src/services/speech/text_to_speech.py

import base64
import random
import re
from io import BytesIO
from typing import List, Generator

from gtts import gTTS
from gtts.tokenizer import pre_processors


class TextToSpeech:
    def __init__(self):
        self.preprocessing = [pre_processors.tone_marks, pre_processors.end_of_line, pre_processors.word_sub,
                              pre_processors.abbreviations]

    def _convert_sentence(self, text: str, lang: str, tld: str) -> bytes:
        tts = gTTS(text=text, lang=lang, slow=False, tld=tld, pre_processor_funcs=self.preprocessing)
        mp3_fp = BytesIO()
        tts.write_to_fp(mp3_fp)
        mp3_fp.seek(0)
        return mp3_fp.getvalue()

    def _split_corpus(self, corpus: str) -> List[str]:
        sentences = re.split(r'(?<=[.!?]) +', corpus)
        return sentences

    def sentence_audio_generator(self, paragraph: str, lang: str, tld: str) -> Generator[str, None, None]:
        sentences = self._split_corpus(paragraph)
        for sentence in sentences:
            mp3 = self._convert_sentence(sentence, lang, tld)
            base64_audio = base64.b64encode(mp3).decode("utf-8")
            yield base64_audio