Dividing a word into syllables

[peregilk]

Is it possible to use the library for dividing words into syllables?

[razzius]

Based on my understanding, words are divided into syllables by vowels, as you can see in these 2 examples:

>>> import pronouncing
>>> pronouncing.phones_for_word('screeched')
['S K R IY1 CH T']
>>> pronouncing.phones_for_word('syllables')
['S IH1 L AH0 B AH0 L Z']

Since the phones in CMUdict (which this project uses) start with vowels if they themselves are vowels, you can come up with a function like so:

def phone_is_vowel(phone):
    return phone[0] in 'AEIOU'

And split words using an approach like follows:

def split_on(condition, l):
    result = []
    current = []
    for element in l:
        current.append(element)
        if condition(element):
            result.append(current)
            current = []
    return result

def split_into_vowels(phones):
    return split_on(phone_is_vowel, phones.split())

Finally, since a single word can have multiple pronunciations, you'll have to deal with that somehow:

def split_word_into_syllables(word):
    # Naive approach: just take the first pronuncation
    phones = pronouncing.phones_for_word(word)[0]
    return split_into_vowels(phones)

Putting it all together you get something like so:

>>> split_word_into_syllables('history')
[['HH', 'IH1'], ['S', 'T', 'ER0'], ['IY0']]

However depending on your use case, you might have to consider things like which vowel/syllable the consonants belong to.

You can see an example of this kind of use case in this library which can be used to find multisyllabic rhymes: https://github.com/Tareq62/multisyllabic-rhymer/blob/master/multisyllabic_rhymer.py#L25