At one point it seems we've changed how a regular expression is formed from the user input in the Tokenizer processor, but we didn't update the validation. This means that, in some edge cases, a valid input could lead to a form error or some invalid expression would be accepted.

Comments

drunken monkey created an issue. See original summary.

drunken monkey’s picture

Status: Active » Needs review
StatusFileSize
new951 bytes

This should fix it.

Status: Needs review » Needs work

The last submitted patch, 2: 2871497-2--tokenizer_regex_validation.patch, failed testing.

drunken monkey’s picture

Status: Needs work » Needs review
StatusFileSize
new743 bytes
new1.65 KB

Oh, seems we actually have a test for the wrong behavior!

Status: Needs review » Needs work

The last submitted patch, 4: 2871497-4--tokenizer_regex_validation.patch, failed testing.

drunken monkey’s picture

Status: Needs work » Needs review
StatusFileSize
new589 bytes
new1.66 KB

Huh, turns out PREG is amazingly fault-tolerant when it comes to character classes. Tried a lot of nonsense, without any errors. But this one should work.

borisson_’s picture

Status: Needs review » Reviewed & tested by the community

  • drunken monkey committed 265beff on 8.x-1.x
    Issue #2871497 by drunken monkey: Fixed validation of "Whitespace...
drunken monkey’s picture

Status: Reviewed & tested by the community » Fixed

Thanks for reviewing!
Committed.

Status: Fixed » Closed (fixed)

Automatically closed - issue fixed for 2 weeks with no activity.