src/split_audio/text_utils.py

# IPA Phonemizer: https://github.com/bootphon/phonemizer

_pad = "$"
_punctuation = ';:,.!?¡¿—…"«»“” '
_letters = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'
_letters_ipa = "ɑɐɒæɓʙβɔɕçɗɖðʤəɘɚɛɜɝɞɟʄɡɠɢʛɦɧħɥʜɨɪʝɭɬɫɮʟɱɯɰŋɳɲɴøɵɸθœɶʘɹɺɾɻʀʁɽʂʃʈʧʉʊʋⱱʌɣɤʍχʎʏʑʐʒʔʡʕʢǀǁǂǃˈˌːˑʼʴʰʱʲʷˠˤ˞↓↑→↗↘'̩'ᵻ"
_extend = "-124567̪"

# Export all symbols:
symbols = [_pad] + list(_punctuation) + list(_letters) + list(_letters_ipa) + list(_extend)

dicts = {}
for i in range(len((symbols))):
    dicts[symbols[i]] = i

class TextCleaner:
    def __init__(self, dummy=None):
        self.word_index_dictionary = dicts
        # print(len(dicts))
    def __call__(self, text):
        indexes = []
        for char in text:
            try:
                indexes.append(self.word_index_dictionary[char])
            except KeyError:
                # print(text)
                print(f"Unknown character {char} in text. Please add this character and extend the dictionary layer of the model.")
        return indexes