RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-26 08:27:45 +00:00
Code Issues Activity Actions

LibRegex: Make parse_disjunction() consume all disjunctions in one frame

Browse source 
This helps us not blow up when too many disjunctions are chained togther
in the regex we're parsing.
Fixes #12615.
This commit is contained in:
Ali Mohammad Pur 2022-02-19 17:34:55 +03:30 • committed by Andreas Kling
parent 627bbee055
commit 4be7239626
2 changed files with 33 additions and 21 deletions
Download patch file Download diff file Expand all files Collapse all files

7
Tests/LibRegex/Regex.cpp
View file

@ -498,6 +498,8 @@ TEST_CASE(posix_extended_nested_capture_group)
EXPECT_EQ(result.capture_group_matches[0][2].view, "llo"sv); EXPECT_EQ(result.capture_group_matches[0][2].view, "llo"sv);
} }
auto parse_test_case_long_disjunction_chain = String::repeated("a|"sv, 10000);
TEST_CASE(ECMA262_parse) TEST_CASE(ECMA262_parse)
{ {
struct _test { struct _test {
@ -506,7 +508,7 @@ TEST_CASE(ECMA262_parse)
regex::ECMAScriptFlags flags {}; regex::ECMAScriptFlags flags {};
}; };
constexpr _test tests[] { _test const tests[] {
{ "^hello.$"sv }, { "^hello.$"sv },
{ "^(hello.)$"sv }, { "^(hello.)$"sv },
{ "^h{0,1}ello.$"sv }, { "^h{0,1}ello.$"sv },
@ -599,7 +601,8 @@ TEST_CASE(ECMA262_parse)
{ "(?<$$_$$>a)"sv }, { "(?<$$_$$>a)"sv },
{ "(?<ÿ>a)"sv }, { "(?<ÿ>a)"sv },
{ "(?<𝓑𝓻𝓸𝔀𝓷>a)"sv }, { "(?<𝓑𝓻𝓸𝔀𝓷>a)"sv },
{ "((?=lg)?[vl]k\\-?\\d{3}) bui| 3\\.[-\\w; ]{10}lg?-([06cv9]{3,4})"sv, regex::Error::NoError, ECMAScriptFlags::BrowserExtended } // #12373, quantifiable assertions. { "((?=lg)?[vl]k\\-?\\d{3}) bui| 3\\.[-\\w; ]{10}lg?-([06cv9]{3,4})"sv, regex::Error::NoError, ECMAScriptFlags::BrowserExtended }, // #12373, quantifiable assertions.
{ parse_test_case_long_disjunction_chain.view() }, // A whole lot of disjunctions, should not overflow the stack.
}; };
for (auto& test : tests) { for (auto& test : tests) {

47
Userland/Libraries/LibRegex/RegexParser.cpp
View file

@ -956,28 +956,37 @@ bool ECMA262Parser::parse_pattern(ByteCode& stack, size_t& match_length_minimum,
bool ECMA262Parser::parse_disjunction(ByteCode& stack, size_t& match_length_minimum, bool unicode, bool named) bool ECMA262Parser::parse_disjunction(ByteCode& stack, size_t& match_length_minimum, bool unicode, bool named)
{ {
ByteCode left_alternative_stack; size_t total_match_length_minimum = NumericLimits<size_t>::max();
size_t left_alternative_min_length = 0; Vector<ByteCode> alternatives;
auto alt_ok = parse_alternative(left_alternative_stack, left_alternative_min_length, unicode, named); do {
if (!alt_ok) ByteCode alternative_stack;
return false; size_t alternative_minimum_length = 0;
auto alt_ok = parse_alternative(alternative_stack, alternative_minimum_length, unicode, named);
if (!alt_ok)
return false;
if (!match(TokenType::Pipe)) { alternatives.append(move(alternative_stack));
stack.extend(left_alternative_stack); total_match_length_minimum = min(alternative_minimum_length, total_match_length_minimum);
match_length_minimum = left_alternative_min_length;
return alt_ok; if (!match(TokenType::Pipe))
break;
consume();
} while (true);
Optional<ByteCode> alternative_stack {};
for (auto& alternative : alternatives) {
if (alternative_stack.has_value()) {
ByteCode target_stack;
target_stack.insert_bytecode_alternation(alternative_stack.release_value(), move(alternative));
alternative_stack = move(target_stack);
} else {
alternative_stack = move(alternative);
}
} }
consume(); stack.extend(alternative_stack.release_value());
ByteCode right_alternative_stack; match_length_minimum = total_match_length_minimum;
size_t right_alternative_min_length = 0; return true;
auto continuation_ok = parse_disjunction(right_alternative_stack, right_alternative_min_length, unicode, named);
if (!continuation_ok)
return false;
stack.insert_bytecode_alternation(move(left_alternative_stack), move(right_alternative_stack));
match_length_minimum = min(left_alternative_min_length, right_alternative_min_length);
return continuation_ok;
} }
bool ECMA262Parser::parse_alternative(ByteCode& stack, size_t& match_length_minimum, bool unicode, bool named) bool ECMA262Parser::parse_alternative(ByteCode& stack, size_t& match_length_minimum, bool unicode, bool named)