From aa902aba30857b8a543614ea2f0f0925abb7762a Mon Sep 17 00:00:00 2001 From: rany Date: Tue, 11 May 2021 00:38:46 +0300 Subject: [PATCH] Use NLP instead of Regex to get each sentence --- edge-tts.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/edge-tts.py b/edge-tts.py index f149974..ec2efa4 100755 --- a/edge-tts.py +++ b/edge-tts.py @@ -5,6 +5,7 @@ import uuid import argparse import urllib.request import websocket # pip install websocket-client +from nltk.tokenize import wordpunct_tokenize from xml.sax.saxutils import quoteattr as escape try: import thread @@ -81,7 +82,7 @@ if __name__ == "__main__": volumeString = args.volume sentenceBoundaryEnabled = 'True' if args.enable_sentence_boundary else 'False' wordBoundaryEnabled = 'True' if args.disable_word_boundary else 'False' - text = args.text.replace(chr(9), " ").replace(chr(13), " ").replace(chr(32), " ") - run_tts() + for text in wordpunct_tokenize(args.text.replace(chr(9), " ").replace(chr(13), " ").replace(chr(32), " ")): + run_tts() elif args.list_voices is True: list_voices()