zompist bboard

So, I have a rule:

tis/c/V(C)_#

This rule requires two categories:

C=mnptḱkǩbƂdƋǵgĝğǥs₁₂₃rlywʦcčjǰĵɉšṧfvʎ
V=eéēḗoóōṓuúūùiíīìṛṝḷḹṃṇaáāà

Let's use this as our example input:

pótis

If we put all of this into SCA²...

pótis

It doesn't work. /tis/ does not become /c/ as the rule clearly states it should.
However, if we take the C line out of the categories...

póc

Suddenly, it does work.
Clearly, there's something wrong with my C category, like maybe a diacritic that's actually a separate character.
Or maybe it's in the V category? Yet the rule works with V alone.
Anyways, there isn't any such problem. I checked them all, they're all single characters. No arrant character combining here, to my knowledge.
Even if the damn thing was somehow interpreting ó as two characters, the vowel o and the diacritic, the rule should still work, whether it's interpreting the diacritic as part of V or C.
I suppose if it is interpreting it as two characters, it could be not registering the diacritic as part of either V or C, but that's ridiculous, because the diacritic appears on letters in both categories!

So yeah, I think I found a bug.

Matrix wrote:So, I have a rule:

tis/c/V(C)_#

This rule requires two categories:

C=mnptḱkǩbƂdƋǵgĝğǥs₁₂₃rlywʦcčjǰĵɉšṧfvʎ
V=eéēḗoóōṓuúūùiíīìṛṝḷḹṃṇaáāà

Let's use this as our example input:

pótis

If we put all of this into SCA²...

pótis

It doesn't work. /tis/ does not become /c/ as the rule clearly states it should.
However, if we take the C line out of the categories...

póc

Suddenly, it does work.
Clearly, there's something wrong with my C category, like maybe a diacritic that's actually a separate character.
Or maybe it's in the V category? Yet the rule works with V alone.
Anyways, there isn't any such problem. I checked them all, they're all single characters. No arrant character combining here, to my knowledge.
Even if the damn thing was somehow interpreting ó as two characters, the vowel o and the diacritic, the rule should still work, whether it's interpreting the diacritic as part of V or C.
I suppose if it is interpreting it as two characters, it could be not registering the diacritic as part of either V or C, but that's ridiculous, because the diacritic appears on letters in both categories!

So yeah, I think I found a bug.

I seem to recall that brackets can be an issue, particularly before the _. Have you tried:
tis/c/VC_#
tis/c/V_#
?


On the bug front, though: rules that result in introducing new characters into strings under certain conditions often seem to only introduce the character the first time that the conditions arise in the string, not throughout the word. But repeating the rule makes it work, even when it "should" just result in continuing to double and triple etc the character at the first instance. Other times, though, the rule works fine. I've given up on trying to work out why and just keep repeating any sort of rule like that and then if necessary using rules to delete accidental geminate.

Salmoneus wrote:I seem to recall that brackets can be an issue, particularly before the _. Have you tried:
tis/c/VC_#
tis/c/V_#
?

Alright, that works, thanks!