sepehr
7398cae359
All checks were successful
Deploy to Production / Build and Deploy (push) Successful in 2m25s
98 lines
4.1 KiB
Python
98 lines
4.1 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
Migrate and translate all non-multilingual glossaries in the database to multilingual glossaries.
|
|
It uses the free Google Translate provider to fill in the translations for target languages
|
|
(de, es, it, pt, nl, ru, ja, ko, zh, ar, fa, en) in the glossary_terms table.
|
|
"""
|
|
|
|
import sys
|
|
import logging
|
|
from pathlib import Path
|
|
|
|
ROOT = Path(__file__).resolve().parent.parent
|
|
sys.path.insert(0, str(ROOT))
|
|
|
|
from database.connection import get_sync_session
|
|
from database.models import Glossary, GlossaryTerm
|
|
from services.providers.google_provider import get_legacy_google_adapter
|
|
|
|
logging.basicConfig(
|
|
level=logging.INFO,
|
|
format="%(asctime)s [%(levelname)s] %(message)s",
|
|
)
|
|
logger = logging.getLogger("translate_non_multilingual_glossaries")
|
|
|
|
TARGET_LANGUAGES = ["de", "es", "it", "pt", "nl", "ru", "ja", "ko", "zh", "ar", "fa"]
|
|
|
|
def translate_term(adapter, text_to_translate: str, src_lang: str, target_langs: list[str]) -> dict:
|
|
translations = {}
|
|
for lang in target_langs:
|
|
try:
|
|
translated = adapter.translate(text_to_translate, target_language=lang, source_language=src_lang)
|
|
translations[lang] = translated.strip()
|
|
except Exception as e:
|
|
logger.error(f"Failed to translate '{text_to_translate}' to {lang}: {e}")
|
|
translations[lang] = ""
|
|
return translations
|
|
|
|
def main():
|
|
logger.info("Connecting to database and fetching non-multilingual glossaries...")
|
|
|
|
adapter = get_legacy_google_adapter()
|
|
|
|
with get_sync_session() as session:
|
|
# Find glossaries where target_language != 'multi'
|
|
non_multi_glossaries = session.query(Glossary).filter(Glossary.target_language != 'multi').all()
|
|
|
|
if not non_multi_glossaries:
|
|
logger.info("All glossaries in database are already multilingual ('multi').")
|
|
return 0
|
|
|
|
logger.info(f"Found {len(non_multi_glossaries)} non-multilingual glossaries to convert.")
|
|
|
|
for glossary in non_multi_glossaries:
|
|
logger.info(f"Processing glossary '{glossary.name}' (ID: {glossary.id})")
|
|
|
|
# Change target_language to 'multi'
|
|
glossary.target_language = 'multi'
|
|
|
|
# Rename the glossary to indicate it's now multilingual (e.g. replace "Anglais" with "Multilingue")
|
|
if "Anglais" in glossary.name:
|
|
glossary.name = glossary.name.replace("Anglais", "Multilingue")
|
|
elif "English" in glossary.name:
|
|
glossary.name = glossary.name.replace("English", "Multilingue")
|
|
else:
|
|
glossary.name = f"{glossary.name} → Multilingue"
|
|
|
|
src_lang = glossary.source_language or "fr"
|
|
|
|
terms = session.query(GlossaryTerm).filter(GlossaryTerm.glossary_id == glossary.id).all()
|
|
logger.info(f"Translating {len(terms)} terms for '{glossary.name}'...")
|
|
|
|
for idx, term in enumerate(terms):
|
|
translations = term.translations or {}
|
|
|
|
# Check if we already have translations for the target languages
|
|
missing_langs = [lang for lang in TARGET_LANGUAGES if lang not in translations or not translations[lang]]
|
|
|
|
if missing_langs:
|
|
# Translate
|
|
logger.info(f" [{idx+1}/{len(terms)}] Translating '{term.source}' to {missing_langs}...")
|
|
new_translations = translate_term(adapter, term.source, src_lang, missing_langs)
|
|
translations.update(new_translations)
|
|
|
|
# Ensure the original default target (e.g. English translation) is in the translations dict under 'en'
|
|
if 'en' not in translations or not translations['en']:
|
|
translations['en'] = term.target
|
|
|
|
term.translations = translations
|
|
|
|
session.commit()
|
|
logger.info(f"Glossary '{glossary.name}' successfully converted to multilingual.")
|
|
|
|
logger.info("Migration complete!")
|
|
return 0
|
|
|
|
if __name__ == '__main__':
|
|
sys.exit(main())
|