Feature idea: extract text

I am playing with the idea of extracting the texts from the SVG file to provide glossaries for language learners. We might also use this as an accessibility feature here. So, I have started writing a python script.

For dictionary access, we could simply add a Wordlink button, and presto pronto, instantly clickable text with a dictionary! It looks like this:

http://multidict.net/wordlink/?url=https%3A%2F%2Fwww.peppercarrot.com%2Fen%2Fstatic8%2Fwiki&sl=en

If there is no interest for this feature here, I can always host a copy of the Gaelic files on my own page - I am running a Gaelic Learner's site and I think that it would be a fun resource.

Now to the technical bits:

I can't get who is speaking and the order of elements from the SVG, neither can I guess how to combine multiple text elements for sounds etc. So, I have come up with the idea of creating a reference file for the en locale. So, for the English reference file

Pepper|1|1|...and the last touch
Pepper|3|1|"...mmm probably not strong enough"
Sound|2|2|SHH SHH
Sound|4|2|PLOP PLOP

I am getting

Pepper|1|1| ...agus beagan gleus oirre
Pepper|3|1| ...cha chreid mi nach eil i caran lag fhathast
Sound|2|2| SSS SSS
Sound|4|2| PLUB PLUB

File format is: Speaker|display order|segments combined|text

Now, all I need to do is to parse the LibreOffice Calc file to get the names translations, and then the result still needs to be reparsed for whatever output we want.