Strange Context Match behavior

Hi,
We have recently come across a strange behavior of our TM when changing Context Matches.

We use Trados Studio 2015 and the file to translate is a .po file. Our TM was built on the basis of .po files previously translated by another company.

The prepared (pre-translated) .po.sdlxliff file looks as follows in the Editor view:

Trados Studio Editor view showing pre-translated segments with AT flags and a Context Match in the translation results.

The pre-translated segments are assigned the AT flag, which seems normal to me, because they come from a TM built from automated translation files.

In the translation results view, the selected segment appears as a Context Match, which seems normal to me, as well, because the files from which the TM was created have the same (context) structure as the one to be translated.

But when I change one of these Context Matches (I added '_TEST' in the example) and confirm the translation, the Context Match is not updated (which is what I would expect) but two 100% matches are created:

Trados Studio translation results with a segment edited to add '_TEST' not updating the Context Match but creating two 100% matches.

When I change a segment that is marked as a 100% match and confirm the translation, it appears as a Context Match and the former translation still appears as a 100% match in the translation results:

Trados Studio showing a segment marked as a 100% match, edited and confirmed, now appearing as a Context Match with the former translation still as a 100% match.

So if I want to get only one translation per segment (i.e., the changed one), I have to delete the existing one, which is s very annoying and time-consuming.

Is this an expected behavior? Or is there a TM option we overlooked? Exporting the TM content and importing it into a completely new one doesn't change anything.

Thank you  very much!



Generated Image Alt-Text
[edited by: Trados AI at 2:15 PM (GMT 0) on 28 Feb 2024]
emoji
  • Hi ,

    I played around with this in 2015 and 2017, investigating different user IDs and settings. I can reproduce your behaviour exactly in Studio 2015 but not in 2017. I believe this may have been a behaviour that was changed in 2017 onwards and don't expect to see this changed in 2015 now.

    If it helps... and it may not... you could make your changes without updating the TM as you edit the file, and then use the option to overwrite existing TUs when you update your TM through a batch task at the end. That will overwrite them.

    Paul Filkin | RWS Group

    ________________________
    Design your own training!

    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub

  • Hi Paul,
    Thank you for your answer. We have now upgraded to 2017, but it's still the same.
    I will forward your suggestion to the colleagues concerned, but I guess that they would prefer to update the TM directly in order to get the correct match in further occurrences within the file.
    Our current workaround is to use the 'Apply Translation' option for the pre-translated segments. The AT flag then disappears (while the CM or 100% label appears) and the segment can be updated correctly in our TM (a lot of fumbling around, but at least, it works).

    We will keep you informed of any new findings ;-)
  • Hi there, I've been having this problem for years. I was just wondering if 2019 handled this better? Thank you!