Mr. Andi,
my person will try to explain so that you possible gain an understanding of th issue (yet not much into this in regard of IT).
The khmer-script (it might be a matter for other scripts as well) uses vocals, vowels and diacritics around a main vowel. And, since sentences are written without visible space it's usual to use unvisible spaces between words to make them searchable. Certain software has problems to display the sometimes complicated arrangement (up to 3, 4 levels around a vowel) of characters or is not able to compensate certain slight errors coming when correcting word-parts afterwards.
One of this apperances is the renderer character to make a footer-vowel out of a vovel. A simple word, "
citta", "cett" spelled in Khmer, would be arranged like this:
i
CT = ចិត្ត typing ច + ិិ + ត + ្ + ត
t
្ the assist-character o
្ renders the vowel ត into a footer under another vowel ត្ត .
Often software is not able to render the display prober and do in some cominations show the "help-character" and the vowel sperate, do not render it into a footer-vowel, as one appearance. As such can also come by "wrong" typing, if looking the "show differences" view (which seems to be not performed to display the script rightly), one can not be sure in regard of this case when using this tool.
My person likes to trust some here is of use for you and your good undertakings so far and trusts that it is understood that he is "not able" avoids to take on merely faith of what is not given, using others resources and register, take part, on such as github or other "open" third part possibilities (enough and not always 100% with conscious to handle with Mr. Andi's and other's in fellow of him gifts here.)
May you also have good mutual success to introduce your work in Cambodia, since it is still very poor, yet also very open in many directions, and undeveloped in regard of accessable software.