problem with uploading translation to the source document (json)



first of all, thank you for launching the new version of - uploading .json file is very comfortable for us.

Now, we have project Crystal (version 2018-05-24) in English and we tried to upload the translated text (in Spanish), but only dozens of pairs of translations were matched even though the whole translation has hundreds of translations which is ca half of the current source text. Please see attached files. For example, item "word sketch" is matched, but the item with the key "wordlist" is not.
The number of lines (and their order) does not correspond, but I would expect the translation works only with the keys which are unique.

Thank you.




Michal Cukr
July 19, 2018, 9:52 AM

Thanks, I'm looking forward to the version 4.6

Damian Jansen
July 19, 2018, 5:41 AM

sorry for the late response. I have a fix under QA that should address the issue, it required a rewrite of the adapter and a reconsideration of how we upload raw translation files. Trying to get it into the impending 4.6 release.

Michal Cukr
May 29, 2018, 2:16 PM

Can you provide us an approximate time of fixing it?

Sean Flanigan
May 29, 2018, 4:34 AM

see for the offline po issue.

Michal Cukr
May 28, 2018, 7:38 AM

Moreover, I have detected that uploading of .po file doesn't even work. I downloaded this file from Zanata for offline translation. I have translated a few items, e.g. Corpus description and uploaded back to Zanata. But no items were translated or fuzzy.

I am attaching the file with translation made offline (ca from the line 1320 et seq. there are translated items), but the translated items were not matched after re-uploading to Zanata. Why? This is the core function. How can we continue in working when we prepare a new version of texts for localization but we would not upload the previous translation...

Ready for Release


Damian Jansen


Michal Cukr


Tested Version/s



Fix versions

Affects versions