CafeTran

I've found a workflow that enables me to extract only the segments with the status "Not Translated" from a Transit NXT project, in order to translate them in CafeTran.

It's not perfect, since it requires a change of file format to TMX. But it's better than nothing ;).

You need Transit for this.

First: Filter on segments with the status "Not Translated".

Then assign the segment status "Checked 2" to the displayed segments only.

In the button bar below, choose:

Then:

Make sure that "Also export unaltered segments" is checked.

Translate the TMX in CafeTran. Import it Transit. You'll have to put the dots at the i.

Back in Transit:

Import the TMX file, Transit will create the corresponding language pairs:

Add the folder with the re-imported language files to the Project's reference folder (Project settings, Tab Ref).

Filter on Checked 2.

Set status to Not Translated.

Start translating and notice that all segments that you've translated in CafeTran will be taken from the TM.

Note that you'll probably have to adjust some markups. Hey, I didn't say that it was perfect ;).

See also:

http://www.proz.com/forum/transit_support/317977-how_to_insert_a_string_at_the_start_or_end_of_all_segments.html

Which you can use for an alternative workflow: add the string "Not Translated" to all untranslated segments in the target language file and translate this in CafeTran (in CafeTran, filter on the string "Not Translated").

Of course, this can "spoil" your TM with the  target language-target language segments.

The alternative workflow (adding a unique string to untranslated segments in Transit, filter on this string in CafeTran) will have the advantage over the TMX workflow that you can insert all tags (in Transit: markups) in CafeTran. So when you export your Transit language file from CafeTran, you can directly use it in Transit. No tedious corrections of markups necessary.