I was only able to get so far. It’s just interesting what the outcome turned out to be. Some words are recognized, while others are not. I think this may have something to do woth the way the encoding is exercised while creating the file from a word list.
I downloaded some of the dictionaries from this github repo
And I created my own word list and named the file myapell.dic
On Aspell Creating an Individual Word List it says to specify the base on the current directory, or else it will create the file on the word list dictionaries at /usr/lib/aspell/.
I had made the mistake of creating the master file without the prefix, and it wasn’t until later when I realized it.
The reason was that I had come across this
post from linuxquestions.org in which the op didn’t specify it. But it’s fine. Since I could always remove it.
$ aspell --lang=es --encoding=utf-8 create master mydic.rws <
Then back on spell.kak
to line 31 and 113 of the file.
But unfortunately, some of the words are not even recognized. e.g., thistle and thistles are fine, third and thirds are not
I also read some of the explanation given by the chromium Project about some of the rules that are in use with the
.aff files fromHunspell. See
This should apply as well to Aspell
.aff files and
.dat files from Aspell.
But I also tweaked around with
dump master while trying to modify the word list from Aspell itself, especially after reading
https://superuser.com/a/137990 , Aspell seemed to depend largely on the base dictionary itself.
So while creating the file from say both en and es dictionaries,
--lang applies to only one. The end result is that it recognizes only one dictionary.
One of the hints by which I have relied, is by having (without modifying any of the dic files) under English the word ‘thirds’ and in Spanish the word ‘perseverarías’ , both properly recognized.
If anyone has a working solution by going this route, please let me know.