lostecho/edge-tts
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Improve message handling

Browse Source
This commit is contained in:
rany
2021-06-06 20:12:49 +03:00
parent 65cb204f97
commit 9f8a965a1f
2 changed files with 15 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
setup.cfg
View File

@@ -1,6 +1,6 @@
[metadata] [metadata]
name = edge-tts name = edge-tts
version = 1.1.3 version = 1.1.4
author = rany author = rany
author_email = ranygh@riseup.net author_email = ranygh@riseup.net
description = Microsoft Edge's TTS description = Microsoft Edge's TTS

23
src/edgeTTS/__init__.py
View File

@@ -63,7 +63,6 @@ def bool_to_lower_str(x): return 'true' if x else 'false'
async def run_tts(msg, sentenceBoundary=False, wordBoundary=False, codec="audio-24khz-48kbitrate-mono-mp3"): async def run_tts(msg, sentenceBoundary=False, wordBoundary=False, codec="audio-24khz-48kbitrate-mono-mp3"):
sentenceBoundary = bool_to_lower_str(sentenceBoundary) sentenceBoundary = bool_to_lower_str(sentenceBoundary)
wordBoundary = bool_to_lower_str(wordBoundary) wordBoundary = bool_to_lower_str(wordBoundary)
logging.debug("Doing %s!" % msg)
# yes, the connectid() in websockets.connect is different # yes, the connectid() in websockets.connect is different
async with websockets.connect( async with websockets.connect(
wssUrl + "&ConnectionId=" + connectId(), wssUrl + "&ConnectionId=" + connectId(),
@@ -80,16 +79,22 @@ async def run_tts(msg, sentenceBoundary=False, wordBoundary=False, codec="audio-
message='X-Timestamp:'+formatdate()+'\r\nContent-Type:application/json; charset=utf-8\r\nPath:speech.config\r\n\r\n' message='X-Timestamp:'+formatdate()+'\r\nContent-Type:application/json; charset=utf-8\r\nPath:speech.config\r\n\r\n'
message+='{"context":{"synthesis":{"audio":{"metadataoptions":{"sentenceBoundaryEnabled":"'+sentenceBoundary+'","wordBoundaryEnabled":"'+wordBoundary+'"},"outputFormat":"' + codec + '"}}}}\r\n' message+='{"context":{"synthesis":{"audio":{"metadataoptions":{"sentenceBoundaryEnabled":"'+sentenceBoundary+'","wordBoundaryEnabled":"'+wordBoundary+'"},"outputFormat":"' + codec + '"}}}}\r\n'
await ws.send(message) await ws.send(message)
logging.debug("> %s" % message)
await ws.send(msg) await ws.send(msg)
logging.debug("> %s" % msg) download = False
async for recv in ws: async for recv in ws:
recv = recv.encode('utf-8') if type(recv) is not bytes else recv if type(recv) is str:
logging.debug("< %s" % recv) if 'turn.start' in recv:
if b'turn.end' in recv: download = True
await ws.close() elif 'turn.end' in recv:
elif b'Path:audio\r\n' in recv: download = False
yield b"".join(recv.split(b'Path:audio\r\n')[1:]) await ws.close()
# TODO: add some sort of captioning based on audio:metadata. It's just JSON with offset.
# WordBoundary is the only thing supported. SentenceBoundary does nothing.
#elif 'audio.metadata' in recv:
# print("".join(recv.split('Path:audio.metadata\r\n\r\n')[1:]), file=sys.stderr)
elif type(recv) is bytes:
if download:
yield b"".join(recv.split(b'Path:audio\r\n')[1:])
# Based on https://github.com/pndurette/gTTS/blob/6d9309f05b3ad26ca356654732f3b5b9c3bec538/gtts/utils.py#L13-L54 # Based on https://github.com/pndurette/gTTS/blob/6d9309f05b3ad26ca356654732f3b5b9c3bec538/gtts/utils.py#L13-L54
# Modified to measure based on bytes rather than number of characters # Modified to measure based on bytes rather than number of characters