@mathroc what do you think?

that's working for me :)
👍

+1

I'm trying to add unit tests to cover this (#366), as the return line is never being hit by the current testsuite. However, I have a few questions about this code:

1. What is the purpose of the strpos conditional? As written, this will always return true. Unless this is intended to work around some kind of bug in the PHP core, strpos should always return an integer offset >= 0, or boolean false.
2. The case cited in the comment is already in the testsuite, but never actually hits this - the relevant test case seems to be handled by add rfc3339 helper class #238, which was merged several months prior to this PR. I assume that means there's a reason for this that isn't covered by the rfc3339 helper thing, but I can't figure out what that might be.
3. This code appears to be intended for checking date-time constraints, but that constraint is entirely handled by the rfc3339 helper. Does this code perhaps need to be moved and integrated with that?

I would appreciate it if someone could clarify what this code is actually for, and whether it still serves a purpose, needs moving, has been superseded, or was simply merged in error after the problem was already solved.

        // handles the case where a non-6 digit microsecond datetime is passed
        // which will fail the above string comparison because the passed
        // $datetime may be '2000-05-01T12:12:12.123Z' but format() will return
        // '2000-05-01T12:12:12.123000Z'
        if ((strpos('u', $format) !== -1) && (preg_match('/\.\d+Z$/', $datetime))) {
            return true;
        }