hohn/codeql
1
0
Fork 0
mirror of https://github.com/github/codeql.git synced 2026-04-18 13:34:02 +02:00
Code Issues Packages Projects Releases Wiki Activity

Added support for character class union in regex processing

Browse Source
This commit is contained in:
Napalys
2025-03-03 08:26:48 +01:00
parent 1e05f327d6
commit fe6de2f672
6 changed files with 325 additions and 257 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
javascript/extractor/src/com/semmle/js/parser/RegExpParser.java
View File

@@ -9,6 +9,7 @@ import com.semmle.js.ast.regexp.CharacterClassEscape;
import com.semmle.js.ast.regexp.CharacterClassQuotedString;
import com.semmle.js.ast.regexp.CharacterClassRange;
import com.semmle.js.ast.regexp.CharacterClassSubtraction;
import com.semmle.js.ast.regexp.CharacterClassUnion;
import com.semmle.js.ast.regexp.Constant;
import com.semmle.js.ast.regexp.ControlEscape;
import com.semmle.js.ast.regexp.ControlLetter;
@@ -568,6 +569,7 @@ public class RegExpParser {
STANDARD,
INTERSECTION,
SUBTRACTION,
UNION
}
// ECMA 2024 `v` flag allows nested character classes.
@@ -599,12 +601,26 @@ public class RegExpParser {
}
}
boolean containsComplex = elements.stream().anyMatch(term -> term instanceof UnicodePropertyEscape ||
term instanceof CharacterClassQuotedString ||
term instanceof CharacterClass);
// Set type to UNION only if:
// 1. We haven't already determined a specific type (intersection/subtraction)
// 2. We have more than one element
// 3. We have at least one complex element (i.e. a nested character class or a UnicodePropertyEscape)
if (containsComplex && classType == CharacterClassType.STANDARD && elements.size() > 1) {
classType = CharacterClassType.UNION;
}
// Create appropriate RegExpTerm based on the detected class type
switch (classType) {
case INTERSECTION:
return this.finishTerm(new CharacterClass(loc, Collections.singletonList(new CharacterClassIntersection(loc, elements)), inverted));
case SUBTRACTION:
return this.finishTerm(new CharacterClass(loc, Collections.singletonList(new CharacterClassSubtraction(loc, elements)), inverted));
case UNION:
return this.finishTerm(new CharacterClass(loc, Collections.singletonList(new CharacterClassUnion(loc, elements)), inverted));
case STANDARD:
default:
return this.finishTerm(new CharacterClass(loc, elements, inverted));