On Sun, Oct 18, 2015 at 8:30 AM, Dirk Hohndel <[email protected]> wrote: > >> On Oct 17, 2015, at 10:20 PM, Miika Turkia <[email protected]> wrote: >>>>> >>>>> I'll fix that on Transifex. And I'll check the other languages as well, >>>>> Benjamin. >>>> >>>> We are slowly getting a nice list of formatting options that must be >>>> rejected before release. Time for some scripting, I assume. >>>> Unfortunately all % signs are not marking an argument but might be >>>> literal % in the output... >>> >>> It should be a valid requirement to say that the number of '%' in the >>> original >>> string and the translation must be the same. >>> >>> That would have been enough to catch this one. >> >> And % followed by the same char as the original. Remember %S, and I >> think we have had an issue with %1 being translated with %x or >> similar. > > You have to be careful because it is not at all unusual that you have > "bla %d something %s" but the translation is "huff %s puff %d" :-) > > But by all means, PLEASE create a script that can help us validate > the translation. Even if it has false positives, that's easy enough to > manually check...
I think we should first check the Transifex validation preferences for the project (which I don't have right to look at). Documentation indicates that there might be some settings that could prevent these bugs: https://www.transifex.com/blog/2013/customize-translation-checks-for-your-files/ miika _______________________________________________ subsurface mailing list [email protected] http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
