Re: [PATCH 2/2] Add C_LOCALE_OUTPUT prereq to test cases that require English text matching

From: Jonathan Nieder <jrnieder@gmail.com>
Date: Jun 25, 2012, 11:39 UTC
Message-ID: <20120625113956.GB9270@burratino>
In-Reply-To: <CACsJy8DpOL+V=sVZt3B-7OQ=POHtu1bovFB-sQ9HwPpmdKmN-A@mail.gmail.com>

Nguyen Thai Ngoc Duy wrote:

> On Sun, Jun 24, 2012 at 11:28 PM, Jonathan Nieder <jrnieder@gmail.com> wrote:
>> Nguyễn Thái Ngọc Duy wrote:

Show 14 quoted lines

>>> -     test_expect_${3:-success} "relative date ($2)" "
>>> +     test_expect_${3:-success} C_LOCALE_OUTPUT "relative date ($2)" "
>>>       test-date show $t >actual &&
>>>       test_cmp expect actual
>>>       "
>>
>> Could use test_i18ncmp so we catch if test-date crashes, but anyway,
>> yeah, we can't expect to be able to meaningfully test date formatting
>> in another language.
>
> This reminds me of an inkscape crash I encountered long ago because of
> a mistake in Vietnamese translation. Too bad we can't test all locales
> for similar bugs (or we would need to i18n-ize expected results in the
> test suite too)

Sure we can. :) We just need to make sure we are using test_i18ncmp where appropriate so the tests run and check as much as they can without relying on details that are left to the translators' disgression.

But yes, the test harness doesn't know how to do that yet. I'm hoping that soon-ish there will be an option to run tests in your chosen language, so the test result can more closely reflect bugs one is likely to run into in practice (and so developers can more easily test the edge cases such as Turkic locales where tolower('I') != 'i'). Sorry I haven't had enough time to bring this kind of thing into reality, and thanks for stepping in with the partial work in that direction before.

Jonathan