Alan Twigg

I think this could be possible using the new Multlingual XML filetype.  For example, I can take this file I created:

4722.cells.docx

Then unzip it and tackle the underlying XML like this:

This should get me this:

Which, if the file remained consistent in the format you provided may be an elegant solution.  I think I could even exclude the grey shaded cells with a little work as the XPath starts to get a bit complex... but didn't go that far yet as I just wanted to prove the concept.

However... I cheated a bit here because my preview is actually based on an XML filetype rather than the Multilingual one.  I seem to have a problem with namespaces which I'll have to address with the developer on Monday.  But XML did allow me to prove the idea at least:

So perhaps more on this on Monday after I get to the bottom of this.  But an interesting one anyway!

Alan Twigg

I'm not sure if you are still interested in this... but I was ;-)

I got a lot further using the multilingual XML filetype so I've created a short video just to explain to you properly how this would work as it is a new filetype and deserves some explanation:

Alan Twigg

I feel like I'm in this alone now ;-)  However, it's been an interesting exercise and we have made some improvements to the plugin to be able to handle the namespaces correctly, and I also enhanced the XPath query to only pick out the rows that needed to be translated, so I exclude the grey cells.  In fact I enjoyed it so much I used this as an example of how this new filetype is so incredibly useful:

https://multifarious.filkin.com/2021/12/01/so-simple-yet-so-useful/

I hope you are able to come back at some time as you seemed to have given up on this interesting task... but perhaps this solution is something you can use.