
|
|
|
|
static void execute(Pos data, size_t size, Pos & res_data, size_t & res_size)
|
|
|
|
|
{
|
|
|
|
|
res_data = data;
|
|
|
|
|
res_size = 0;
|
|
|
|
|
|
|
|
|
|
Pos pos = data;
|
|
|
|
|
Pos end = pos + size;
|
|
|
|
|
|
|
|
|
|
Pos tmp;
|
|
|
|
|
size_t protocol_length;
|
|
|
|
|
ExtractProtocol::execute(data, size, tmp, protocol_length);
|
|
|
|
|
pos += protocol_length + 3;
|
|
|
|
|
|
2016-11-21 02:39:37 +00:00
|
|
|
|
if (pos >= end || pos[-1] != '/' || pos[-2] != '/')
|
2012-07-16 03:42:36 +00:00
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
if (pos + 4 < end && !strncmp(pos, "www.", 4))
|
|
|
|
|
{
|
|
|
|
|
res_data = pos;
|
|
|
|
|
res_size = 4;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
struct ExtractURLParameterImpl
|
|
|
|
|
{
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector(const ColumnString::Chars_t & data,
|
2017-03-13 18:01:46 +00:00
|
|
|
|
const ColumnString::Offsets_t & offsets,
|
|
|
|
|
std::string pattern,
|
2013-09-15 05:51:43 +00:00
|
|
|
|
ColumnString::Chars_t & res_data, ColumnString::Offsets_t & res_offsets)
|
2013-03-05 12:12:47 +00:00
|
|
|
|
{
|
|
|
|
|
res_data.reserve(data.size() / 5);
|
|
|
|
|
res_offsets.resize(offsets.size());
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
pattern += '=';
|
|
|
|
|
const char * param_str = pattern.c_str();
|
|
|
|
|
size_t param_len = pattern.size();
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
size_t prev_offset = 0;
|
|
|
|
|
size_t res_offset = 0;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
for (size_t i = 0; i < offsets.size(); ++i)
|
|
|
|
|
{
|
|
|
|
|
size_t cur_offset = offsets[i];
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2015-07-24 19:23:22 +00:00
|
|
|
|
const char * str = reinterpret_cast<const char *>(&data[prev_offset]);
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2015-07-24 19:23:22 +00:00
|
|
|
|
const char * pos = nullptr;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
const char * begin = strpbrk(str, "?#");
|
2015-07-24 19:23:22 +00:00
|
|
|
|
if (begin != nullptr)
|
2013-03-05 12:12:47 +00:00
|
|
|
|
{
|
2015-07-24 19:23:22 +00:00
|
|
|
|
pos = begin + 1;
|
|
|
|
|
while (true)
|
2013-03-05 12:12:47 +00:00
|
|
|
|
{
|
2015-07-24 19:23:22 +00:00
|
|
|
|
pos = strstr(pos, param_str);
|
|
|
|
|
|
|
|
|
|
if (pos == nullptr)
|
|
|
|
|
break;
|
|
|
|
|
|
2016-01-26 21:24:09 +00:00
|
|
|
|
if (pos[-1] != '?' && pos[-1] != '#' && pos[-1] != '&')
|
2015-07-24 19:23:22 +00:00
|
|
|
|
{
|
|
|
|
|
pos += param_len;
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
pos += param_len;
|
|
|
|
|
break;
|
|
|
|
|
}
|
2013-03-05 12:12:47 +00:00
|
|
|
|
}
|
2015-07-24 19:23:22 +00:00
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos != nullptr)
|
2013-03-05 12:12:47 +00:00
|
|
|
|
{
|
2015-07-24 19:23:22 +00:00
|
|
|
|
const char * end = strpbrk(pos, "&#");
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (end == nullptr)
|
2013-03-05 12:12:47 +00:00
|
|
|
|
end = pos + strlen(pos);
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
res_data.resize(res_offset + (end - pos) + 1);
|
2016-04-15 00:33:21 +00:00
|
|
|
|
memcpySmallAllowReadWriteOverflow15(&res_data[res_offset], pos, end - pos);
|
2013-03-05 12:12:47 +00:00
|
|
|
|
res_offset += end - pos;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
res_data.resize(res_offset + 1);
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
res_data[res_offset] = 0;
|
|
|
|
|
++res_offset;
|
|
|
|
|
res_offsets[i] = res_offset;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 12:12:47 +00:00
|
|
|
|
prev_offset = cur_offset;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
struct CutURLParameterImpl
|
|
|
|
|
{
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector(const ColumnString::Chars_t & data,
|
2017-03-13 18:01:46 +00:00
|
|
|
|
const ColumnString::Offsets_t & offsets,
|
|
|
|
|
std::string pattern,
|
2013-09-15 05:51:43 +00:00
|
|
|
|
ColumnString::Chars_t & res_data, ColumnString::Offsets_t & res_offsets)
|
2013-03-18 10:27:45 +00:00
|
|
|
|
{
|
|
|
|
|
res_data.reserve(data.size());
|
|
|
|
|
res_offsets.resize(offsets.size());
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
pattern += '=';
|
|
|
|
|
const char * param_str = pattern.c_str();
|
|
|
|
|
size_t param_len = pattern.size();
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
size_t prev_offset = 0;
|
|
|
|
|
size_t res_offset = 0;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
for (size_t i = 0; i < offsets.size(); ++i)
|
|
|
|
|
{
|
|
|
|
|
size_t cur_offset = offsets[i];
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
const char * url_begin = reinterpret_cast<const char *>(&data[prev_offset]);
|
|
|
|
|
const char * url_end = reinterpret_cast<const char *>(&data[cur_offset]) - 1;
|
|
|
|
|
const char * begin_pos = url_begin;
|
|
|
|
|
const char * end_pos = begin_pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
do
|
|
|
|
|
{
|
2016-01-26 21:24:09 +00:00
|
|
|
|
const char * begin = strpbrk(url_begin, "?#");
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (begin == nullptr)
|
2013-03-18 10:27:45 +00:00
|
|
|
|
break;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
const char * pos = strstr(begin + 1, param_str);
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos == nullptr)
|
2013-03-18 10:27:45 +00:00
|
|
|
|
break;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
|
|
|
|
|
if (pos[-1] != '?' && pos[-1] != '#' && pos[-1] != '&')
|
2013-03-18 10:27:45 +00:00
|
|
|
|
{
|
2014-04-08 07:31:51 +00:00
|
|
|
|
pos = nullptr;
|
2013-03-18 10:27:45 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
begin_pos = pos;
|
|
|
|
|
end_pos = begin_pos + param_len;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
/// Пропустим значение.
|
2016-01-26 21:24:09 +00:00
|
|
|
|
while (*end_pos && *end_pos != '&' && *end_pos != '#')
|
2013-03-18 10:27:45 +00:00
|
|
|
|
++end_pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-01-26 21:24:09 +00:00
|
|
|
|
/// Захватим '&' до или после параметра.
|
|
|
|
|
if (*end_pos == '&')
|
2013-03-18 10:27:45 +00:00
|
|
|
|
++end_pos;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
else if (begin_pos[-1] == '&')
|
2013-03-18 10:27:45 +00:00
|
|
|
|
--begin_pos;
|
|
|
|
|
} while (false);
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:39:47 +00:00
|
|
|
|
size_t cut_length = (url_end - url_begin) - (end_pos - begin_pos);
|
|
|
|
|
res_data.resize(res_offset + cut_length + 1);
|
2016-04-15 00:33:21 +00:00
|
|
|
|
memcpySmallAllowReadWriteOverflow15(&res_data[res_offset], url_begin, begin_pos - url_begin);
|
|
|
|
|
memcpySmallAllowReadWriteOverflow15(&res_data[res_offset] + (begin_pos - url_begin), end_pos, url_end - end_pos);
|
2013-03-18 10:39:47 +00:00
|
|
|
|
res_offset += cut_length + 1;
|
2013-03-18 10:27:45 +00:00
|
|
|
|
res_data[res_offset - 1] = 0;
|
|
|
|
|
res_offsets[i] = res_offset;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-18 10:27:45 +00:00
|
|
|
|
prev_offset = cur_offset;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
class ExtractURLParametersImpl
|
|
|
|
|
{
|
|
|
|
|
private:
|
|
|
|
|
Pos pos;
|
|
|
|
|
Pos end;
|
|
|
|
|
bool first;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
public:
|
2014-11-12 17:23:26 +00:00
|
|
|
|
static constexpr auto name = "extractURLParameters";
|
|
|
|
|
static String getName() { return name; }
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-12-29 19:38:10 +00:00
|
|
|
|
static size_t getNumberOfArguments() { return 1; }
|
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
static void checkArguments(const DataTypes & arguments)
|
|
|
|
|
{
|
2016-05-23 00:40:28 +00:00
|
|
|
|
if (!typeid_cast<const DataTypeString *>(&*arguments[0]))
|
|
|
|
|
throw Exception("Illegal type " + arguments[0]->getName() + " of first argument of function " + getName() + ". Must be String.",
|
|
|
|
|
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
|
2013-03-05 13:30:23 +00:00
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
void init(Block & block, const ColumnNumbers & arguments) {}
|
2013-08-07 11:25:02 +00:00
|
|
|
|
|
|
|
|
|
/// Возвращает позицию аргумента, являющегося столбцом строк
|
|
|
|
|
size_t getStringsArgumentPosition()
|
|
|
|
|
{
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
/// Вызывается для каждой следующей строки.
|
|
|
|
|
void set(Pos pos_, Pos end_)
|
|
|
|
|
{
|
|
|
|
|
pos = pos_;
|
|
|
|
|
end = end_;
|
|
|
|
|
first = true;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
/// Получить следующий токен, если есть, или вернуть false.
|
|
|
|
|
bool get(Pos & token_begin, Pos & token_end)
|
2013-03-06 09:00:58 +00:00
|
|
|
|
{
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos == nullptr)
|
2013-03-06 09:00:58 +00:00
|
|
|
|
return false;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
if (first)
|
|
|
|
|
{
|
|
|
|
|
first = false;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
pos = strpbrk(pos, "?#");
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos == nullptr)
|
2013-03-06 09:00:58 +00:00
|
|
|
|
return false;
|
|
|
|
|
++pos;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-01-26 21:24:09 +00:00
|
|
|
|
while (true)
|
|
|
|
|
{
|
|
|
|
|
token_begin = pos;
|
|
|
|
|
pos = strpbrk(pos, "=&#?");
|
|
|
|
|
if (pos == nullptr)
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
if (*pos == '?')
|
|
|
|
|
{
|
|
|
|
|
++pos;
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (*pos == '&' || *pos == '#')
|
|
|
|
|
{
|
2013-03-06 11:22:17 +00:00
|
|
|
|
token_end = pos++;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
++pos;
|
|
|
|
|
pos = strpbrk(pos, "&#");
|
|
|
|
|
if (pos == nullptr)
|
|
|
|
|
token_end = end;
|
|
|
|
|
else
|
|
|
|
|
token_end = pos++;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
2013-08-02 13:55:43 +00:00
|
|
|
|
class ExtractURLParameterNamesImpl
|
|
|
|
|
{
|
|
|
|
|
private:
|
|
|
|
|
Pos pos;
|
|
|
|
|
Pos end;
|
|
|
|
|
bool first;
|
|
|
|
|
|
|
|
|
|
public:
|
2014-11-12 17:23:26 +00:00
|
|
|
|
static constexpr auto name = "extractURLParameterNames";
|
|
|
|
|
static String getName() { return name; }
|
2013-08-02 13:55:43 +00:00
|
|
|
|
|
2016-12-29 19:38:10 +00:00
|
|
|
|
static size_t getNumberOfArguments() { return 1; }
|
|
|
|
|
|
2013-08-02 13:55:43 +00:00
|
|
|
|
static void checkArguments(const DataTypes & arguments)
|
|
|
|
|
{
|
2016-05-23 00:40:28 +00:00
|
|
|
|
if (!typeid_cast<const DataTypeString *>(&*arguments[0]))
|
|
|
|
|
throw Exception("Illegal type " + arguments[0]->getName() + " of first argument of function " + getName() + ". Must be String.",
|
|
|
|
|
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
|
2013-08-02 13:55:43 +00:00
|
|
|
|
}
|
|
|
|
|
|
2013-08-07 11:25:02 +00:00
|
|
|
|
/// Возвращает позицию аргумента, являющегося столбцом строк
|
|
|
|
|
size_t getStringsArgumentPosition()
|
|
|
|
|
{
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2013-08-02 13:55:43 +00:00
|
|
|
|
void init(Block & block, const ColumnNumbers & arguments) {}
|
|
|
|
|
|
|
|
|
|
/// Вызывается для каждой следующей строки.
|
|
|
|
|
void set(Pos pos_, Pos end_)
|
|
|
|
|
{
|
|
|
|
|
pos = pos_;
|
|
|
|
|
end = end_;
|
|
|
|
|
first = true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Получить следующий токен, если есть, или вернуть false.
|
|
|
|
|
bool get(Pos & token_begin, Pos & token_end)
|
|
|
|
|
{
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos == nullptr)
|
2013-08-02 13:55:43 +00:00
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
if (first)
|
|
|
|
|
{
|
|
|
|
|
first = false;
|
2016-01-26 21:24:09 +00:00
|
|
|
|
pos = strpbrk(pos, "?#");
|
2013-08-02 13:55:43 +00:00
|
|
|
|
}
|
2013-08-05 08:40:56 +00:00
|
|
|
|
else
|
2016-01-26 21:24:09 +00:00
|
|
|
|
pos = strpbrk(pos, "&#");
|
2013-08-05 08:40:56 +00:00
|
|
|
|
|
2014-04-08 07:31:51 +00:00
|
|
|
|
if (pos == nullptr)
|
2013-08-05 08:40:56 +00:00
|
|
|
|
return false;
|
2013-08-02 13:55:43 +00:00
|
|
|
|
++pos;
|
|
|
|
|
|
2016-01-26 21:24:09 +00:00
|
|
|
|
while (true)
|
|
|
|
|
{
|
|
|
|
|
token_begin = pos;
|
2013-08-02 13:55:43 +00:00
|
|
|
|
|
2016-01-26 21:24:09 +00:00
|
|
|
|
pos = strpbrk(pos, "=&#?");
|
|
|
|
|
if (pos == nullptr)
|
|
|
|
|
return false;
|
|
|
|
|
else
|
|
|
|
|
token_end = pos;
|
|
|
|
|
|
|
|
|
|
if (*pos == '?')
|
|
|
|
|
{
|
|
|
|
|
++pos;
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
}
|
2013-08-02 13:55:43 +00:00
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
};
|
2013-03-06 09:00:58 +00:00
|
|
|
|
|
|
|
|
|
class URLHierarchyImpl
|
|
|
|
|
{
|
|
|
|
|
private:
|
2013-03-06 11:22:17 +00:00
|
|
|
|
Pos begin;
|
2013-03-06 09:00:58 +00:00
|
|
|
|
Pos pos;
|
|
|
|
|
Pos end;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
public:
|
2014-11-12 17:23:26 +00:00
|
|
|
|
static constexpr auto name = "URLHierarchy";
|
|
|
|
|
static String getName() { return name; }
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2016-12-29 19:38:10 +00:00
|
|
|
|
static size_t getNumberOfArguments() { return 1; }
|
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
static void checkArguments(const DataTypes & arguments)
|
|
|
|
|
{
|
2016-05-23 00:40:28 +00:00
|
|
|
|
if (!typeid_cast<const DataTypeString *>(&*arguments[0]))
|
|
|
|
|
throw Exception("Illegal type " + arguments[0]->getName() + " of first argument of function " + getName() + ". Must be String.",
|
|
|
|
|
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
|
2013-03-06 09:00:58 +00:00
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
void init(Block & block, const ColumnNumbers & arguments) {}
|
2013-08-07 11:25:02 +00:00
|
|
|
|
|
|
|
|
|
/// Возвращает позицию аргумента, являющегося столбцом строк
|
|
|
|
|
size_t getStringsArgumentPosition()
|
|
|
|
|
{
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
/// Вызывается для каждой следующей строки.
|
|
|
|
|
void set(Pos pos_, Pos end_)
|
|
|
|
|
{
|
2013-03-06 11:22:17 +00:00
|
|
|
|
begin = pos = pos_;
|
2013-03-06 09:00:58 +00:00
|
|
|
|
end = end_;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 09:00:58 +00:00
|
|
|
|
/// Получить следующий токен, если есть, или вернуть false.
|
|
|
|
|
bool get(Pos & token_begin, Pos & token_end)
|
2013-03-05 13:30:23 +00:00
|
|
|
|
{
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/// Код из URLParser.
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
if (pos == end)
|
2013-03-05 13:30:23 +00:00
|
|
|
|
return false;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
if (pos == begin)
|
2013-03-05 13:30:23 +00:00
|
|
|
|
{
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/// Распарсим всё, что идёт до пути
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/// Предположим, что протокол уже переведён в нижний регистр.
|
|
|
|
|
while (pos < end && ((*pos > 'a' && *pos < 'z') || (*pos > '0' && *pos < '9')))
|
|
|
|
|
++pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/** Будем вычислять иерархию только для URL-ов, в которых есть протокол, и после него идут два слеша.
|
|
|
|
|
* (http, file - подходят, mailto, magnet - не подходят), и после двух слешей ещё хоть что-нибудь есть
|
|
|
|
|
* Для остальных просто вернём полный URL как единственный элемент иерархии.
|
|
|
|
|
*/
|
|
|
|
|
if (pos == begin || pos == end || !(*pos++ == ':' && pos < end && *pos++ == '/' && pos < end && *pos++ == '/' && pos < end))
|
|
|
|
|
{
|
|
|
|
|
pos = end;
|
|
|
|
|
token_begin = begin;
|
|
|
|
|
token_end = end;
|
|
|
|
|
return true;
|
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/// Доменом для простоты будем считать всё, что после протокола и двух слешей, до следующего слеша или до ? или до #
|
|
|
|
|
while (pos < end && !(*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
if (pos != end)
|
|
|
|
|
++pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
token_begin = begin;
|
|
|
|
|
token_end = pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
return true;
|
2013-03-05 13:30:23 +00:00
|
|
|
|
}
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
/// Идём до следующего / или ? или #, пропуская все те, что вначале.
|
|
|
|
|
while (pos < end && (*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
|
|
|
|
if (pos == end)
|
2013-03-05 13:30:23 +00:00
|
|
|
|
return false;
|
2013-03-06 11:22:17 +00:00
|
|
|
|
while (pos < end && !(*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
if (pos != end)
|
|
|
|
|
++pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-06 11:22:17 +00:00
|
|
|
|
token_begin = begin;
|
|
|
|
|
token_end = pos;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2013-03-05 13:30:23 +00:00
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2014-02-11 19:18:38 +00:00
|
|
|
|
class URLPathHierarchyImpl
|
|
|
|
|
{
|
|
|
|
|
private:
|
|
|
|
|
Pos begin;
|
|
|
|
|
Pos pos;
|
|
|
|
|
Pos end;
|
|
|
|
|
Pos start;
|
|
|
|
|
|
|
|
|
|
public:
|
2014-11-12 17:23:26 +00:00
|
|
|
|
static constexpr auto name = "URLPathHierarchy";
|
|
|
|
|
static String getName() { return name; }
|
2014-02-11 19:18:38 +00:00
|
|
|
|
|
2016-12-29 19:38:10 +00:00
|
|
|
|
static size_t getNumberOfArguments() { return 1; }
|
|
|
|
|
|
2014-02-11 19:18:38 +00:00
|
|
|
|
static void checkArguments(const DataTypes & arguments)
|
|
|
|
|
{
|
2016-05-23 00:40:28 +00:00
|
|
|
|
if (!typeid_cast<const DataTypeString *>(&*arguments[0]))
|
|
|
|
|
throw Exception("Illegal type " + arguments[0]->getName() + " of first argument of function " + getName() + ". Must be String.",
|
|
|
|
|
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
|
2014-02-11 19:18:38 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void init(Block & block, const ColumnNumbers & arguments) {}
|
|
|
|
|
|
|
|
|
|
/// Возвращает позицию аргумента, являющегося столбцом строк
|
|
|
|
|
size_t getStringsArgumentPosition()
|
|
|
|
|
{
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Вызывается для каждой следующей строки.
|
|
|
|
|
void set(Pos pos_, Pos end_)
|
|
|
|
|
{
|
|
|
|
|
begin = pos = pos_;
|
|
|
|
|
start = begin;
|
|
|
|
|
end = end_;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Получить следующий токен, если есть, или вернуть false.
|
|
|
|
|
bool get(Pos & token_begin, Pos & token_end)
|
|
|
|
|
{
|
|
|
|
|
/// Код из URLParser.
|
|
|
|
|
|
|
|
|
|
if (pos == end)
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
if (pos == begin)
|
|
|
|
|
{
|
|
|
|
|
/// Распарсим всё, что идёт до пути
|
|
|
|
|
|
|
|
|
|
/// Предположим, что протокол уже переведён в нижний регистр.
|
|
|
|
|
while (pos < end && ((*pos > 'a' && *pos < 'z') || (*pos > '0' && *pos < '9')))
|
|
|
|
|
++pos;
|
|
|
|
|
|
|
|
|
|
/** Будем вычислять иерархию только для URL-ов, в которых есть протокол, и после него идут два слеша.
|
|
|
|
|
* (http, file - подходят, mailto, magnet - не подходят), и после двух слешей ещё хоть что-нибудь есть
|
|
|
|
|
* Для остальных просто вернём пустой массив.
|
|
|
|
|
*/
|
|
|
|
|
if (pos == begin || pos == end || !(*pos++ == ':' && pos < end && *pos++ == '/' && pos < end && *pos++ == '/' && pos < end))
|
|
|
|
|
{
|
|
|
|
|
pos = end;
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Доменом для простоты будем считать всё, что после протокола и двух слешей, до следующего слеша или до ? или до #
|
|
|
|
|
while (pos < end && !(*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
|
|
|
|
|
|
|
|
|
start = pos;
|
|
|
|
|
|
|
|
|
|
if (pos != end)
|
|
|
|
|
++pos;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/// Идём до следующего / или ? или #, пропуская все те, что вначале.
|
|
|
|
|
while (pos < end && (*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
|
|
|
|
if (pos == end)
|
|
|
|
|
return false;
|
|
|
|
|
while (pos < end && !(*pos == '/' || *pos == '?' || *pos == '#'))
|
|
|
|
|
++pos;
|
|
|
|
|
|
|
|
|
|
if (pos != end)
|
|
|
|
|
++pos;
|
|
|
|
|
|
|
|
|
|
token_begin = start;
|
|
|
|
|
token_end = pos;
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
/** Выделить кусок строки, используя Extractor.
|
|
|
|
|
*/
|
|
|
|
|
template <typename Extractor>
|
|
|
|
|
struct ExtractSubstringImpl
|
|
|
|
|
{
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector(const ColumnString::Chars_t & data, const ColumnString::Offsets_t & offsets,
|
|
|
|
|
ColumnString::Chars_t & res_data, ColumnString::Offsets_t & res_offsets)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
size_t size = offsets.size();
|
|
|
|
|
res_offsets.resize(size);
|
2013-04-17 12:52:55 +00:00
|
|
|
|
res_data.reserve(size * Extractor::getReserveLengthForElement());
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
size_t prev_offset = 0;
|
|
|
|
|
size_t res_offset = 0;
|
|
|
|
|
|
|
|
|
|
/// Выделенный кусок.
|
|
|
|
|
Pos start;
|
|
|
|
|
size_t length;
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
for (size_t i = 0; i < size; ++i)
|
|
|
|
|
{
|
|
|
|
|
Extractor::execute(reinterpret_cast<const char *>(&data[prev_offset]), offsets[i] - prev_offset - 1, start, length);
|
2014-06-26 00:58:14 +00:00
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
res_data.resize(res_data.size() + length + 1);
|
2016-04-15 00:33:21 +00:00
|
|
|
|
memcpySmallAllowReadWriteOverflow15(&res_data[res_offset], start, length);
|
2012-07-16 03:42:36 +00:00
|
|
|
|
res_offset += length + 1;
|
|
|
|
|
res_data[res_offset - 1] = 0;
|
|
|
|
|
|
|
|
|
|
res_offsets[i] = res_offset;
|
|
|
|
|
prev_offset = offsets[i];
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void constant(const std::string & data,
|
|
|
|
|
std::string & res_data)
|
|
|
|
|
{
|
|
|
|
|
Pos start;
|
|
|
|
|
size_t length;
|
|
|
|
|
Extractor::execute(data.data(), data.size(), start, length);
|
|
|
|
|
res_data.assign(start, length);
|
|
|
|
|
}
|
|
|
|
|
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector_fixed(const ColumnString::Chars_t & data, size_t n,
|
|
|
|
|
ColumnString::Chars_t & res_data)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
throw Exception("Column of type FixedString is not supported by URL functions", ErrorCodes::ILLEGAL_COLUMN);
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/** Удалить кусок строки, используя Extractor.
|
|
|
|
|
*/
|
|
|
|
|
template <typename Extractor>
|
|
|
|
|
struct CutSubstringImpl
|
|
|
|
|
{
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector(const ColumnString::Chars_t & data, const ColumnString::Offsets_t & offsets,
|
|
|
|
|
ColumnString::Chars_t & res_data, ColumnString::Offsets_t & res_offsets)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
res_data.reserve(data.size());
|
|
|
|
|
size_t size = offsets.size();
|
|
|
|
|
res_offsets.resize(size);
|
|
|
|
|
|
|
|
|
|
size_t prev_offset = 0;
|
|
|
|
|
size_t res_offset = 0;
|
|
|
|
|
|
|
|
|
|
/// Выделенный кусок.
|
|
|
|
|
Pos start;
|
|
|
|
|
size_t length;
|
|
|
|
|
|
|
|
|
|
for (size_t i = 0; i < size; ++i)
|
|
|
|
|
{
|
|
|
|
|
const char * current = reinterpret_cast<const char *>(&data[prev_offset]);
|
|
|
|
|
Extractor::execute(current, offsets[i] - prev_offset - 1, start, length);
|
|
|
|
|
size_t start_index = start - reinterpret_cast<const char *>(&data[0]);
|
|
|
|
|
|
|
|
|
|
res_data.resize(res_data.size() + offsets[i] - prev_offset - length);
|
2016-04-15 00:33:21 +00:00
|
|
|
|
memcpySmallAllowReadWriteOverflow15(
|
|
|
|
|
&res_data[res_offset], current, start - current);
|
|
|
|
|
memcpySmallAllowReadWriteOverflow15(
|
|
|
|
|
&res_data[res_offset + start - current], start + length, offsets[i] - start_index - length);
|
2012-07-16 03:42:36 +00:00
|
|
|
|
res_offset += offsets[i] - prev_offset - length;
|
|
|
|
|
|
|
|
|
|
res_offsets[i] = res_offset;
|
|
|
|
|
prev_offset = offsets[i];
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void constant(const std::string & data,
|
|
|
|
|
std::string & res_data)
|
|
|
|
|
{
|
|
|
|
|
Pos start;
|
|
|
|
|
size_t length;
|
|
|
|
|
Extractor::execute(data.data(), data.size(), start, length);
|
2014-12-05 11:44:37 +00:00
|
|
|
|
res_data.reserve(data.size() - length);
|
|
|
|
|
res_data.append(data.data(), start);
|
|
|
|
|
res_data.append(start + length, data.data() + data.size());
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|
|
|
|
|
|
2013-09-15 05:51:43 +00:00
|
|
|
|
static void vector_fixed(const ColumnString::Chars_t & data, size_t n,
|
|
|
|
|
ColumnString::Chars_t & res_data)
|
2012-07-16 03:42:36 +00:00
|
|
|
|
{
|
|
|
|
|
throw Exception("Column of type FixedString is not supported by URL functions", ErrorCodes::ILLEGAL_COLUMN);
|
|
|
|
|
}
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
2016-12-15 12:05:05 +00:00
|
|
|
|
/// Percent decode of url data.
|
2016-12-12 06:09:00 +00:00
|
|
|
|
struct DecodeURLComponentImpl
|
2016-12-10 21:04:58 +00:00
|
|
|
|
{
|
|
|
|
|
static void vector(const ColumnString::Chars_t & data, const ColumnString::Offsets_t & offsets,
|
2016-12-15 12:05:05 +00:00
|
|
|
|
ColumnString::Chars_t & res_data, ColumnString::Offsets_t & res_offsets);
|
2016-12-10 21:04:58 +00:00
|
|
|
|
|
|
|
|
|
static void constant(const std::string & data,
|
2016-12-15 12:05:05 +00:00
|
|
|
|
std::string & res_data);
|
2016-12-10 21:04:58 +00:00
|
|
|
|
|
|
|
|
|
static void vector_fixed(const ColumnString::Chars_t & data, size_t n,
|
2016-12-15 12:05:05 +00:00
|
|
|
|
ColumnString::Chars_t & res_data);
|
2016-12-10 21:04:58 +00:00
|
|
|
|
};
|
|
|
|
|
|
2012-07-16 03:42:36 +00:00
|
|
|
|
}
|