wolfang/anki-hsk-creator
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

test auding pauses toaudio genberation

Browse Source
This commit is contained in:
Wolfang Torres
2026-06-24 20:05:51 +08:00
parent bff05fca01
commit 5334f3aefc
2 changed files with 10 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/anki_hsk_creator/constants.py
View File

@@ -22,6 +22,10 @@ PHRASES_TYPE = ".phrases"
DICT_TYPE = ".dictionary" DICT_TYPE = ".dictionary"
DICTATION_TYPE = ".dictation" DICTATION_TYPE = ".dictation"
# time dalays
WORD = 0.1
COMMA = 0.5
POINT = 1.0
class LANGUAGES: class LANGUAGES:
"""Available laguages for translation""" """Available laguages for translation"""

10
src/anki_hsk_creator/proccessor.py
View File

@@ -8,7 +8,7 @@ import argostranslate.translate
import torchaudio import torchaudio
# Local # Local
from .constants import LANGUAGES from .constants import LANGUAGES, WORD, COMMA
from .utility import CCCEDICT, TTS, DictionaryResult, ProcessFile, TranslationResult from .utility import CCCEDICT, TTS, DictionaryResult, ProcessFile, TranslationResult
# Constants # Constants
@@ -27,7 +27,8 @@ def dictation_process(
for n, line in enumerate(text_lines): for n, line in enumerate(text_lines):
line = line.strip() line = line.strip()
line = " ".join(line.split()) line = " ".join(line.split())
line = line.replace("，", "， 。。。 ") audio_line = f"[pause:{WORD}s]".join(line)
audio_line = audio_line.replace("，", f"，[pause:{COMMA}s]")
audio_path = process_file.resources / f"N{n:03n}.wav" audio_path = process_file.resources / f"N{n:03n}.wav"
if not audio_path.exists(): if not audio_path.exists():
audio = TTS.MODEL.generate( audio = TTS.MODEL.generate(
@@ -51,11 +52,12 @@ def translator_process(
for n, line in enumerate(text_lines): for n, line in enumerate(text_lines):
line = line.strip() line = line.strip()
line = " ".join(line.split()) line = " ".join(line.split())
line = line.replace("，", "， 。。。 ") audio_line = "[pause:0.1s]".join(line)
audio_line = audio_line.replace("，", "，[pause:0.5s]")
audio_path = process_file.resources / f"N{n:03n}.wav" audio_path = process_file.resources / f"N{n:03n}.wav"
if not audio_path.exists(): if not audio_path.exists():
audio = TTS.MODEL.generate( audio = TTS.MODEL.generate(
f"{line}。", language_id=LANGUAGES.CN, **TTS.DEFAULTS f"{audio_line}。", language_id=LANGUAGES.CN, **TTS.DEFAULTS
) )
torchaudio.save(audio_path, audio, TTS.MODEL.sr) torchaudio.save(audio_path, audio, TTS.MODEL.sr)
translated = argostranslate.translate.translate( translated = argostranslate.translate.translate(