Converter to sound classes

[PhyloStar]

IPA to ASJP, IPA to SCA. Add a separate class for converting the original IPA data to sound classes.

[Anaphory]

While it isn't a class, there is the ipa2asjp module, which would in principle be the right place to have those functions. (and ipa2asjp and ipa2sca exist in there, as well as effectively two ipa2ipa functions for tokenizing IPA strings cleanly, with and without removing syllable boundary markers etc.)

On the other hand, the readers in read_data_cldf are currently hard-coded for doing either IPA or ASJP transcription, and the logic is to take it from a column, with a fallback of automatically converting IPA to ASJP. I'm not sure what the best logic for those functions would be – what do you think of supplying them with a column name ('IPA' or 'ASJP') and a segmenter (lingpy.ipa2tokens, ipa2asjp.tokenize_word_reversibly, ipa2asjp.ipa2sca or None for “Just use the string as a sequence”)?