[Feature Request] Arguments

[Qwerty-Space]

Hello,

I'd like to request arguments for phones_for_word such as stress_marks=False. That would be brilliant.

And also the ability to convert multiple words in a single string. Perhaps a for_phone_in_words function

[aparrish]

These are great suggestions, thank you! In the first case, it's easy enough to implement on your own, something like:

>>> import pronouncing, re
>>> [re.sub(r"\d", "", item) for item in pronouncing.phones_for_word("perfect")]
['P ER F EH K T', 'P ER F IH K T']

In the second case, I'm hesitant to introduce word tokenization logic into the library, since it's such a can of worms, and I feel like most of the bug reports I would get would be about the word tokenization algorithm being wrong or insufficient (instead of what the library is supposed to be about, e.g., getting pronunciation information about words). I'd encourage users instead to make a decision about their own method of tokenizing words for their particular use case and work from there. E.g., with nltk:

>>> import nltk
>>> ' '.join([pronouncing.phones_for_word(item)[0] for item in nltk.word_tokenize("this is a test!") if item in pronouncing.lookup])
'DH IH1 S IH1 Z AH0 T EH1 S T'

Though on second thought that is kind of a mouthful. I'll take it into consideration and would also be happy to consider a pull request with this feature...

[Qwerty-Space]

Thanks for the fast reply!

I only ask because I'm trying to write a syllable cipher, and ARPA seems like the easiest system to use for it.

So I'd like to make an (en|de)cryption program for it.

And sorry, I'm don't think I'm good enough at python yet to make a PR like this.

[Acryptarch]

Is this module still being developed? I feel like I could implement the feature described

[aparrish]

Is this module still being developed? I feel like I could implement the feature described

Yes! It is still being developed. As noted above, I'm willing to entertain PRs for these features, but I'm more inclined to leave these as downstream implementation tasks for users of the package, rather than incorporating them into the package itself.

[Acryptarch]

That's a fair point. What other features do you think would be useful? Maybe an implementation of the logios tool in python would be a good addition to this library.