stdcxx-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Martin Sebor (JIRA)" <>
Subject [jira] Updated: (STDCXX-535) std::time_put can generate output that is not parseable by std::time_get facet
Date Sun, 06 Apr 2008 23:00:24 GMT


Martin Sebor updated STDCXX-535:

              Severity: Incorrect Behavior
         Fix Version/s: 4.3
              Assignee: Martin Sebor
    Remaining Estimate: 4h
     Original Estimate: 4h

Sounds like POSIX doesn't intend for {{strftime()}} to produce output that's not parseable
by {{strptime()}} in this case. We need to file an Aardvark with POSIX and tweak our code
to conform to the intended POSIX behavior.

Scheduled for 4.3.

> std::time_put can generate output that is not parseable by std::time_get facet
> ------------------------------------------------------------------------------
>                 Key: STDCXX-535
>                 URL:
>             Project: C++ Standard Library
>          Issue Type: Bug
>          Components: 22. Localization
>    Affects Versions: 4.1.2, 4.1.3, 4.1.4
>            Reporter: Travis Vitek
>            Assignee: Martin Sebor
>             Fix For: 4.3
>   Original Estimate: 4h
>  Remaining Estimate: 4h
> std::time_get<>::get_date() is required to be able to parse the output produced
by std::time_put<>::put(..., 'x'). For some locales, the '%x' format specifier expands
out to '%e.%m.%Y'. When a date is formatted using this, there will be a leading space, and
that leading space causes the time_get<>::get_date() operation to fail.
> The root of the problem is that the POSIX strftime() function [which we emulate in time_get]
requires that the '%e' specifier generate whitespace for single digit monthdays, and the POSIX
strptime() function says that the number may be padded on the left with 0s. It does not appear
to specify that whitespace is allowed. The strptime() implementation on some platforms [sun,
linux, compaq, aix] allow this whitespace, while others [hp, freebsd] do not.
> Discussion here.
> []

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

View raw message