Red Hat Bugzilla – Bug 901401
RFE: Implement support for InDesign documents
Last modified: 2015-11-09 19:10:25 EST
inDesign is a leading Desktop publishing software used broadly in Marketing campaigns for product launches.
Translation of these material is usually done from the inDesign file format to guarantee the same layout in all the languages.
To make possible the use of Zanata to translate this sort of material we need to have support for inDesign.
I've done some preliminary work on this in https://github.com/zanata/zanata/commit/af4405848067b9b78abe7692f5db25d9bafc69b2
but we will need to find a way of handling large files. We could split them into multiple BLOBs, but it might be better if we kept large files out of the DB. (Large files can slow down DB backups, DB storage is more expensive than filesystem storage, etc.)
Also, we should probably implement some way for project maintainers to choose Okapi options, instead of hard-coding them in the adapter class.
Increasing the max blob size on the DB works fine. I tested a with a few real-world InDesign (IDML) files by exporting them from a binary original, pushing to zanata and then exporting to InDesign again.
The process worked smoothly, with a few items we need to look at:
- The order of the entries in Zanata does not seem to match the order in the original InDesign file. The specific example that I noticed had the document's footer near the top of the list which could be confusing.
- There seem to be comments and instructions for the designers in the document itself, which may or may not need to be translated. If they don't need to be translated they might throw off the statistics. So we may need to look at a way to ignore those types of entries.
Regarding the order of the entries in the document, we may not be able to do much about it, short of implementing a preview feature (with a copy of InDesign). As with PDF files, I gather the order shown on the page doesn't always match the order in the underlying file.
We should be able to filter out the designer comments in most cases by modifying the Okapi IDML filter, assuming they have a sane representation in the IDML file.
I found a bit of information about the order problem here: http://stackoverflow.com/questions/8346961/logical-order-of-story-xml-in-idml
Closing VERIFIED bugs which predate Zanata 3.1.2.