process cuneiform signs

[chiarcos]

So far, the analyzer can generate and analyse underlying forms, but not actual written forms (transcripts). This is because we haven't be able to map syllables to characters, yet. A bit tricky because the writing system is defective and not every (assumed) phoneme is written.

[chiarcos]

working with transliterations, now. remaining todos:

• add morphology to experimental/analyze.py
• replace ./analyze.py
• populate signs.tsv