[Please note that some of the entries are very old.*]

Saturday, 1 August 2015

Converting Huge TMX Resources to... TMX Resources

With the free version:
  • Download and install the free version of CafeTran. It’s a ridiculously small download, but you may have to install Java as well
  • Run CafeTran, the Dashboard will pop up
  • Drop a document (e.g. *.docx) or a project (e.g. SDLXLIFF) on the Dashboard, and check the settings for the source and target language
  • In the Menu, select Total Recall | New Memory Table…
  • In the dialogue that pops up, give the table a name, optionally delete “Client” and “Subject,” and check the “Language Column Names”. Click OK.
  • In the Menu, select Total Recall | Memory Tables, and select your newly created table
  • In the Menu, select Total Recall | Cat Tools Exchange | Load from TMX Memory…, and browse to the TMX memory you want to add
  • Select it, and the import will start. This may take a while, because CafeTran will index the table for blisteringly fast search. Not necessary if you want to end up with a TMX file again, but I don’t think you can avoid the indexing part
  • You now have your two languages in a simple table. To turn it into a TMX file again, you'll need a tool to open your SQLite database to export the table to a format that can be imported in a CAT tool, like CSV or Excel, to get a TMX file again. I use SQLite Browser - again free - for that purpose.
Click to enlarge
  • Close CafeTran
  • Download and install the free SQLite browser
  • Browse to your newly created database (in my case here: ~/CafeTran.app/Contents/Java/resources/databases/SQLiteMemoryBase.db) and open it with the browser
  • Click the tab Browse Data, and in the browser menu, go to File | Export | Table(s) as CSV File. That's a format a CAT tool should be able to import in a TMX memory.
Of course, with the paid version of CafeTran, things are a lot easier:
  • Open or create a project in the right language pairIn the Menu, select Memory | Open Memory, browse to your huge TMX file, and double-click it. Only the segments in the selected language pair will be imported
  • See to it that the new memory is active in the tabbed pane by selecting it if necessary, and in the Menu, go to Memory | Save Memory as...
UPDATE: It's actually just as easy using the free version. Just drop the huge TMX file on the Dashboard, a settings dialogue will pop up. Select the language pair you need. In the menu, Project | Convert Project | To TMX Memory... You may want to assign as much RAM as you can in the Menu, Edit | Options | Memory