Don't know what phones belong to which word

bootphon / phonemizer

Simple text to phones converter for multiple languages

https://bootphon.github.io/phonemizer/

GNU General Public License v3.0

1.23k stars 172 forks source link

Don't know what phones belong to which word #56

Closed RuABraun closed 3 years ago

RuABraun commented 4 years ago

I wanted to get pronunciations for a vocab of 200k words, but after calling phonemizer the output is 199999 lines long. So I can't just paste the files, because that would lead to a mismatch.

I don't understand why the output does not have as the first column/field the word ?

mmmaat commented 4 years ago

(a hello or a thank you is always welcomed, I'm a human)

Hi,

Are you sure you don't have an empty line in your output file? You should have one.
If not this is a bug and I'll need more information : the exact command and parameters you used, the output of phonemize --version and, if you can find it, the missing word in output...

I don't understand why the output does not have as the first column/field the word ?

We generally apply the phonemizer to whole sentences, not single words.

mmmaat commented 4 years ago

You should have an empty line because sometimes espeak (I'm supposing you're using espeak backend) drops words. In that case there is nothing we can do at phonemizer level, this is an espeak bug.

RuABraun commented 4 years ago

Sorry for being rude.

Thank you for the quick response, will respond tomorrow (just left work) with more information.