-
Notifications
You must be signed in to change notification settings - Fork 47
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Allow localizers extracting raw translations they’ve pushed during a cycle #569
Comments
I can see a use case for a view displaying the strings changed between two branches, like we do for Gaia, extending https://transvision.mozfr.org/channelcomparison/ with 'Added strings' and 'Removed strings' On the other hand, I'm not completely sure of how many people would use a file with only these strings, especially in an automated way, given the number of false positives. |
A view would be perfect. So maybe we could make the Gaia view generic enough to work for any repo and remove the Channel comparison? + add a way to query any of these set of strings via the API? (Just a thought) |
Making the Gaia view more generic would definitely work for me, there are other interesting things in there (e.g. changed English strings). An API call would work for me. |
Definitely agree for English strings, that would be more consistent as currently we have to check https://transvision.mozfr.org/channelcomparison/?locale=en-US&chan1=central&chan2=aurora to have the same data and that’s less obvious. |
Talked with Théo today and it seems that improving the channelcomparison view would suffice (basically it only lacks added strings). I would say it's doable during the hackathon. |
I’ve been doing that manually for some time now ( Editing https://github.com/mozfr/transvision/blob/master/app/scripts/tmxmaker.py#L54 locally to only return the text, e.g. one string per line, without its ID). Doing so allows me to check all the strings with a grammar checker (mine is http://www.antidote.info/antidote for both French and English). This also works when localizers are using a simple spell checker.
Currently it’s a pain to check 30,000 strings with lots of false positive where I could only check what changed between two branches. I also don’t do it for each release because it’s time consuming. Helping automate this would help catching typos like http://hg.mozilla.org/l10n-central/fr/rev/6cc9480baf4b One could even think of running a Jenkins server with a spell checker.
The data I need are basically the green and orange strings exposed by Elmo on https://l10n.mozilla.org/source/diff/?from=661304c55fbc&to=d9c01cf0f9e9&repo=l10n-central%2Ffr but without string IDs.
In Transvision world that would be New strings + strings displayed in Channel Comparison.
We could expose this in download view if you think several people would be interested in this, or just exposing an API if you don’t want to clutter the UI.
What do you guys think?
The text was updated successfully, but these errors were encountered: