eliott/ladybird
1
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird synced 2026-04-28 10:37:17 +02:00
Code Issues Packages Projects Releases Wiki Activity

LibRegex: Support UTF-16 RegexStringView and improve Unicode matching

Browse Source 
When the Unicode option is not set, regular expressions should match
based on code units; when it is set, they should match based on code
points. To do so, the regex parser must combine surrogate pairs when
the Unicode option is set. Further, RegexStringView needs to know if
the flag is set in order to return code point vs. code unit based
string lengths and substrings.
This commit is contained in:
Timothy Flynn
2021-07-20 22:33:00 -04:00
committed by Linus Groh
parent 2e45e52993
commit 47f6bb38a1
Notes: sideshowbarker 2024-07-18 08:26:31 +09:00 
Author: https://github.com/trflynn89
Commit: https://github.com/SerenityOS/serenity/commit/47f6bb38a1b
Pull-request: https://github.com/SerenityOS/serenity/pull/8931
Reviewed-by: https://github.com/linusg ✅
5 changed files with 167 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
Userland/Libraries/LibRegex/RegexParser.cpp
View File

@@ -10,6 +10,7 @@
#include <AK/String.h>
#include <AK/StringBuilder.h>
#include <AK/StringUtils.h>
#include <AK/Utf16View.h>
namespace regex {
@@ -1440,13 +1441,31 @@ bool ECMA262Parser::parse_atom_escape(ByteCode& stack, size_t& match_length_mini
if (try_skip("u")) {
if (auto code_point = read_digits(ReadDigitsInitialZeroState::Allow, true, 4); code_point.has_value()) {
// FIXME: The minimum length depends on the mode - should be utf8-length in u8 mode.
// In Unicode mode, we need to combine surrogate pairs into a single code point. But we also need to be
// rather forgiving if the surrogate pairs are invalid. So if a second code unit follows this code unit,
// but doesn't form a valid surrogate pair, insert bytecode for both code units individually.
Optional<u32> low_surrogate;
if (unicode && Utf16View::is_high_surrogate(*code_point) && try_skip("\\u")) {
low_surrogate = read_digits(ReadDigitsInitialZeroState::Allow, true, 4);
if (!low_surrogate.has_value()) {
set_error(Error::InvalidPattern);
return false;
}
if (Utf16View::is_low_surrogate(*low_surrogate)) {
*code_point = Utf16View::decode_surrogate_pair(*code_point, *low_surrogate);
low_surrogate.clear();
}
}
match_length_minimum += 1;
StringBuilder builder;
builder.append_code_point(code_point.value());
// FIXME: This isn't actually correct for ECMAScript.
auto u8_encoded = builder.string_view();
stack.insert_bytecode_compare_string(u8_encoded);
stack.insert_bytecode_compare_values({ { CharacterCompareType::Char, (ByteCodeValueType)code_point.value() } });
if (low_surrogate.has_value()) {
match_length_minimum += 1;
stack.insert_bytecode_compare_values({ { CharacterCompareType::Char, (ByteCodeValueType)low_surrogate.value() } });
}
return true;
} else if (!unicode) {
// '\u' is allowed in non-unicode mode, just matches 'u'.