The contents in square brackets [] are NOT tags - so you need to parse them manually.

Yes, this is another issue. What I am concerned now is that the CDATA content is not segmented separately but as a whole segment. Previously I got this result:

But now I am not able to reproduce it...

Here is another example with incorrectly parsed embedded content:

Sotir Rangelov

Perhaps check that you are really using the filetype you think you are?

Indeed, this was the issue! I had to disable the other xliff filters:

Thank you very much, Paul!

Sotir Rangelov

Excellent... sometimes the solution is right under your nose :-)

Would you please advise how to do this - parse them manually the content in square brackets?

You have to add some regex to your embedded content processing. As you have used HTML5 as embedded content processor, I'm afraid it wont be possible. In this case you could use the app Cleanup Source and add the regex for square brackets there.

Or you develop a new xml file type based on legacy embedded content and add the tags and square brackets there one by one. Not so simple, but this way you can get also content like this [textarea textarea-249 placeholder "Съобщение"] processed. However, giving all the necessary regex here in the forum will be a very demanding task.

Thank you, @Jerzy! I will try this out.