Sotir Rangelov

You could set the segmentation hint for br tags to "Exclude" and this may be somewhat better (apologies for the Bulgarian... a sample file would have been useful ;-)):

I used a file like this:

<?xml version="1.0" encoding="UTF-8"?>
<rootelement>
  <to>офис.<br class="xlf-newline" /><br class="xlf-newline" />отправить оплату за товар в офис по адрес:<br class="xlf-newline" /><br class="xlf-newline" />бул.</to>
  <to>6 септември, № 152, партер, офис - 8<br class="xlf-newline" /><br class="xlf-newline" />ап.</to>
  <to>Пловдивски 4000, България</to>
</rootelement>

But even this isn't really what you want in a perfect world.  But then I don't know what your XML looks like.  I can get to this:

If the XML is like this I can use an exception to the full stop rule and achieve this:

<?xml version="1.0" encoding="UTF-8"?>
<rootelement>
  <to>офис.<br class="xlf-newline" /><br class="xlf-newline" />отправить оплату за товар в офис по адрес:<br class="xlf-newline" /><br class="xlf-newline" />бул.</to>
  <to>6 септември, № 152, партер, офис - 8<br class="xlf-newline" /><br class="xlf-newline" />ап. Пловдивски 4000, България</to>
</rootelement>

If none of this helps perhaps you can share a small sample like this so we can see what you're working with?

Here is my xml file, which is an xliff file from WPML.

job-43.zip

I created a custom XML filetype filter as per your instructions in youtube. No special settings, only Note is set to untranslatable:

So how to "set the segmentation hint for br tags to "Exclude"? Where should I do this - in the Translation memory settings?

Another issue is that the Title and href attributes of the A element are not translatable:

Whereas I must translate them... Do you have any suggestion how to correct the filter to be able to translate the title and the href within the link?

Sotir Rangelov

ok... I think you need to wind back a little first because I'm not sure you have grasped the task here.  You have created your filetype translating everything in it apart from the note.  This is clearly incorrect.  An XLIFF is a bilingual file so you should only be translating the target element, maybe something like this:

Note I disabled everything, added a rule to make sure nothing was translated and set the target as always translatable.

Next I set the embedded content processor to handle the CDATA sections using the HTML processor like this:

This alone gives me this preview:

This looks fairly neat and the only inline tags are the break tags.  If these tags should be excluded I can do this to get this:

Maybe better... you'll be a better judge of this than me.  But to do this you need to remember that these tags are being handled by the embedded content processor, in this case the HTML 5.2.0.0 processor.  So you need to go to this processor:

Select the br parser rule and change it to reflect the change you want... in this to set it as structure so that you break at every br tag:

So in this case I didn't bother setting the segmentation hint, I just made the rule structure which achieves the same thing.

See if this helps you?

Thank you, Paul. This works as expected!
I also managed to edit the parser so that I am able to translate the URLs of the links

I marked href for translation and prepared the file again.