Re: Status of time_get and time_put facets

[Martin Sebor <sebor at roguewave dot com>]

Paolo Carlini wrote:
>
...
> Ok, thanks for your explanation and hints: I plan to investigate in
> detail the issue in the next days.
>
> For sure the current testsuite lacks testcases for the behaviour
> you expect...
Here's another test case that translates the name of a weekday
read from stdin in a locale given by the first argument to a
locale given by the second command line argument. It uses
a single locale object with time_get and time_put each from
a distinct named locale. Sample output is here:

    $ echo "Mittwoch" | ./t de_DE fr_FR
    mercredi

In order for this to work correctly, the stream object cannot
be used to retrieve the facet data. Each time facet has to have
its own.

With gcc/libstdc++ 3.2.1 it fails to compile due to a const
correctness bug (locale template ctor takes any Facet*, i.e.,
including const std::time_get<char>*). After fixing the
compilation error, the program segfaults.

Regards
Martin


#include <iostream>
#include <locale>

int main (int argc, const char *argv[])
{
    const std::locale l1 (argc > 1 ? argv [1] : "");
    const std::locale l2 (argc > 2 ? argv [2] : "");

    const std::locale l1_l2 =
        std::locale (std::locale (std::locale::classic (),
                                  &std::use_facet<std::time_get<char> 
>(l1)),
                     &std::use_facet<std::time_put<char> >(l2));

    const std::time_get<char> &tg =
        std::use_facet<std::time_get<char> >(l1_l2);

    const std::time_put<char> &tp =
        std::use_facet<std::time_put<char> >(l1_l2);

    std::tm t = std::tm ();

    std::ios::iostate err = std::ios::goodbit;

    // benign but unnecessary
    // std::cin.imbue (l1_l2);
    // std::cout.imbue (l1_l2);

    tg.get_weekday (std::cin.rdbuf (), std::istreambuf_iterator<char>(),
                    std::cin, err, &t);

    if (err & std::ios::failbit)
        return 1;

    tp.put (std::cout.rdbuf (), std::cout, ' ', &t, 'A') = '\n';
}