Open temainowu opened 10 months ago
I am fairly sure that the onset function now works great, it finds the element with highest sonority and checks if there is anything before it, this doesn't fix all the issues but it is improvement. also the function will break if given empty syllables.
a few months ago, I was wanting to include syntax trees in hopes of using syntactic constraints to analyse full sentences. alas this did not come to pass, but if I did manage this, then extending tree representations to syllable structure would likely fix this issue.
no one knows what a syllable is so it is really hard to tell haskell what it is, the onset constraint thinks that "sto" has no onset because it starts with a high sonority, lowers in sonority and then rises again which is a similar pattern to "ots" which it correctly identifies as having no onset (and thus issues a violation for it). I need to start considering the magnitude of the segment 's phonologies rather than just the general shape of the syllable 's sonority contour. (I know that noCoda has issues that are specifically because of sizeOfCoda but I haven't tested it nor thought about what they would be so idk what they are)