[pacman-dev] [PATCH] Tidy up missing changelog error message

[Allan McRae]

Xavier wrote:
> Allan McRae wrote:
>   
>> Dan McGee wrote:
>>     
>>> On Sat, May 10, 2008 at 11:31 AM, Allan McRae <mcrae_allan at hotmail.com> wrote:
>>>   
>>>       
>>>>        if((fp = alpm_pkg_changelog_open(pkg)) == NULL) {
>>>> -               /* TODO after string freeze use pm_fprintf */
>>>> -               fprintf(stderr, _("error: no changelog available for '%s'.\n"),
>>>> +               pm_fprintf(stderr, PM_LOG_ERROR, "no changelog available for '%s'.\n",
>>>>     
>>>>         
>>> Ooo, nice.
>>>
>>> Any chance you want to fold this into a bigger patch, hopefully
>>> slightly tested, finding more of these types of things? This one
>>> wasn't alone.
>>>   
>>>       
>> Before, I continue on with this, does the patched version of the string 
>> above get flagged for translation?  I ask because the only other use of 
>> the pm_fprintf function I found was in pacman/pacman.c.
>>
>> pm_fprintf(stderr, PM_LOG_ERROR, "Internal pacman error: Segmentation 
>> fault.\n"
>>         "Please submit a full bug report with --debug if appropriate.\n");
>>
>> Now, I can't find those strings in the files in the po/ folder.  These 
>> lines are in pacman-3.1.4 so should have been translated.  Am I missing 
>> something obvious or should the strings be surrounded with "_( )"?
>>
>>
>>     
>
> Yes, _ is what indicates which strings can be translated :
> src/pacman/util.h:#define _(str) gettext(str)
>
> In the case of an error that is always displayed (not debug output), we 
> want it to be translated. So we need to keep _ there indeed.
>   
So just to clarify, was the "Internal pacman error..." message above not 
flagged for translation for a reason?  Otherwise, I will fix this as I 
update all other fprintf lines.

Allan