Hello Greg, thanks for your help.
It works greatly

However, I exaggerated a bit with the simplification:

what I called "first word non in italic" is actually a combination of 1-n words going from the first word non in italic to a special character (§) that I can easily add to the original text by find and replace.

That's beacause my first idea was to find a way to add "§§" between the first and the second word of each line only when the second word is not in italic.
Then I could convert text into table parting columns with §, and get an empty cell when there is not a second word in italic.

Also, the table should have: a fourth column whose content can be defined in one of the following ways (they will give the same results, choose the simplest to code)
- the word comprised between the previous § and the next §
- the first 4 characters after the previous §

The rest of the line will be contained, unbroken, in a fifth column

Thanks for your suggestions