Hi Stepan Konev

Here's what I do if I want to split a single translation unit (TU) in a TM...

I create a project containing a simple text file containing the Source TU text that I want to split.

In the Editor window, I split the SDLXLIFF segment into the required amount of segments and then I translate and confirm those segments to the TM. The TM will then contain the number of individual TUs that correspond to the individual segments that I have just translated.

Then I can delete the TU that I wanted to split, either via Concordance in Editor, or with the TM open in the Translation Memories view.

Obviously if you have LOTS of TUs that you want to split, this method may be too time-consuming.

All the best

Ali

Hi Alison Field,
Thank you for your input. However, with this method, you translate the newly split segments from scratch. But I have them all already translated in my TM. The problem is that TUs in my TM go in blocks of segments, 2 to 6 sentences per block (they used paragraph-based segmentation I guess). My client changed their workflow somehow so that they broke those blocks into individual segments, and now I need to "translate" the same text as individual segments.

Ah, I see. Does Fragment Matching not offer you what you need? 

Hi Stepan Konev

I'm sorry... I didn't read your initial post properly...

You can export the TM to TMX then create a project from it using the TMX filetype definition here:

https://appstore.sdl.com/language/app/file-type-definition-for-tmx/317/

That should give you the opportunity to split and merge as you wish, once you have the TMX  open as an sdlxliff.

Then you can create a new TM and import the sdlxliff to that.

You may need to adjust settings to get this working the way you want it, for example export to TMX as 'text only' if the tags are different between the old client format and the new. 

Good luck!

Ali

Oh, that looks promising and easier than aligning. Great! Thank you.