Files
Auto_Bangumi/backend/src/module/parser/title_parser.py
Estrella Pan fa24ec4e2a fix(core): stop blocking the event loop in OpenAI parsing and static routing
OpenAIParser now uses AsyncOpenAI — the previous sync SDK call ran directly
on the event loop (submitting to a ThreadPoolExecutor and immediately
blocking on .result() serialized it anyway). TitleParser.raw_parser and its
callers become async accordingly.

The SPA catch-all route listed dist/ on every request; snapshot it once at
startup. The module-level bangumi TTL cache is now lock-guarded — it is
written from asyncio.to_thread workers in notification paths while the event
loop reads it.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014w1Z6Nxy6XTRgkFXqPr9Zh
2026-07-02 11:39:10 +02:00

115 lines
4.1 KiB
Python

import logging
from module.conf import settings
from module.models import Bangumi
from module.models.bangumi import Episode
from module.parser.analyser import (
OpenAIParser,
mikan_parser,
raw_parser,
tmdb_parser,
torrent_parser,
)
logger = logging.getLogger(__name__)
class TitleParser:
def __init__(self):
pass
@staticmethod
def torrent_parser(
torrent_path: str,
torrent_name: str | None = None,
season: int | None = None,
file_type: str = "media",
):
try:
return torrent_parser(torrent_path, torrent_name, season, file_type)
except Exception as e:
logger.warning(f"Cannot parse {torrent_path} with error {e}")
@staticmethod
async def tmdb_parser(title: str, season: int, language: str):
tmdb_info = await tmdb_parser(title, language)
if tmdb_info:
logger.debug("TMDB Matched, official title is %s", tmdb_info.title)
tmdb_season = tmdb_info.last_season if tmdb_info.last_season else season
return tmdb_info.title, tmdb_season, tmdb_info.year, tmdb_info.poster_link
else:
logger.warning(f"Cannot match {title} in TMDB. Use raw title instead.")
logger.warning("Please change bangumi info manually.")
return title, season, None, None
@staticmethod
async def tmdb_poster_parser(bangumi: Bangumi):
tmdb_info = await tmdb_parser(
bangumi.official_title, settings.rss_parser.language
)
if tmdb_info:
logger.debug("TMDB Matched, official title is %s", tmdb_info.title)
bangumi.poster_link = tmdb_info.poster_link
else:
logger.warning(
f"Cannot match {bangumi.official_title} in TMDB. Use raw title instead."
)
logger.warning("Please change bangumi info manually.")
@staticmethod
async def raw_parser(raw: str) -> Bangumi | None:
language = settings.rss_parser.language
try:
# use OpenAI ChatGPT to parse raw title and get structured data
if settings.experimental_openai.enable:
kwargs = settings.experimental_openai.dict(exclude={"enable"})
gpt = OpenAIParser(**kwargs)
episode_dict = await gpt.parse(raw, asdict=True)
episode = Episode(**episode_dict)
else:
episode = raw_parser(raw)
if episode is None:
return None
titles = {
"zh": episode.title_zh,
"en": episode.title_en,
"jp": episode.title_jp,
}
title_raw = episode.title_en or episode.title_zh or episode.title_jp
if titles[language]:
official_title = titles[language]
elif titles["zh"]:
official_title = titles["zh"]
elif titles["en"]:
official_title = titles["en"]
elif titles["jp"]:
official_title = titles["jp"]
else:
official_title = title_raw
if not title_raw:
logger.warning("Cannot extract title_raw from '%s', skipping", raw)
return None
_season = episode.season
logger.debug("RAW:%s >> %s", raw, title_raw)
return Bangumi(
official_title=official_title,
title_raw=title_raw,
season=_season,
season_raw=episode.season_raw,
group_name=episode.group,
dpi=episode.resolution,
source=episode.source,
subtitle=episode.sub,
eps_collect=False if episode.episode > 1 else True,
offset=0,
filter=",".join(settings.rss_parser.filter),
)
except (ValueError, AttributeError, TypeError) as e:
logger.warning(f"Cannot parse '{raw}': {type(e).__name__}: {e}")
return None
@staticmethod
async def mikan_parser(homepage: str) -> tuple[str, str]:
return await mikan_parser(homepage)