Description of problem: This bug is relates to the download portion of https://bugzilla.redhat.com/show_bug.cgi?id=725319. Language team translators should be able to download PO files for each individual document or the whole project. Version-Release number of selected component (if applicable): 1.4.x
Added a "Files" section where individual files (po only at this time) can be downloaded. There's also an option to download all files for the iteration / language. Merged into 1.5.x tracking branch.
Some po files appear to have the content type charset in the header set to "CHARSET", whereas others have the correct value "UTF8"
The charset=CHARSET problem is fixed in bug 748727. There's a few more changes for this story in https://github.com/zanata/zanata/commit/a9ee3c796325c155070442de0450b2296071deca and https://github.com/zanata/zanata/commit/9032882c3f5f8f2f3c41c7202c762f12d3bdf976. But we do need a slight change to the algorithm which generates some of the PO headers: if you upload a PO, *don't* change anything via the webtrans editor, and then download the PO, you should get back the lastTranslator/date that was uploaded.
The system will now check for the original (as in when pushed) PO-Revision-Date header value. If this date is more current than any of the 'Last Changed' dates in any of the translations, the PO files produced will keep the original values for PO-Revision-Date and Last-Translator header values. Otherwise, the system will generate values according to the internal data. For changes see: https://github.com/zanata/zanata/commit/8ab71d1a3b2b2d7d72d7cc195ac5ffa466372912 https://github.com/zanata/zanata/commit/8376864ffa0045e6f28027b069761d260fc6d3be
Was reporting revision date and last translator based on last approved string in document, rather than most recent approved date. Fixed in https://github.com/zanata/zanata/commit/73449eed5cf276a2b4ac1b654563aaf443af5143
verified in 1.5