Re: PHP-5.5.10RC1 is available
Hi,
On Fri, February 21, 2014 11:50, Julien Pauli wrote:
> On Thu, Feb 20, 2014 at 3:36 PM, Remi Collet <[email protected]>
> wrote:
>
>> Le 20/02/2014 12:55, Julien Pauli a écrit :
>>
>>> You can test it and report any bugs or regressions that you may
>>> notice using the bug system at http://bugs.php.net.
>>
>> From fresh test build:
>>
>>
>>
>> TEST FAILURE: ../ext/pcre/tests/match_flags3.diff --
>> 024+ Warning: preg_match(): Numeric named subpatterns are not allowed in
>> /builddir/build/BUILD/php-5.5.10RC1/ext/pcre/tests/match_flags3.php on
>>
> line
>> 14
>> 024- Warning: preg_match(): Compilation failed: group name must start
>> with a non-digit at offset %d in %smatch_flags3.php on line %d
>>
>> TEST FAILURE: ../ext/pcre/tests/grep2.diff --
>> 022+ array(0) {
>> 022- array(3) {
>> 023- [5]=>
>> 024- string(1) "a"
>> 025- ["xyz"]=>
>> 026- string(2) "q6"
>> 027- [6]=>
>> 028- string(3) "h20"
>> 024+ bool(true)
>> 030- bool(false)
>>
>>
>> TEST FAILURE: ../ext/pcre/tests/bug37911.diff --
>> 020+ Warning: preg_replace_callback(): Numeric named subpatterns are not
>> allowed in
> /builddir/build/BUILD/php-5.5.10RC1/ext/pcre/tests/bug37911.php o
>
>> n line 14 020- Warning: preg_replace_callback(): Compilation failed:
>> group name must start with a non-digit at offset %d in %sbug37911.php on
>> line %d
>>
>>
>>
>> I understand bundled libpcre have been update to 8.34 (and my build use
>> system pcre 8.32), and test have also been updates for changes.
>>
>> But this means we have behavior change (grep2.phpt) in a minor update.
>>
>>
>> Thoughts ?
>>
>
> Putting Anatol in the CC as he made this change.
>
>
> Julien.P
>
the match_flags3.phpt and bug37911.phpt are a more precise warning for
what is already handled that way. Looks like there's more internally, but
a warning text change is only what differs to outside. It's the entry 19.
for PCRE 8.34 here http://pcre.org/changelog.txt .
With the grep2.phpt - I don't see an exact item there, 36. for 8.33 could
be that. This is the only point where one could be in doubt actually. I'd
see that more like acceptable side effect of fixing a recursion limit
issue. Taking in account the amount of fixes not affecting the current
functionality, I think they prevail that one peccadillo.
Regards
Anatol
Thread (7 messages)