Russ Allbery wrote: > Sure, that sounds like a good idea to me. Should we trigger on anything > in a lintian template that looks like ${DISK}? Would $DISK also be > valid, or are the curly braces required?
The braces are required. > I assume that something like the following pseudocode would work: It should really count the number of times each variable occurs in the original string and check that it occurs the same number of times in the translated string. It should also ignore any strings that are marked "fuzzy" In D-I this does sometimes result in false negatives as a translator can conceivably condense the text in such a way that the repetition of the variable is not needed in his translation, but those are rare exceptions. This check is really a check for the translators, and not for package maintainers. As such adding it to Lintian is not immediately obvious. Also, correcting the error in certain translations may be non-trivial if the translation uses a totally different character set than the maintainer uses. Therefore IMO the explanation for the Lintian error should explain that maintainers should only correct errors in translations themselves if: - the error is completely obvious; examples: - simple typo in variable name - variable in the wrong case (upper v.s. lower) - braces forgotten - normal braces used instead of curly ones - (s)he's reasonable sure there are no other errors in the translation (for example: if the translation is significantly shorter than the original, it is quite likely that the translation is just plain wrong) - (s)he can edit the translation without causing encoding problems In all other cases the maintainer should IMO either contact the translator (preferred), or mark the string fuzzy. Personally I feel that this check would be more usefully added on the web pages that show translation status for the translators themselves. Even better would be if we could develop some kind of "translation update submission system" that did such checks would be done there (and where the translators would have the option of saying "Yes, I'm really sure this is OK"). Cheers, FJP -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]