xliff file imports, but removes all text boxes

We are five years later, but it seems the bug persists in Rise:
If an xliff includes <mrk> tags, the text content between those tags is simple ignored and the element of the course in which this text should go seems ignored.
I noticed that with the course title, I did not dig further.

for the records, the exported xliff include all content as this sequence of tags:
<trans-unit><source>[content in original language]</source></trans-unit>

The translators add translations like this

<trans-unit>
  <source>[content in original language]</source>
  <target>[content in target language]</target>
</trans-unit>

But some translators transform the xliff code the following way, compatible with the norm xliff 1.2:

<trans-unit>
  <source>[content in original language]</source>
  <seg-source>
    <mrk mtype="seg" mid="1">[content part 1]</mrk>
    <mrk mtype="seg" mid="2">[content part 2]</mrk>
  </seg-source>
  <target>
    <mrk mtype="seg" mid="1">[translated content part 1]</mrk>
    <mrk mtype="seg" mid="2">[translated content part 2]</mrk>
</target>
</trans-unit>

Indeed the file works again if the <mrk> tags are removed. But it is not clear if some space should be added if the slicing occured int the middle of a sentence.
the encoutered mrk are incremented in the 'mid' parameter throughout the file. Not clear what it is used for.

The best would be that those file could be imported a such in Rise without modification...