materialize less strings when ranking states

Author: erik-krogh

javascript/ql/test/query-tests/Security/CWE-400/ReDoS/PolynomialBackTracking.expected

@@ -579,4 +579,4 @@ edges
 | polynomial-redos.js:130:2:130:31 | modifie ... g, "b") | polynomial-redos.js:5:16:5:32 | req.query.tainted | polynomial-redos.js:130:2:130:9 | modified | This $@ that depends on $@ may run slow on strings starting with 'c' and with many repetitions of 'c'. | polynomial-redos.js:130:21:130:22 | c+ | regular expression | polynomial-redos.js:5:16:5:32 | req.query.tainted | a user-provided value |
 | polynomial-redos.js:133:2:133:32 | modifie ... g, "b") | polynomial-redos.js:5:16:5:32 | req.query.tainted | polynomial-redos.js:133:2:133:10 | modified2 | This $@ that depends on $@ may run slow on strings starting with 'f' and with many repetitions of 'f'. | polynomial-redos.js:133:22:133:23 | f+ | regular expression | polynomial-redos.js:5:16:5:32 | req.query.tainted | a user-provided value |
 | polynomial-redos.js:136:5:136:35 | modifie ... g, "b") | polynomial-redos.js:5:16:5:32 | req.query.tainted | polynomial-redos.js:136:5:136:13 | modified3 | This $@ that depends on $@ may run slow on strings starting with 'h' and with many repetitions of 'h'. | polynomial-redos.js:136:25:136:26 | h+ | regular expression | polynomial-redos.js:5:16:5:32 | req.query.tainted | a user-provided value |
-| polynomial-redos.js:138:5:138:326 | tainted ... )C.*X/) | polynomial-redos.js:5:16:5:32 | req.query.tainted | polynomial-redos.js:138:5:138:11 | tainted | This $@ that depends on $@ may run slow on strings starting with 'AAAAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC'. | polynomial-redos.js:138:322:138:323 | .* | regular expression | polynomial-redos.js:5:16:5:32 | req.query.tainted | a user-provided value |
+| polynomial-redos.js:138:5:138:326 | tainted ... )C.*X/) | polynomial-redos.js:5:16:5:32 | req.query.tainted | polynomial-redos.js:138:5:138:11 | tainted | This $@ that depends on $@ may run slow on strings starting with 'AAAAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC'. | polynomial-redos.js:138:322:138:323 | .* | regular expression | polynomial-redos.js:5:16:5:32 | req.query.tainted | a user-provided value |

javascript/ql/test/query-tests/Security/CWE-400/ReDoS/PolynomialReDoS.expected

@@ -2,4 +2,4 @@
 | test.py:9:14:9:29 | Str | test.py:9:27:9:29 | \\d+ | Strings starting with '0.9' and with many repetitions of '99' can start matching anywhere after the start of the preceeding \\d+ |
 | test.py:11:22:11:33 | Str | test.py:11:31:11:33 | \\s+ | Strings with many repetitions of ' ' can start matching anywhere after the start of the preceeding \\s+$ |
 | test.py:18:14:18:25 | Str | test.py:18:23:18:25 | \\s+ | Strings with many repetitions of ' ' can start matching anywhere after the start of the preceeding \\s+$ |
-| test.py:20:23:20:274 | Str | test.py:20:273:20:274 | .* | Strings starting with 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' can start matching anywhere after the start of the preceeding (AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)C.*Y |
+| test.py:20:23:20:274 | Str | test.py:20:273:20:274 | .* | Strings starting with 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' can start matching anywhere after the start of the preceeding (AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)(AA\|BB)C.*Y |

python/ql/test/query-tests/Security/CWE-730-PolynomialReDoS/PolynomialBackTracking.expected

@@ -29,4 +29,4 @@ subpaths
 | test.py:9:32:9:35 | ControlFlowNode for text | test.py:2:26:2:32 | ControlFlowNode for ImportMember | test.py:9:32:9:35 | ControlFlowNode for text | This $@ that depends on a $@ may run slow on strings starting with '0.9' and with many repetitions of '99'. | test.py:9:27:9:29 | \\d+ | regular expression | test.py:2:26:2:32 | ControlFlowNode for ImportMember | user-provided value |
 | test.py:12:17:12:20 | ControlFlowNode for text | test.py:2:26:2:32 | ControlFlowNode for ImportMember | test.py:12:17:12:20 | ControlFlowNode for text | This $@ that depends on a $@ may run slow on strings with many repetitions of ' '. | test.py:11:31:11:33 | \\s+ | regular expression | test.py:2:26:2:32 | ControlFlowNode for ImportMember | user-provided value |
 | test.py:16:24:16:30 | ControlFlowNode for my_text | test.py:2:26:2:32 | ControlFlowNode for ImportMember | test.py:16:24:16:30 | ControlFlowNode for my_text | This $@ that depends on a $@ may run slow on strings with many repetitions of ' '. | test.py:18:23:18:25 | \\s+ | regular expression | test.py:2:26:2:32 | ControlFlowNode for ImportMember | user-provided value |
-| test.py:21:18:21:21 | ControlFlowNode for text | test.py:2:26:2:32 | ControlFlowNode for ImportMember | test.py:21:18:21:21 | ControlFlowNode for text | This $@ that depends on a $@ may run slow on strings starting with 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC'. | test.py:20:273:20:274 | .* | regular expression | test.py:2:26:2:32 | ControlFlowNode for ImportMember | user-provided value |
+| test.py:21:18:21:21 | ControlFlowNode for text | test.py:2:26:2:32 | ControlFlowNode for ImportMember | test.py:21:18:21:21 | ControlFlowNode for text | This $@ that depends on a $@ may run slow on strings starting with 'AAAAAAAAAAAAAAAAAAAABBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC' and with many repetitions of 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAC'. | test.py:20:273:20:274 | .* | regular expression | test.py:2:26:2:32 | ControlFlowNode for ImportMember | user-provided value |

python/ql/test/query-tests/Security/CWE-730-PolynomialReDoS/PolynomialReDoS.expected

@@ -147,6 +147,8 @@ module Make<RegexTreeViewSig TreeImpl> {
   /**
    * Gets a string for the full location of `t`.
    */
+  bindingset[t]
+  pragma[inline_late]
   string getTermLocationString(RegExpTerm t) {
     exists(string file, int startLine, int startColumn, int endLine, int endColumn |
       t.hasLocationInfo(file, startLine, startColumn, endLine, endColumn) and

shared/regex/codeql/regex/nfa/NfaUtils.qll

@@ -354,13 +354,13 @@ module Make<RegexTreeViewSig TreeImpl> {
    */
   private int rankState(State state) {
     state =
-      rank[result](State s |
-        exists(StateTuple tuple |
-          tuple = getARelevantStateTuple(_, _) and
-          s = [tuple.getFirst(), tuple.getSecond(), tuple.getThird()]
+      rank[result](State s, int startLine, int endLine, int startColumn, int endColumn |
+        exists(StateTuple tuple | tuple = getARelevantStateTuple(_, _) |
+          s = [tuple.getFirst(), tuple.getSecond(), tuple.getThird()] and
+          s.getRepr().hasLocationInfo(_, startLine, startColumn, endLine, endColumn)
         )
       |
-        s order by getTermLocationString(s.getRepr())
+        s order by startLine, startColumn, endLine, endColumn
       )
   }