regexp: skip backtracker for long programs

This update makes maxBacktrackLen return 0 if
len(prog.Inst) > maxBacktrackProg. This prevents an attempt to
backtrack against a nil bitstate.

Fixes #10319

Change-Id: Icdbeb2392782ccf66f9d0a70ea57af22fb93f01b
Reviewed-on: https://go-review.googlesource.com/8473
Reviewed-by: Brad Fitzpatrick <[email protected]>

Author: mattyb

src/regexp/backtrack.go

@@ -47,21 +47,30 @@ var notBacktrack *bitState = nil
 // maxBitStateLen returns the maximum length of a string to search with
 // the backtracker using prog.
 func maxBitStateLen(prog *syntax.Prog) int {
+	if !shouldBacktrack(prog) {
+		return 0
+	}
 	return maxBacktrackVector / len(prog.Inst)
 }
 
 // newBitState returns a new bitState for the given prog,
 // or notBacktrack if the size of the prog exceeds the maximum size that
 // the backtracker will be run for.
 func newBitState(prog *syntax.Prog) *bitState {
-	if len(prog.Inst) > maxBacktrackProg {
+	if !shouldBacktrack(prog) {
 		return notBacktrack
 	}
 	return &bitState{
 		prog: prog,
 	}
 }
 
+// shouldBacktrack reports whether the program is too
+// long for the backtracker to run.
+func shouldBacktrack(prog *syntax.Prog) bool {
+	return len(prog.Inst) <= maxBacktrackProg
+}
+
 // reset resets the state of the backtracker.
 // end is the end position in the input. ncap and reqcap are the number
 // of the machine's capture registers and the number of user-requested

src/regexp/exec_test.go

@@ -713,3 +713,15 @@ func TestLongest(t *testing.T) {
 		t.Errorf("longest match was %q, want %q", g, w)
 	}
 }
+
+// TestProgramTooLongForBacktrace tests that a regex which is too long
+// for the backtracker still executes properly.
+func TestProgramTooLongForBacktrack(t *testing.T) {
+	longRegex := MustCompile(`(one|two|three|four|five|six|seven|eight|nine|ten|eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen|eighteen|nineteen|twenty|twentyone|twentytwo|twentythree|twentyfour|twentyfive|twentysix|twentyseven|twentyeight|twentynine|thirty|thirtyone|thirtytwo|thirtythree|thirtyfour|thirtyfive|thirtysix|thirtyseven|thirtyeight|thirtynine|forty|fortyone|fortytwo|fortythree|fortyfour|fortyfive|fortysix|fortyseven|fortyeight|fortynine|fifty|fiftyone|fiftytwo|fiftythree|fiftyfour|fiftyfive|fiftysix|fiftyseven|fiftyeight|fiftynine|sixty|sixtyone|sixtytwo|sixtythree|sixtyfour|sixtyfive|sixtysix|sixtyseven|sixtyeight|sixtynine|seventy|seventyone|seventytwo|seventythree|seventyfour|seventyfive|seventysix|seventyseven|seventyeight|seventynine|eighty|eightyone|eightytwo|eightythree|eightyfour|eightyfive|eightysix|eightyseven|eightyeight|eightynine|ninety|ninetyone|ninetytwo|ninetythree|ninetyfour|ninetyfive|ninetysix|ninetyseven|ninetyeight|ninetynine|onehundred)`)
+	if !longRegex.MatchString("two") {
+		t.Errorf("longRegex.MatchString(\"two\") was false, want true")
+	}
+	if longRegex.MatchString("xxx") {
+		t.Errorf("longRegex.MatchString(\"xxx\") was true, want false")
+	}
+}