Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix(espeak_wrapper): avoid stuck process on windows 

Fixes #24
This commit is contained in:
Enno Hermann 2024-05-28 21:43:35 +02:00
parent 203f60f1e1
commit 49fcbd908b
1 changed files with 13 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
TTS/tts/utils/text/phonemizers/espeak_wrapper.py
View File

@ -50,7 +50,7 @@ else:
_DEF_ESPEAK_VER = None _DEF_ESPEAK_VER = None
def _espeak_exe(espeak_lib: str, args: list) -> list[bytes]: def _espeak_exe(espeak_lib: str, args: list) -> list[str]:
"""Run espeak with the given arguments.""" """Run espeak with the given arguments."""
cmd = [ cmd = [
espeak_lib, espeak_lib,
@ -59,25 +59,18 @@ def _espeak_exe(espeak_lib: str, args: list) -> list[bytes]:
"1", # UTF8 text encoding "1", # UTF8 text encoding
] ]
cmd.extend(args) cmd.extend(args)
logger.debug("espeakng: executing %s", repr(cmd)) logger.debug("Executing: %s", repr(cmd))
with subprocess.Popen( p = subprocess.run(cmd, capture_output=True, encoding="utf8", check=True)
cmd, for line in p.stderr.strip().split("\n"):
stdout=subprocess.PIPE, if line.strip() != "":
stderr=subprocess.PIPE, logger.warning("%s: %s", espeak_lib, line.strip())
) as p: res = []
res = iter(p.stdout.readline, b"") for line in p.stdout.strip().split("\n"):
err = iter(p.stderr.readline, b"") if line.strip() != "":
for line in err: logger.debug("%s: %s", espeak_lib, line.strip())
logger.warning("espeakng: %s", line.decode("utf-8").strip()) res.append(line.strip())
res2 = list(res) return res
p.stdout.close()
if p.stderr:
p.stderr.close()
if p.stdin:
p.stdin.close()
p.wait()
return res2
class ESpeak(BasePhonemizer): class ESpeak(BasePhonemizer):
@ -195,8 +188,6 @@ class ESpeak(BasePhonemizer):
# compute phonemes # compute phonemes
phonemes = "" phonemes = ""
for line in _espeak_exe(self.backend, args): for line in _espeak_exe(self.backend, args):
logger.debug("line: %s", repr(line))
ph_decoded = line.decode("utf8").strip()
# espeak: # espeak:
# version 1.48.15: " p_ɹ_ˈaɪ_ɚ t_ə n_oʊ_v_ˈɛ_m_b_ɚ t_w_ˈɛ_n_t_i t_ˈuː\n" # version 1.48.15: " p_ɹ_ˈaɪ_ɚ t_ə n_oʊ_v_ˈɛ_m_b_ɚ t_w_ˈɛ_n_t_i t_ˈuː\n"
# espeak-ng: # espeak-ng:
@ -206,7 +197,7 @@ class ESpeak(BasePhonemizer):
# "sɛʁtˈɛ̃ mˈo kɔm (en)fˈʊtbɔːl(fr) ʒenˈɛʁ de- flˈaɡ də- lˈɑ̃ɡ." # "sɛʁtˈɛ̃ mˈo kɔm (en)fˈʊtbɔːl(fr) ʒenˈɛʁ de- flˈaɡ də- lˈɑ̃ɡ."
# phonemize needs to remove the language flags of the returned text: # phonemize needs to remove the language flags of the returned text:
# "sɛʁtˈɛ̃ mˈo kɔm fˈʊtbɔːl ʒenˈɛʁ de- flˈaɡ də- lˈɑ̃ɡ." # "sɛʁtˈɛ̃ mˈo kɔm fˈʊtbɔːl ʒenˈɛʁ de- flˈaɡ də- lˈɑ̃ɡ."
ph_decoded = re.sub(r"\(.+?\)", "", ph_decoded) ph_decoded = re.sub(r"\(.+?\)", "", line)
phonemes += ph_decoded.strip() phonemes += ph_decoded.strip()
return phonemes.replace("_", separator) return phonemes.replace("_", separator)
@ -226,13 +217,11 @@ class ESpeak(BasePhonemizer):
args = ["--voices"] args = ["--voices"]
langs = {} langs = {}
for count, line in enumerate(_espeak_exe(_DEF_ESPEAK_LIB, args)): for count, line in enumerate(_espeak_exe(_DEF_ESPEAK_LIB, args)):
line = line.decode("utf8").strip()
if count > 0: if count > 0:
cols = line.split() cols = line.split()
lang_code = cols[1] lang_code = cols[1]
lang_name = cols[3] lang_name = cols[3]
langs[lang_code] = lang_name langs[lang_code] = lang_name
logger.debug("line: %s", repr(line))
return langs return langs
def version(self) -> str: def version(self) -> str: