[ast] Spelling differences between LibreOffice and desktop/web modules

I found out that LibreOffice LT module works fine with Asturian spelling, but Desktop LT and Web LT language testings misrecognice words with apostrophations, treating them as two independent words, and generating a spelling typo on the apostrophed word. How can this be possible, as all of them use the same binary files? I’m using version 6.4 for LibreOffice. Could anybody give me some tip of what I am forgetting to check? Thanks a lot!
I attach screen captures of sample text from LibreOffice, MS Word and web, with words marked as errors:

LibreOffice

MSWord


.

Hi Xesús.
Please copy some examples here in plain text format that we can use for testing.
Does the problem occur with straight apostrophes, curly apostrophes, or with both?

Hi Jaume,
Of course, here is a text specially arranged to show all possible apostrophations in Asturian (except for “que”, which for the moment is not implemented because of space-saving and performance purpose). The problem occurs with both apostrophes (straight and curly).
The text is orthographically and grammatically correct. On LibreOffice, this text doesn’t generate any error. But does in Desktop LT and on the website: m’aguanto (me), t’aguantes (te), s’aguanta (se), d’esi (de), n’espresiones (en), revisase’l (el, backwards), l’usu (el, onwards), p’averase ( pa).

Nun m’aguanto pa cuntá-yoslo. Nun t’aguantes nin a ti mesmu. Nun s’aguanta per ella sola. Una de les funciones d’esi enantar ye la de llocalizar el productu. Podemos apurrir el programa con too tracamundiao n’espresiones regulares. Con cada versión nueva que sal, vuelve revisase’l testu orixinal. Otru aspeutu que pescudar ye l’usu didáuticu que se-y pue dar nel enseñu. Marchó coles oreyes gaches p’averase a la ponte.

One more thing. Reviewing this issue, I’ve found out, too, that this apostrophed word pairs are marked on LibreOffice as typos if they are at the bottom of the phrase (followed by a full stop), but not in the middle (followed by space or comma). Does that behaviour is related to this issue, perhaps? Here is a text to check that:

Como davezu, el rapazu tenía la mano alzada n’aire. Caltién n’aire la zapica.

The first “n’aire” (followed by a full stop) is marked as typo. The second “n’aire” (followed by a space) does not.