"Fix text" making false positives in Spanish spa_OCRFixReplaceList_User.xml. -- Solution given in ticket.

[fraternl]

"Fix text" turns embedded l (right) into i (wrong) in Spanish.

I turned off my own spa_OCRFixReplaceList_User.xml to make sure it wasn't my own regex.
It would be nice if you could see it.
For instance to make User separate.

[fraternl]

It's this regex in spa_OCRFixReplaceList.xml

    <!-- "i" en vez de "l" en terminaciones «clón» -->
    <RegEx find="clón\b" replaceWith="ción" />

better to lose the regex.

ciclón — cyclone
anticiclón — anticyclone
canclón — a regional word
choclón — regional
carranclón — regional

[fraternl]

Instead of dropping the rule, you can make it narrower and replace it with this.

<!-- "i" en vez de "l" en terminaciones «cclón» or «uclón» -->
<RegEx find="([cu])clón\b" replaceWith="${1}ción" />

ChatGPT couldn't find words ending with -cclón and -uclón

The rule that's in spa_OCRFixReplaceList_User.xml is making more damage than it solves.
There is also the spell check.

These will not get altered:

ciclón
anticiclón
motociclón
biciclón
triciclón
siclón
teclón
chanclón
chiclón
choclón
aclón
ñoclón
sonclón
reclón
ranclón