happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

common : use regex_search with anchoring for partial matching

This commit is contained in:
Alde Rojas 2025-12-24 01:39:41 -06:00
parent 6b757458da
commit 2a7bf2ede2
1 changed files with 13 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
common/regex-partial.cpp
View File

@ -27,7 +27,7 @@ common_regex_match common_regex::search(const std::string & input, size_t pos, b
return res;
}
std::match_results<std::string::const_reverse_iterator> srmatch;
if (std::regex_match(input.rbegin(), input.rend() - pos, srmatch, rx_reversed_partial)) {
if (std::regex_search(input.rbegin(), input.rend() - pos, srmatch, rx_reversed_partial, std::regex_constants::match_continuous)) {
auto group = srmatch[1].str();
if (group.length() != 0) {
auto it = srmatch[1].second.base();
@ -55,18 +55,18 @@ common_regex_match common_regex::search(const std::string & input, size_t pos, b
to see if a string ends with a partial regex match, but but it's not in std::regex yet.
Instead, we'll the regex into a partial match regex operating as a full match on the reverse iterators of the input.
- /abcd/ -> (dcba|cba|ba|a).* -> ((?:(?:(?:(?:d)?c)?b)?a).*
- /a|b/ -> (a|b).*
- /abcd/ -> ^(dcba|cba|ba|a) -> ^((?:(?:(?:(?:d)?c)?b)?a)
- /a|b/ -> ^(a|b)
- /a*?/ -> error, could match ""
- /a*b/ -> ((?:b)?a*+).* (final repetitions become eager)
- /.*?ab/ -> ((?:b)?a).* (merge .*)
- /a.*?b/ -> ((?:b)?.*?a).* (keep reluctant matches)
- /a(bc)d/ -> ((?:(?:d)?(?:(?:c)?b))?a).*
- /a(bc|de)/ -> ((?:(?:(?:e)?d)?|(?:(?:c)?b)?)?a).*
- /ab{2,4}c/ -> abbb?b?c -> ((?:(?:(?:(?:(?:c)?b)?b)?b?)?b?)?a).*
- /a*b/ -> ^((?:b)?a*+) (final repetitions become eager)
- /.*?ab/ -> ^((?:b)?a) (omit .*)
- /a.*?b/ -> ^((?:b)?.*?a) (keep reluctant matches)
- /a(bc)d/ -> ^((?:(?:d)?(?:(?:c)?b))?a)
- /a(bc|de)/ -> ^((?:(?:(?:e)?d)?|(?:(?:c)?b)?)?a)
- /ab{2,4}c/ -> ^cbbb?b?a -> ^((?:(?:(?:(?:(?:c)?b)?b)?b?)?b?)?a)
The regex will match a reversed string fully, and the end of the first (And only) capturing group will indicate the reversed start of the original partial pattern
(i.e. just where the final .* starts in the inverted pattern; all other groups are turned into non-capturing groups, and reluctant quantifiers are ignored)
The regex will match a reversed string fully, and the end of the first (And only) capturing group will indicate the reversed start of the original partial pattern.
All other groups are turned into non-capturing groups, and reluctant quantifiers are ignored.
*/
std::string regex_to_reversed_partial_regex(const std::string & pattern) {
auto it = pattern.begin();
@ -177,7 +177,7 @@ std::string regex_to_reversed_partial_regex(const std::string & pattern) {
}
}
// /abcd/ -> (dcba|cba|ba|a).* -> ((?:(?:(?:d)?c)?b)?a).*
// /abcd/ -> ^(dcba|cba|ba|a) -> ^((?:(?:(?:d)?c)?b)?a)
// if n(=4) parts, opening n-1(=3) non-capturing groups after the 1 capturing group
// We'll do the outermost capturing group and final .* in the enclosing function.
std::vector<std::string> res_alts;
@ -200,5 +200,5 @@ std::string regex_to_reversed_partial_regex(const std::string & pattern) {
throw std::runtime_error("Unmatched '(' in pattern");
}
return "(" + res + ")[\\s\\S]*";
return "^(" + res + ")";
}