I already see two problems:
- It does not differentiate proper noun and normal noun. Example: ‘train’ is both the vehicle
\tʁɛ̃\and a city in germany
- It extracts old french pronunciation. Example: ‘de’ was pronounce
\dam\in old french, but it should be pronunces
For proper noun, we could either not extract proper noun pronunciation at all, or match the case, and keep the capital letter.