RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-25 19:37:35 +00:00
Code Issues Activity Actions

Revert "LibRegex: Implement an ECMA262 Regex quirk with negative loo..."

Browse source 
This partially reverts commit c11be92e23.
That commit fixes one thing and breaks many more, a next commit will
implement this quirk in a more sane way.
This commit is contained in:
Ali Mohammad Pur 2022-01-21 20:10:51 +03:30 • committed by Linus Groh
parent e0d3a8f6ce
commit 704e0654b3
2 changed files with 2 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

1
Tests/LibRegex/Regex.cpp
View file

@ -682,7 +682,6 @@ TEST_CASE(ECMA262_match)
{ "[\\0]"sv, "\0"sv, true, combine_flags(ECMAScriptFlags::Unicode, ECMAScriptFlags::BrowserExtended) }, { "[\\0]"sv, "\0"sv, true, combine_flags(ECMAScriptFlags::Unicode, ECMAScriptFlags::BrowserExtended) },
{ "[\\01]"sv, "\1"sv, true, ECMAScriptFlags::BrowserExtended }, { "[\\01]"sv, "\1"sv, true, ECMAScriptFlags::BrowserExtended },
{ "(\0|a)"sv, "a"sv, true }, // #9686, Should allow null bytes in pattern { "(\0|a)"sv, "a"sv, true }, // #9686, Should allow null bytes in pattern
{ "(.*?)a(?!(a+)b\\2c)\\2(.*)"sv, "baaabaac"sv, true }, // #6042, Groups inside lookarounds may be referenced outside, but their contents appear empty if the pattern in the lookaround fails.
{ "a|$"sv, "x"sv, true, (ECMAScriptFlags)regex::AllFlags::Global }, // #11940, Global (not the 'g' flag) regexps should attempt to match the zero-length end of the string too. { "a|$"sv, "x"sv, true, (ECMAScriptFlags)regex::AllFlags::Global }, // #11940, Global (not the 'g' flag) regexps should attempt to match the zero-length end of the string too.
}; };
// clang-format on // clang-format on

15
Userland/Libraries/LibRegex/RegexMatcher.cpp
View file

@ -489,14 +489,7 @@ Optional<bool> Matcher<Parser>::execute(MatchInput const& input, MatchState& sta
return true; return true;
case ExecutionResult::Failed: case ExecutionResult::Failed:
if (!states_to_try_next.is_empty()) { if (!states_to_try_next.is_empty()) {
auto next_state = states_to_try_next.take_last(); state = states_to_try_next.take_last();
// Note: ECMA262 quirk: Note 3, https://tc39.es/ecma262/#sec-runtime-semantics-canonicalize-ch
// capture groups defined in lookarounds "leak" outside the regex,
// but their contents are empty if the lookaround fails.
// This is done by manually clearing the groups where needed, and leaking their contents here.
if constexpr (IsSame<Parser, ECMA262>)
swap(next_state.capture_group_matches, state.capture_group_matches);
state = move(next_state);
continue; continue;
} }
return false; return false;
@ -506,11 +499,7 @@ Optional<bool> Matcher<Parser>::execute(MatchInput const& input, MatchState& sta
return {}; return {};
return false; return false;
} }
auto next_state = states_to_try_next.take_last(); state = states_to_try_next.take_last();
// See note above about an ECMA262 quirk.
if constexpr (IsSame<Parser, ECMA262>)
swap(next_state.capture_group_matches, state.capture_group_matches);
state = move(next_state);
++recursion_level; ++recursion_level;
continue; continue;
} }