Linn Duncker ,

When you analyze XLIFFs, do you have this checkbox Report internal fuzzy match leverage ticked:

Yes, I do have that box ticked. 

Actually, I think the problem is that the fuzzy matches are imported as 100 % matches or context matches.

Linn Duncker ,

Ok, so fuzzies don’t come from internal matches.

Linn Duncker said:

I get no fuzzy matches when analyzing files

You confirm that opening the report that is at the top of the list on the Report view, right?

And 0 fuzzies means 0 words in these bands, right?

Could you confirm in the report that the right TMs are attached to your project:

Linn Duncker said:

Still I can see in the bilingual file that there are several fuzzy matches.

You checked this point in the Editor, right? I guess no TMs have been detached or disabled meanwhile, right?

I guess the Fuzzy bands are the default ones, right?

Last, and before I run out of ideas, could you please confirm that your Trados Studio software is updated? It should be SR2, CU12, 9338.

Thanks for the answers! So I did an update, this is the version I´m currently running: Trados Studio 2021 SR2 - 16.2.12.9338.

And yes, I use the report at the top, I have checked that I have the right TM, the Fuzzy bands are default, but still there are no fuzzy matches between 50-99...

This is the clients report and my report of the same file. So I get no fuzzy matches at all.

When I import the XLIFF into the TM I tick Draft to get the fuzzy matches into the TM. Right?

Could this have something to do with the fact that I work with two language pairs, Finnish -> Swedish (Finland) and Finnish -> Swedish (Sweden)? This has caused confusion before. The XLIFF I want to import is in Swedish (Sweden) while the TM seems to be set to Swedish (Finland). Problem is I´m not sure what to do about it. I do have the two language pairs installed in Studio.

Linn Duncker ,

I didn’t know of this step. :-) You don’t need to do this step, you can just open the XLIFF in Trados Studio and translate or review it.

Anyway, doing the import that way, you are not importing draft segments and that may explain the difference if you don’t have the most up-to-date TM.

Linn Duncker ,

Linn Duncker said:

TM seems to be set to Swedish (Finland)

You can confirm the language combination. To do so, open the TM in the Translations Memory view. Then right-click and select Settings.

To import an XLIFF file, its language combination should match the one of the TM you are importing to. Otherwise, there will be 0 segments imported.

If you confirm language combinations differ and you really need to import the XLIFF, then you’ll need to change language in the TM. Basically you export your TM to TMX and carefully you replace the old language code with the new one (there will be an instance per translation unit saved in the TM, so once you know for sure the language code you need, a replace all operation will be very convenient). Then you create an empty TM and import the TMX with the edited language code. Last, attach the new TM to your project.

If you don’t need to import the XLIFF to the TM, I’d use the AnyTM app, free from the SDL AppStore. This app let’s you get matches even if the language combination doesn’t exactly match.

Linn Duncker ,

Linn Duncker said:

TM seems to be set to Swedish (Finland)

You can confirm the language combination. To do so, open the TM in the Translations Memory view. Then right-click and select Settings.

To import an XLIFF file, its language combination should match the one of the TM you are importing to. Otherwise, there will be 0 segments imported.

If you confirm language combinations differ and you really need to import the XLIFF, then you’ll need to change language in the TM. Basically you export your TM to TMX and carefully you replace the old language code with the new one (there will be an instance per translation unit saved in the TM, so once you know for sure the language code you need, a replace all operation will be very convenient). Then you create an empty TM and import the TMX with the edited language code. Last, attach the new TM to your project.

If you don’t need to import the XLIFF to the TM, I’d use the AnyTM app, free from the SDL AppStore. This app let’s you get matches even if the language combination doesn’t exactly match.

Hi!

Thanks for the answers, but this still didn´t resolve my problem. When I get the XLIFF from the client, I always open it in Trados Studio (Editor), and that way I can see all the matches, both 100 % and fuzzy matches. The problem is that I can´t analyze the XLIFF for fuzzy matches, and I need the analyze to give discounts for 85-99 % matches. But as you can see in the attached reports above I get no fuzzy matches in the analyze report. This worked before I updated to Studio 2021.

What I usually do is I create an empty TM (have now checked that the language pairs of the TM match the imported XLIFF, and I have the AnyTM option), and then I import the clients XLIFF into that TM just to make the analyze. But Trados now seems to import all the fuzzy matches as 100 % matches. 

So my question is, why can´t I analyze the clients XLIFF for fuzzy matches (85-99 %)?

Hi Linn Duncker ,

Linn Duncker said:

Trados now seems to import all the fuzzy matches as 100 % matches

That’s the expected behavior. You import an XLIFF to a TM and the matching information is not kept in the TM, just in the XLIFF. Anything imported will be a 100 % match. Draft segments will appeared as 100 % after importing them to the TM, so if I understand your workflow correctly, you shouldn’t tick the Report internal fuzzy match leverage checkbox. Please try unticking it.

Linn Duncker said:

why can´t I analyze the clients XLIFF for fuzzy matches (85-99 %)?

If the above doesn’t help, are you sure you and your client have the same TMs? I mean, same name and same content?

Also there are slight differences in total number of segments, words and characters. Different TMs may explain the difference in segments and words, but I believe that the difference in characters can be explained only by a difference in the document being analyzed…