Dymas/abogen
1
0
Fork 0
mirror of https://github.com/denizsafak/abogen.git synced 2026-09-20 11:40:57 +02:00
Code Issues Packages Projects Releases Wiki Activity

Fix spaCy unknown language error: map Kokoro single-letter codes to Language enum in get_spacy_model

Browse Source
This commit is contained in:
Deniz Şafak
2026-07-23 03:33:06 +03:00
parent 27f88b759d
commit 342ea0dfac
1 changed files with 21 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
abogen/spacy_utils.py
+21 -5
View File
@@ -21,6 +21,19 @@ SPACY_MODELS = {
Language.HI: "xx_sent_ud_sm", Language.HI: "xx_sent_ud_sm",
} }
# Kokoro single-letter codes -> Language enum (inverse of pipeline_factory._KOKORO_LANG_MAP)
_KOKORO_TO_LANGUAGE = {
"a": Language.EN_US,
"b": Language.EN_GB,
"e": Language.ES,
"f": Language.FR,
"h": Language.HI,
"i": Language.IT,
"j": Language.JA,
"p": Language.PT_BR,
"z": Language.ZH,
}
def _load_spacy(): def _load_spacy():
"""Lazy load spaCy module.""" """Lazy load spaCy module."""
@@ -61,11 +74,14 @@ def get_spacy_model(lang_code, log_callback=None):
# Normalize to Language enum # Normalize to Language enum
if not isinstance(lang_code, Language): if not isinstance(lang_code, Language):
try: if isinstance(lang_code, str) and lang_code in _KOKORO_TO_LANGUAGE:
lang_code = Language.from_str(lang_code) lang_code = _KOKORO_TO_LANGUAGE[lang_code]
except ValueError: else:
log(f"\nspaCy: Unknown language '{lang_code}'...") try:
return None lang_code = Language.from_str(lang_code)
except ValueError:
log(f"\nspaCy: Unknown language '{lang_code}'...")
return None
# Check if model is cached # Check if model is cached
if lang_code in _nlp_cache: if lang_code in _nlp_cache: