Skip to content

Enable lookarounds to influence atomicity #118153

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
stephentoub merged 2 commits into from
Jul 29, 2025
Merged

Enable lookarounds to influence atomicity #118153

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
69d99ca
Enable lookarounds to influence atomicity
stephentoub Jul 29, 2025
c7f42f1
Update src/libraries/System.Text.RegularExpressions/tests/UnitTests/R…
stephentoub Jul 29, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions src/libraries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/RegexNode.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2326,11 +2326,17 @@ private static bool CanBeMadeAtomic(RegexNode node, RegexNode subsequent, bool i
return true;
}

// If this node is a {one/notone/set}loop, see if it overlaps with its successor in the concatenation.
// If it doesn't, then we can upgrade it to being a {one/notone/set}loopatomic.
// Doing so avoids unnecessary backtracking.
// If this node is a loop, see if it overlaps with its successor in the concatenation.
// If it doesn't, then we can upgrade it to being atomic to avoid unnecessary backtracking.
switch (node.Kind)
{
case RegexNodeKind when iterateNullableSubsequent && subsequent.Kind is RegexNodeKind.PositiveLookaround:
if (!CanBeMadeAtomic(node, subsequent.Child(0), iterateNullableSubsequent: false, allowLazy: allowLazy))
{
return false;
}
break;

case RegexNodeKind.Oneloop:
case RegexNodeKind.Onelazy when allowLazy:
switch (subsequent.Kind)
Expand Down
54 changes: 54 additions & 0 deletions src/libraries/System.Text.RegularExpressions/tests/FunctionalTests/Regex.Match.Tests.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,9 +93,55 @@ public static IEnumerable<object[]> Match_MemberData()
yield return (@"(?:(?!(b)b)\1a)*", "babababa", RegexOptions.None, 0, 8, true, string.Empty);
yield return (@"(.*?)a(?!(a+)b\2c)\2(.*)", "baaabaac", RegexOptions.None, 0, 8, false, string.Empty);
yield return (@"(?!(abc))+\w\w\w", "abcdef", RegexOptions.None, 0, 6, true, "bcd");
yield return (@"a+(?!c)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a+(?!c)", "aaac", RegexOptions.None, 0, 4, true, "aa");
yield return (@"a*(?!c)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a{2,5}(?!c)", "aaaaac", RegexOptions.None, 0, 6, true, "aaaa");
yield return (@"a+?(?!c)", "aaab", RegexOptions.None, 0, 4, true, "a");
yield return (@"a*?(?!c)", "aaab", RegexOptions.None, 0, 4, true, "");
yield return (@"a{2,5}?(?!c)", "aaaaab", RegexOptions.None, 0, 6, true, "aa");
yield return (@"[ab]*(?!x)", "ababc", RegexOptions.None, 0, 5, true, "abab");
yield return (@"a+(?=b)(?!c)", "aabx", RegexOptions.None, 0, 4, true, "aa");
yield return (@"a+?(?=b)(?!c)", "aabx", RegexOptions.None, 0, 4, true, "aa");

// Zero-width positive lookahead assertion
yield return (@"(?=(abc))?\1", "abc", RegexOptions.None, 0, 3, true, "abc");
yield return (@"(?=(abc))+\1", "abc", RegexOptions.None, 0, 3, true, "abc");
yield return (@"(?=(abc))*\1", "abc", RegexOptions.None, 0, 3, true, "abc");
yield return (@"^.*?(?=.)b", "ab", RegexOptions.None, 0, 2, true, "ab");
yield return (@".*?(?=.)b", "ab", RegexOptions.None, 0, 2, true, "ab");
yield return (@"^(?>.*?)(?=.)b", "ab", RegexOptions.None, 0, 2, false, "");
yield return (@"(?>.*?)(?=.)b", "ab", RegexOptions.None, 0, 2, true, "b");
yield return (@"a+(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a+(?=b)", "aaabc", RegexOptions.None, 0, 5, true, "aaa");
yield return (@"a*(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a{2,5}(?=b)", "aaaaab", RegexOptions.None, 0, 6, true, "aaaaa");
yield return (@"a+?(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a*?(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"a{2,5}?(?=b)", "aaaaab", RegexOptions.None, 0, 6, true, "aaaaa");
yield return (@"a+b+(?=c)", "aabbbc", RegexOptions.None, 0, 6, true, "aabbb");
yield return (@"a+?b+(?=c)", "aabbbc", RegexOptions.None, 0, 6, true, "aabbb");
yield return (@"a+b+?(?=c)", "aabbbc", RegexOptions.None, 0, 6, true, "aabbb");
yield return (@"[ab]+(?=c)", "ababc", RegexOptions.None, 0, 5, true, "abab");
yield return (@"[ab]+?(?=c)", "ababc", RegexOptions.None, 0, 5, true, "abab");
yield return (@"\w+(?=\b)", "hello world", RegexOptions.None, 0, 11, true, "hello");
yield return (@"\w+?(?=\b)", "hello world", RegexOptions.None, 0, 11, true, "hello");
yield return (@"(?>a+)(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(?>a*)(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(?>a{2,5})(?=b)", "aaaaab", RegexOptions.None, 0, 6, true, "aaaaa");
yield return (@"a*(?=a)", "aaa", RegexOptions.None, 0, 3, true, "aa");
yield return (@"a*?(?=a)", "aaa", RegexOptions.None, 0, 3, true, "");
yield return (@"a+(?=a*b)ab", "aaaab", RegexOptions.None, 0, 5, true, "aaaab");
yield return (@"a+?(?=a*b)ab", "aaaab", RegexOptions.None, 0, 5, true, "aaaab");
yield return (@"(a+)+(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(a+?)+(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(a+)+?(?=b)", "aaab", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(a+|b+)(?=c)", "aaac", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(a+?|b+?)(?=c)", "aaac", RegexOptions.None, 0, 4, true, "aaa");
yield return (@"(a+)(?=\1b)", "aaaaaab", RegexOptions.None, 0, 7, true, "aaa");
yield return (@"(a+?)(?=\1b)", "aaaaaab", RegexOptions.None, 0, 7, true, "aaa");
yield return (@"[A-Z]+(?=b)", "AAAb", RegexOptions.IgnoreCase, 0, 4, true, "AAA");
yield return (@"[A-Z]+?(?=b)", "AAAb", RegexOptions.IgnoreCase, 0, 4, true, "AAA");

// Zero-width positive lookbehind assertion
yield return (@"(\w){6}(?<=XXX)def", "abcXXXdef", RegexOptions.None, 0, 9, true, "abcXXXdef");
Expand Down Expand Up @@ -136,13 +182,21 @@ public static IEnumerable<object[]> Match_MemberData()
yield return (@"(?<=(abc)+?)", "123abc", RegexOptions.None, 0, 6, true, "");
yield return (@"(?<=(abc)+?)", "123ab", RegexOptions.None, 0, 5, false, "");
yield return (@"(?<=(abc)+?123)", "abcabc123", RegexOptions.None, 0, 9, true, "");
yield return (@"a+(?!c)(?<=y)", "yaab", RegexOptions.None, 0, 4, false, "");
yield return (@"(?<=a{2,4})b+", "aaabbb", RegexOptions.None, 0, 6, true, "bbb");
yield return (@"(?<=a+)b+?", "aaabbb", RegexOptions.None, 0, 6, true, "b");

// Zero-width negative lookbehind assertion: Actual - "(\\w){6}(?<!XXX)def"
yield return (@"(\w){6}(?<!XXX)def", "XXXabcdef", RegexOptions.None, 0, 9, true, "XXXabcdef");
yield return (@"123(?<!$) abcdef", "123 abcdef", RegexOptions.None, 0, 10, true, "123 abcdef");
yield return (@"(abc)\w(?<!(?(1)e|d))", "abcdabc", RegexOptions.None, 0, 7, true, "abcd");
yield return (@"(abc)\w(?<!(?(cd)e|d))", "abcdabc", RegexOptions.None, 0, 7, true, "abcd");
yield return (@"(?<!(b)a)\1", "bb", RegexOptions.None, 0, 2, false, string.Empty); // negative assertion should not capture
yield return (@"(?<=a)b+c", "abbbbc", RegexOptions.None, 0, 6, true, "bbbbc");
yield return (@"(?<=a+)bc", "aaabc", RegexOptions.None, 0, 5, true, "bc");
yield return (@"(?<!x)a+b", "yaab", RegexOptions.None, 0, 4, true, "aab");
yield return (@"(?<!x)a+b", "xaab", RegexOptions.None, 0, 4, true, "ab");
yield return (@"a+(?=b)(?<!x)", "yaab", RegexOptions.None, 0, 4, true, "aa");

// Nonbacktracking subexpression: Actual - "[^0-9]+(?>[0-9]+)3"
// The last 3 causes the match to fail, since the non backtracking subexpression does not give up the last digit it matched
Expand Down
5 changes: 5 additions & 0 deletions src/libraries/System.Text.RegularExpressions/tests/UnitTests/RegexReductionTests.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -288,6 +288,7 @@ public class RegexReductionTests
[InlineData(@"abc(?=\Z)", @"abc\Z")]
[InlineData(@"abc(?=\A)", @"abc\A")]
[InlineData(@"abc(?=$)", @"abc$")]
[InlineData(@"a*(?=b)bcd", @"(?>a*)(?=b)bcd")]
// Alternation reduction
[InlineData("a|b", "[ab]")]
[InlineData("a|b|c|d|e|g|h|z", "[a-eghz]")]
Expand Down Expand Up @@ -574,6 +575,10 @@ public void PatternsReduceIdentically(string actual, string expected)
[InlineData("(?=(abc))", "(?=abc)")]
[InlineData("(?=a(b*)c)", "(?=ab*c)")]
[InlineData("(?=a((((b))))c)", "(?=abc)")]
[InlineData(@"a*(?=a)", @"(?>a*)(?=a)")]
[InlineData(@"a*(?!b)b", @"(?>a*)(?!b)b")]
[InlineData(@"a*(?<!b)cde", @"(?>a*)(?<!b)cde")]
[InlineData(@"a*(?<=b)cde", @"(?>a*)(?<=b)cde")]
// Loops inside alternation constructs
[InlineData("(abc*|def)chi", "(ab(?>c*)|def)chi")]
[InlineData("(abc|def*)fhi", "(abc|de(?>f*))fhi")]
Expand Down
Loading