PIETRO SFERRINO

I think a bit of regex will make its magic:

1. In Trados Studio, export the TM to TMX
2. Install Notepad++ or any decent editor.
3. Open the TMX with Notepad++. Be patient…
4. In Notepad++, go to the end of the file and select the wrong text. Find the TU corresponding to the segment 20188 and then select until the end of the file.
5. Press CTRL+H to bring the Replace dialog box up.
6. Tick In selection check box.
7. In Find what paste this regex:
   (<tuv xml:lang="en-US">\r\n\s+<seg>)(.+?)(</seg>\r\n\s+</tuv>\r\n\s+<tuv xml:lang="ru-RU">\r\n\s+<seg>)(.+?)(</seg>)
8. In Replace with, paste this regex:
   \1\4\3\2\5
9. Tick the Regular Expression radio button
10. Press Replace button to replace 1 by 1, or Replace All button to replace all in the selected text.

☛ You may need to change the language codes (en-US or/and ru-RU) to match the exact language combination of your TM.

☛ Don't forget to tick In selection check box or Notepad++ will replace all the TUs.

Once the replacement is done:

• save the TMX
• create a new empty TM in Trados Studio
• import the TMX to the new TM
• check it that everything looks good!

PIETRO SFERRINO

I think a bit of regex will make its magic:

1. In Trados Studio, export the TM to TMX
2. Install Notepad++ or any decent editor.
3. Open the TMX with Notepad++. Be patient…
4. In Notepad++, go to the end of the file and select the wrong text. Find the TU corresponding to the segment 20188 and then select until the end of the file.
5. Press CTRL+H to bring the Replace dialog box up.
6. Tick In selection check box.
7. In Find what paste this regex:
   (<tuv xml:lang="en-US">\r\n\s+<seg>)(.+?)(</seg>\r\n\s+</tuv>\r\n\s+<tuv xml:lang="ru-RU">\r\n\s+<seg>)(.+?)(</seg>)
8. In Replace with, paste this regex:
   \1\4\3\2\5
9. Tick the Regular Expression radio button
10. Press Replace button to replace 1 by 1, or Replace All button to replace all in the selected text.

☛ You may need to change the language codes (en-US or/and ru-RU) to match the exact language combination of your TM.

☛ Don't forget to tick In selection check box or Notepad++ will replace all the TUs.

Once the replacement is done:

• save the TMX
• create a new empty TM in Trados Studio
• import the TMX to the new TM
• check it that everything looks good!

Wooow. That's looks cool ... Thank You Jesus! I'll try as I'll have a bit of time and I'll let You know.
Have a great day!
Pietro 

Hello Jesus!

Soooo ... Such a magic stuff ...
Althought ... I am not sure things are going well ... so, even before to try an import I'd like to ask you tour opinion ...

1.
I have found the place where languages are interchanged ... between <tu> tag at row 243521 (wich ends with the </tuv> tag at row 243528

2.
went down to the end of the file, and done as You wrote ...

3.

Now, MY BIG DOUBT IS ... What is that /1/2/3/4/5 I get on the results ... ???

It simply seems to me I have lost my segments ...

Hope You can help.
All this is really SUPER interesting ... But I wish it will work too!

Many thanks.
Pietro

PIETRO SFERRINO

\1 to \5 are regular expression backreferences. They each backreference one of the parenthesis pairs in the regex Find. MOre info here: https://www.regular-expressions.info/replacebackref.html.

You don't have to see \1\2\3\4\5 in the replaced text. If you do, you haven't ticked the Regular Expression checkbox. If your Editor doesn't allow regex replacements, please install Notepad++.

And ... Apart from my previous reply, wich is the most important one ... if I want to check others TM ... is there a fast way to SELECT by origin ???

Look at this screenshot, for instance, please ...

This should be a TM from russian to english ... But, again ... at a certain point ... there is a mistake ...
How can I check this straight from the Studio NOT looking at the TM page by page ???

I suppose I can use the SEARCH field somehow ... right?

MANY thanks one more time!
Pietro

Ohhhh ... I see ... Good yo know that's not what I should see ... I'll try on Windows with NotePad++ and I'll let You know ...

It is strange that ATOM does not allow to change that. Of course I did use the RegEx option, as shown here ... but it seems that is not good in that ...

I'll try again ...

PIETRO SFERRINO

For example, if you need to filter TUs with Cyrillic characters in the Target, insert this regex:

\p{IsCyrillic}

in the Target Text field and tick Use Regular Expressions in the TM view.

PIETRO SFERRINO

OK, the reason is that Atom doesn't use the same regex flavor as Notepad++.

With Atom, you need to use $1$4$3$2$5 instead of \1\4\3\2\5. I can't test it, but it should work.