Gettext "Language" header should separate language and country with underscore
I've noticed a strange issue whereby for Brazilian Portuguese, the "Language" field in the generated gettext file is: "Language: pt-BR\n"
For other county-specific languages, such as Mexican Spanish, the "Language" field in the generated file is: "Language: es_MX\n"
Polish (Poland) is: "Language: pl_PL\n"
We are using android2po to convert between gettext and Android strings.xml (as the feature request to add Android support has been open for some time), and this issue is causing our process to break as it expects an underscore between language and country codes.
Is there as reason pt-BR is different than others? We do not have any aliases set up, and setting one up ("pt_BR") does not resolve the issue.
Using current release branch, v 4.2.2-SNAPSHOT.
It looks like I misread the code when reviewing the PR: it appears that we have option 2, which always uses underscore separation in the Language header (as documented in the Gettext manual).
You should clone our githhub repo, create a fix branch with your changes, and submit github pull request against our zanata-platform repo. Once submitted, one of our developer will review the pull request and merge it when it is done and approved.
Thanks Alex. Can you please clarify what the normal process is for committing/pull requests is? Do I commit on master, or create a separate branch for the changes (as I notice some other ZNTA branches already)? If the latter, I do not have permission to create a branch for this issue.
done. feel free to create a pull request for the change. One of us will review and approve the pull request.
Can somebody please assign this to me? I'm going to put some time into fixing.