add support for escape char classes inside char classes

javascript/ql/src/Performance/ReDoS.ql

@@ -234,7 +234,14 @@ private module CharacterClasses {
       exists(string lo, string hi | child.(RegExpCharacterRange).isRange(lo, hi) |
         lo <= char and char <= hi
       )
-      // TODO: RegExpCharacterClassEscape.
+      or
+      exists(RegExpCharacterClassEscape escape | escape = child |
+        escape.getValue() = escape.getValue().toLowerCase() and
+        classEscapeMatches(escape.getValue(), char)
+        or
+        escape.getValue() = escape.getValue().toUpperCase() and
+        not classEscapeMatches(escape.getValue().toLowerCase(), char)
+      )
     )
   }
 
@@ -248,6 +255,12 @@ private module CharacterClasses {
       child.(RegExpCharacterRange).isRange(result, _)
       or
       child.(RegExpCharacterRange).isRange(_, result)
+      or
+      exists(RegExpCharacterClassEscape escape | child = escape |
+        result = min(string s | classEscapeMatches(escape.getValue().toLowerCase(), s))
+        or
+        result = max(string s | classEscapeMatches(escape.getValue().toLowerCase(), s))
+      )
     )
   }
 
@@ -264,15 +277,7 @@ private module CharacterClasses {
     bindingset[char]
     override predicate matches(string char) { hasChildThatMatches(cc, char) }
 
-    override string choose() {
+    override string choose() { result = min(string c | c = getAMentionedChar(cc)) }
-      result =
-        min(string c |
-          exists(RegExpTerm child | child = cc.getAChild() |
-            c = child.(RegExpConstant).getValue() or
-            child.(RegExpCharacterRange).isRange(c, _)
-          )
-        )
-    }
   }
 
   /**
@@ -293,19 +298,14 @@ private module CharacterClasses {
 
     override string choose() {
       // The next char after the max of the inverted charclass.
-      result =
+      result = nextChar(max(string c | c = getAMentionedChar(cc)))
-        nextChar(max(string c |
-            exists(RegExpTerm child | child = cc.getAChild() |
-              c = child.(RegExpConstant).getValue() or
-              child.(RegExpCharacterRange).isRange(_, c)
-            )
-          ))
     }
   }
 
   /**
    * Holds if the character class escape `clazz` (\d, \s, or \w) matches `char`.
    */
+  pragma[noinline]
   private predicate classEscapeMatches(string clazz, string char) {
     clazz = "d" and
     char = "0123456789".charAt(_)

javascript/ql/test/query-tests/Performance/ReDoS/ReDoS.expected

@@ -20,6 +20,7 @@
 | regexplib/dates.js:66:201:66:208 | NOVEMBER | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'NOVEMBER'. |
 | regexplib/dates.js:66:210:66:217 | DECEMBER | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'DECEMBER'. |
 | regexplib/dates.js:66:234:66:240 | PRESENT | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'PRESENT'. |
+| regexplib/email.js:1:16:1:22 | [-.\\w]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:5:24:5:35 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:5:63:5:74 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:6:10:6:35 | (?:[a-zA-Z0-9][\\.\\-\\+_]?)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
@@ -29,16 +30,18 @@
 | regexplib/email.js:25:212:25:223 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:25:251:25:262 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:32:10:32:25 | (?:\\w[\\.\\-\\+]?)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
+| regexplib/email.js:33:16:33:22 | [-.\\w]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:33:38:33:51 | ([0-9a-zA-Z])+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '00.'. |
+| regexplib/email.js:33:53:33:58 | [-\\w]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:34:24:34:35 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/email.js:34:63:34:74 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/markup.js:3:451:3:453 | .+? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'a '. |
 | regexplib/markup.js:13:6:13:12 | [^"']+? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '('. |
 | regexplib/markup.js:13:14:13:16 | .+? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'a"'. |
-| regexplib/markup.js:37:29:37:56 | [a-zA-Z0-9\|:\|\\/\|=\|-\|.\|\\?\|&]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '='. |
+| regexplib/markup.js:37:29:37:56 | [a-zA-Z0-9\|:\|\\/\|=\|-\|.\|\\?\|&]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '00='. |
 | regexplib/markup.js:40:23:40:25 | \\w+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/markup.js:40:132:40:134 | \\s* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ' @0<""'. |
-| regexplib/markup.js:53:29:53:56 | [a-zA-Z0-9\|:\|\\/\|=\|-\|.\|\\?\|&]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '='. |
+| regexplib/markup.js:53:29:53:56 | [a-zA-Z0-9\|:\|\\/\|=\|-\|.\|\\?\|&]* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '00='. |
 | regexplib/markup.js:56:23:56:25 | \\w+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
 | regexplib/markup.js:56:132:56:134 | \\s* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ' @0<""'. |
 | regexplib/misc.js:15:56:15:118 | (([^\\\\/:\\*\\?"\\\|<>\\. ])\|([^\\\\/:\\*\\?"\\\|<>]*[^\\\\/:\\*\\?"\\\|<>\\. ]))? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '!\\\\}'. |
@@ -68,14 +71,14 @@
 | tst.js:19:71:19:90 | (?:[^)\\\\]\|\\\\\\\\\|\\\\.)+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\\\\\\\\'. |
 | tst.js:31:54:31:55 | .* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\|a\|\\n'. |
 | tst.js:36:23:36:32 | (\\\\\\/\|.)*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\\\\/'. |
-| tst.js:41:27:41:28 | .* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '#'. |
+| tst.js:41:27:41:28 | .* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\t#'. |
 | tst.js:47:25:47:27 | .*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '""'. |
 | tst.js:47:31:47:33 | .*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ''''. |
 | tst.js:52:37:52:39 | .*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ']['. |
 | tst.js:52:70:52:72 | .*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ']['. |
 | tst.js:58:15:58:20 | [a-z]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'a'. |
 | tst.js:60:43:60:54 | [a-zA-Z0-9]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
-| tst.js:66:16:66:31 | [\\w#:.~>+()\\s-]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '#'. |
+| tst.js:66:16:66:31 | [\\w#:.~>+()\\s-]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\t'. |
 | tst.js:66:38:66:40 | .*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ']['. |
 | tst.js:71:19:71:26 | (\\\\?.)*? | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\\\\a'. |
 | tst.js:74:14:74:21 | (b\|a?b)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of 'b'. |
@@ -100,5 +103,9 @@
 | tst.js:149:15:149:24 | (\\s\|[\\f])* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\u000c'. |
 | tst.js:152:15:152:28 | (\\s\|[\\v]\|\\\\v)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\u000b'. |
 | tst.js:155:15:155:24 | (\\f\|[\\f])* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '\u000c'. |
-| tst.js:158:15:158:22 | (\\W\|\\D)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '!'. |
+| tst.js:158:15:158:22 | (\\W\|\\D)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of ' '. |
 | tst.js:161:15:161:22 | (\\S\|\\w)* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
+| tst.js:164:15:164:24 | (\\S\|[\\w])* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
+| tst.js:167:15:167:27 | (1s\|[\\da-z])* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '1s'. |
+| tst.js:170:15:170:23 | (0\|[\\d])* | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |
+| tst.js:173:16:173:20 | [\\d]+ | This part of the regular expression may cause exponential backtracking on strings containing many repetitions of '0'. |

javascript/ql/test/query-tests/Performance/ReDoS/tst.js

@@ -158,4 +158,16 @@ var bad35 = /((\f|[\f])*)"/;
 var bad36 = /((\W|\D)*)"/;
 
 // NOT GOOD
-var bad37 = /((\S|\w)*)"/;
+var bad37 = /((\S|\w)*)"/;
+
+// NOT GOOD
+var bad38 = /((\S|[\w])*)"/;
+
+// NOT GOOD
+var bad39 = /((1s|[\da-z])*)"/;
+
+// NOT GOOD
+var bad40 = /((0|[\d])*)"/;
+
+// NOT GOOD
+var bad41 = /(([\d]+)*)"/;