engineering/uncloseai-speech
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Split on every sentence for streaming (no combining)

Browse source
This commit is contained in:
russell@unturf.com 2026-01-26 19:52:41 -05:00
parent 0a0d023517
commit 8f7f1318a1
1 changed files with 4 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

23
speech.py
View file

@ -519,16 +519,13 @@ def preprocess(raw_input):
def simple_sentence_split(text: str, max_length: int = 500) -> list[str]: def simple_sentence_split(text: str, max_length: int = 500) -> list[str]:
"""Split text into sentences for streaming TTS. """Split text into sentences for streaming TTS.
Splits on sentence boundaries (.!?) for immediate streaming. Splits on every sentence boundary (.!?) for immediate streaming.
Only combines very short sentences (< 50 chars) with the next one.
Long sentences exceeding max_length are split at word boundaries. Long sentences exceeding max_length are split at word boundaries.
""" """
# Split on sentence boundaries, keeping the punctuation # Split on sentence boundaries, keeping the punctuation
parts = re.split(r'([.!?]+\s*)', text) parts = re.split(r'([.!?]+\s*)', text)
result = [] result = []
current = ""
for i in range(0, len(parts), 2): for i in range(0, len(parts), 2):
sentence = parts[i].strip() sentence = parts[i].strip()
punct = parts[i+1] if i+1 < len(parts) else "" punct = parts[i+1] if i+1 < len(parts) else ""
@ -536,21 +533,9 @@ def simple_sentence_split(text: str, max_length: int = 500) -> list[str]:
if not sentence: if not sentence:
continue continue
full_sentence = sentence + punct full_sentence = (sentence + punct).strip()
if full_sentence:
# If sentence is very short, combine with current buffer result.append(full_sentence)
if len(full_sentence) < 50 and current:
current += " " + full_sentence
elif current:
# Save buffered content and start fresh
result.append(current.strip())
current = full_sentence
else:
current = full_sentence
# Add remaining text
if current.strip():
result.append(current.strip())
# Split any sentences that exceed max_length at word boundaries # Split any sentences that exceed max_length at word boundaries
final_result = [] final_result = []