How to treat (undetected) syntax errors in translation strings?

[buhtz]

Hello,

I am an upstream maintainer managing over 40 languages in a Python application. There is one string in my Spanish translation containing to many } and causing exceptions when running. My translation platform (Weblate) was not able to detect that.

How would you suggest to deal with possible errors like that. I can not manually review all translations and also don't have enough skilled reviewers in each of that language.

I think about to somehow test that in a script using every translatable string in every available language.

What do you think?

Best,
Christian

X-Post: izimobil/polib#156

[akx]

Could you show an example of the buggy message (e.g. a link to a commit fixing the .po file)?

[buhtz]

The commit is to big but

1. "foo {bar}}"
2. "foo {{bar}}" (this won't throw an Exception but is a typo in most cases)
3. translated: "{foo} {bar}" original source: "foo {bar}"

I have seen all this cases in my real translations. I have a pending PR fixing the translations and improving my helper script checking such cases in po files.

bit-team/backintime#1829

But there might be a more elegant solution?