stdcxx-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Travis Vitek (JIRA)" <>
Subject [jira] Commented: (STDCXX-676) [MSVC] 22.locale.time.put test fails
Date Wed, 05 Dec 2007 03:43:44 GMT


Travis Vitek commented on STDCXX-676:

Thanks for the feedback. It looks like I need to go back and revise my patch. I _believe_
that this issue will affect any configuration that uses the strftime defined in the MSCRT
library. I don't know much about cygwin, but I believe that platform would be excluded.

As for the supported format specifiers, I only listed the first few that I ran into problems
with. The MSCRT strftime() only supports 'aAbBcdHIjmMpSUwWxyYzZ', I added code for 'CDeFgGhnrRtTuV'.
I didn't add support for the modifiers `O' and `E' because the test doesn't actually rely
on them, it hard-codes the expected values for these instead. Now that I've written this up,
I suppose I could have done the same and used hardcoded values where appropriate. This would
mean removing many of the tests, so it probably isn't a good idea.

Also, I was trying to avoid using std::strcpy(). I feel weird about using the library that
I'm testing in the implementation of the actual test. I could have opted to use the C functions
from <string.h>, but that would have meant modifying some unrelated code to improve
consistency inside the test.

> [MSVC] 22.locale.time.put test fails
> ------------------------------------
>                 Key: STDCXX-676
>                 URL:
>             Project: C++ Standard Library
>          Issue Type: Bug
>          Components: Tests
>    Affects Versions: 4.2.0
>            Reporter: Travis Vitek
>             Fix For: 4.2.1
>         Attachments: stdcxx-223.patch
> In my manual runs I have noticed that the test fails because the MSCRT strftime doesn't
support %T, %e, or %G and asserts internally. This doesn't appear to happen on all configurations
because these tests are running to completion on some of the nightly builds. 
> I think the rw_strftime() that is inside the test could be expanded to avoid the assert
inside the CRT so that the test would pass on all windows configurations. 

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message