Fix extraction of translation messages by Babel - #1047
Conversation
Sorry, something went wrong.
A directory was missing.
Codecov Report
@@ Coverage Diff @@
## master #1047 +/- ##
==========================================
- Coverage 61.98% 61.96% -0.03%
==========================================
Files 227 227
Lines 16405 16405
==========================================
- Hits 10169 10165 -4
- Misses 6236 6240 +4
Continue to review full report at Codecov.
|
Sorry, something went wrong.
stefano-maggiolo
left a comment
There was a problem hiding this comment.
I think the reason we didn't want **py is because of the directory created by setup.py?
Also: is it possible that when I regenerated the po I also deleted some of the existing translation due to this bug? Can you check and if so restore the state of before?
Reviewable status: 0 of 2 files reviewed, all discussions resolved
Sorry, something went wrong.
stefano-maggiolo
left a comment
There was a problem hiding this comment.
Reviewed 2 of 2 files at r1.
Reviewable status:complete! all files reviewed, all discussions resolved
Sorry, something went wrong.
lw
left a comment
There was a problem hiding this comment.
Ok, I will leave the explicit listing of the directories.
I didn't understand what you're referring to: you updated the .pot file, which I regenerated in this PR, but you didn't touch the .po files. (In fact they haven't been updated in at least 6 month, whereas the pot with the missing strings was generated just one month ago). So I don't think any translation was lost due to this issue. Moreover, as we saw in #1045, translations for messages that disappeared from the pot are just commented out, not removed.
Reviewable status:
Sorry, something went wrong.
A directory was missing.
We could honestly just use **.py to extract from all Python files in the repo rather than listing all directories individually. I don't think we would have that many false positives (possibly none at all). I think I originally listed the directories because that was what we were doing before adopting Babel and I wanted to have the files processed in the same order to reduce the diff to the cms.pot file, which however ended being quite large regardless.
This change is