From b62dbbf153c5703ed76cbe2c6fa14b5b290ecad3 Mon Sep 17 00:00:00 2001 From: jxxghp Date: Tue, 29 Sep 2026 22:46:51 +0800 Subject: [PATCH] feat(music): use original torrent evidence in album recognition --- app/chain/media/album.py | 59 +++++- app/chain/media/contract.py | 2 + app/chain/transfer/filter.py | 41 ++-- app/chain/transfer/music.py | 51 +++++ docs/architecture-overview.md | 4 +- docs/refactor/optimization-checklist.md | 2 +- .../architecture/dependency-baseline.json | 5 +- tests/test_music_resource_context.py | 197 ++++++++++++++++++ tests/test_transfer_music_classification.py | 6 +- 9 files changed, 337 insertions(+), 30 deletions(-) create mode 100644 tests/test_music_resource_context.py diff --git a/app/chain/media/album.py b/app/chain/media/album.py index 2e509f442..7dc002683 100644 --- a/app/chain/media/album.py +++ b/app/chain/media/album.py @@ -1,5 +1,6 @@ """专辑目录扫描、曲目对齐与缓存编排 owner。""" +import json import os from copy import deepcopy from pathlib import Path @@ -28,9 +29,47 @@ def _album_directory_cache_key( directory: Path, regions: tuple[str, ...], scripts: tuple[str, ...], + contextual_meta: Optional[MetaMusic] = None, ) -> str: - """返回包含发行偏好的目录键，避免不同手动选择错误共享结果。""" - return "|".join((os.path.abspath(directory), ",".join(regions), ",".join(scripts))) + """将发行偏好及资源证据纳入缓存键，同目录更换种子线索时重新识别。""" + evidence = { + key: getattr(contextual_meta, key, None) + for key in ("album", "artists", "album_artist", "year", "version") + } if contextual_meta else None + if evidence and evidence["album_artist"]: + evidence["artists"] = [evidence["album_artist"]] + if evidence and not any(evidence.values()): + evidence = None + return json.dumps([os.path.abspath(directory), regions, scripts, evidence], ensure_ascii=False, sort_keys=True) + + +def _album_context_with_resource( + directory: Path, + metas: list[MetaMusic], + contextual_meta: Optional[MetaMusic], +) -> MetaMusic: + """以文件自身专辑线索为先，用绑定到本次文件的资源证据补足目录查询。""" + album_meta = MetaMusic.from_album_context(directory.name, metas) + if not contextual_meta: + return album_meta + if contextual_meta.album and not any(meta.album for meta in metas): + album_meta.album = contextual_meta.album + album_meta.title = contextual_meta.album + if not album_meta.artists: + album_meta.artists = ( + [contextual_meta.album_artist] if contextual_meta.album_artist + else list(contextual_meta.artists) + ) + if not album_meta.album_artist and contextual_meta.album_artist: + album_meta.album_artist = contextual_meta.album_artist + for key in ("year", "version"): + local_values = {getattr(meta, key) for meta in metas if getattr(meta, key)} + if not getattr(album_meta, key): + if len(local_values) == 1: + setattr(album_meta, key, next(iter(local_values))) + elif not local_values: + setattr(album_meta, key, getattr(contextual_meta, key)) + return album_meta class MediaAlbumOwner(_MediaOwnerBase): @@ -203,10 +242,11 @@ class MediaAlbumOwner(_MediaOwnerBase): files: list[Path], music_release_regions: Optional[tuple[str, ...]] = None, music_release_scripts: Optional[tuple[str, ...]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """同步汇总本地专辑证据并委托 MusicBrainz 来源链匹配。""" metas = AudioMetadataHelper.read_many(files) - album_meta = MetaMusic.from_album_context(directory.name, metas) + album_meta = _album_context_with_resource(directory, metas, contextual_meta) regions, scripts = self._music_release_preferences( list(music_release_regions) if music_release_regions is not None else None, list(music_release_scripts) if music_release_scripts is not None else None, @@ -227,10 +267,11 @@ class MediaAlbumOwner(_MediaOwnerBase): files: list[Path], music_release_regions: Optional[tuple[str, ...]] = None, music_release_scripts: Optional[tuple[str, ...]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """异步汇总本地专辑证据并委托 MusicBrainz 来源链匹配。""" metas = await run_in_threadpool(AudioMetadataHelper.read_many, files) - album_meta = MetaMusic.from_album_context(directory.name, metas) + album_meta = _album_context_with_resource(directory, metas, contextual_meta) regions, scripts = self._music_release_preferences( list(music_release_regions) if music_release_regions is not None else None, list(music_release_scripts) if music_release_scripts is not None else None, @@ -250,6 +291,7 @@ class MediaAlbumOwner(_MediaOwnerBase): path: Union[str, Path], music_release_regions: Optional[list[str]] = None, music_release_scripts: Optional[list[str]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """按目录级线索批量识别整张专辑并返回文件到曲目的映射。""" directory = Path(path) @@ -263,12 +305,14 @@ class MediaAlbumOwner(_MediaOwnerBase): music_release_scripts, ) custom_preference = music_release_regions is not None or music_release_scripts is not None - key = _album_directory_cache_key(directory, regions, scripts) + key = _album_directory_cache_key(directory, regions, scripts, contextual_meta) signature = self._album_directory_signature(directory, files) matched = self._album_dir_cache.resolve( key, signature, lambda: ( + self._match_music_album_directory(directory, files, regions, scripts, contextual_meta) + if contextual_meta is not None else self._match_music_album_directory(directory, files, regions, scripts) if custom_preference else self._match_music_album_directory(directory, files) @@ -288,6 +332,7 @@ class MediaAlbumOwner(_MediaOwnerBase): path: Union[str, Path], music_release_regions: Optional[list[str]] = None, music_release_scripts: Optional[list[str]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """异步按目录级线索批量识别整张专辑。""" directory = Path(path) @@ -301,12 +346,14 @@ class MediaAlbumOwner(_MediaOwnerBase): music_release_scripts, ) custom_preference = music_release_regions is not None or music_release_scripts is not None - key = _album_directory_cache_key(directory, regions, scripts) + key = _album_directory_cache_key(directory, regions, scripts, contextual_meta) signature = self._album_directory_signature(directory, files) matched = await self._album_dir_cache.async_resolve( key, signature, lambda: ( + self._async_match_music_album_directory(directory, files, regions, scripts, contextual_meta) + if contextual_meta is not None else self._async_match_music_album_directory(directory, files, regions, scripts) if custom_preference else self._async_match_music_album_directory(directory, files) diff --git a/app/chain/media/contract.py b/app/chain/media/contract.py index f14afbc87..1d45d988f 100644 --- a/app/chain/media/contract.py +++ b/app/chain/media/contract.py @@ -171,6 +171,7 @@ if TYPE_CHECKING: path: Union[str, Path], music_release_regions: Optional[list[str]] = None, music_release_scripts: Optional[list[str]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """同步识别音乐专辑目录。""" ... @@ -190,6 +191,7 @@ if TYPE_CHECKING: path: Union[str, Path], music_release_regions: Optional[list[str]] = None, music_release_scripts: Optional[list[str]] = None, + contextual_meta: Optional[MetaMusic] = None, ) -> dict[str, MusicInfo]: """异步识别音乐专辑目录。""" ... diff --git a/app/chain/transfer/filter.py b/app/chain/transfer/filter.py index dcb3f0de7..161549897 100644 --- a/app/chain/transfer/filter.py +++ b/app/chain/transfer/filter.py @@ -19,7 +19,11 @@ from app.chain._contracts import TransferMixinHost from app.chain.media import MediaChain from app.chain.storage import StorageChain from app.chain.transfer.contract import _TransferOwnerBase -from app.chain.transfer.music import prepare_music_batch_context, resolve_music_batch_file_context +from app.chain.transfer.music import ( + prepare_music_batch_context, + resolve_music_batch_file_context, + restore_music_resource_meta, +) from app.domain.context import MediaInfo, MusicAlbumInfo, MusicInfo from app.domain.media import normalize_music_type from app.domain.meta.metamusic import MetaMusic @@ -186,14 +190,12 @@ class _MusicFileFilterBase(_TransferOwnerBase): if file_meta.media_id or getattr(file_item, "storage", "local") != "local": return file_meta, None try: - if music_release_regions is None and music_release_scripts is None: - matched = MediaChain().recognize_music_album_directory(file_path.parent) - else: - matched = MediaChain().recognize_music_album_directory( - file_path.parent, - music_release_regions=music_release_regions, - music_release_scripts=music_release_scripts, - ) + matched = MediaChain().recognize_music_album_directory( + file_path.parent, + music_release_regions=music_release_regions, + music_release_scripts=music_release_scripts, + contextual_meta=file_meta, + ) except Exception as err: logger.debug(f"音乐专辑目录匹配失败：{file_path} - {err}") return file_meta, None @@ -410,20 +412,28 @@ class FileFilterMixin(_TransferOwnerBase): 种子未提供的语义字段由历史中已选媒体补缺；实体类型和来源身份始终 沿用已选媒体，不根据专辑名或文件曲名在单曲与专辑之间转换。 - 多音轨批次误带单曲身份时只保留文件自身标签，避免把同一 recording - 身份传播到整张专辑；调用方随后可使用目录级证据重新匹配专辑。 + 多音轨批次误带单曲身份时保留文件标签和独立的原始种子线索，避免 + 把同一 recording 身份传播到整张专辑；随后可按目录重新匹配专辑。 手动选中多条历史且未要求复用历史身份时，专辑身份也必须丢弃，确保 目录级识别能够重新补齐发行版、分类和规范名称。 """ note = getattr(download_history, "note", None) music_note = note.get("music") if isinstance(note, dict) else None if not isinstance(music_note, dict) or music_note.get("version") != 1: - return None, None + if not download_history or not MediaChain.is_audio_path(file_path): + return None, None + return restore_music_resource_meta(download_history, file_path), None try: saved_meta = MetaMusic.from_dict(music_note.get("meta") or {}) saved_info = MusicInfo.from_dict(music_note.get("media") or {}) except (TypeError, ValueError): - return None, None + resource_meta = ( + restore_music_resource_meta(download_history, file_path) + if MediaChain.is_audio_path(file_path) else None + ) + return resource_meta, None + if not (saved_info.title or saved_meta.title or saved_meta.album): + return restore_music_resource_meta(download_history, file_path), None file_tags = MediaChain.read_path_meta(file_path) should_discard_identity = discard_saved_identity or ( @@ -431,12 +441,11 @@ class FileFilterMixin(_TransferOwnerBase): and saved_info.music_type == MUSIC_ENTITY_RECORDING ) if should_discard_identity: - # 共享 recording 上下文可能包含错误的专辑、年份等字段；整张丢弃， - # 只保留当前文件实际标签，目录级匹配失败时也不会回落到错误身份。 + # 丢弃选错媒体产生的快照，原始资源标题仍可独立作为无身份的查询线索。 file_meta = deepcopy(file_tags) file_meta.org_string = file_path.name file_meta.title = file_meta.title or file_path.stem - return file_meta, None + return restore_music_resource_meta(download_history, file_path, file_meta), None file_meta = deepcopy(saved_meta) # 新旧历史都可能仅在 media 中保留已选专辑；先补缺，再沿用文件标签的 diff --git a/app/chain/transfer/music.py b/app/chain/transfer/music.py index 64efb1778..d4c6383a2 100644 --- a/app/chain/transfer/music.py +++ b/app/chain/transfer/music.py @@ -8,6 +8,7 @@ from pathlib import Path from typing import Any, Optional, Tuple, Union, cast from app.application.audio import AudioMetadataHelper +from app.application.history import DownloadHistorySnapshot from app.chain.transfer.contract import _TransferOwnerBase from app.domain.context import MediaInfo, MusicInfo from app.domain.meta.metamusic import MetaMusic @@ -29,6 +30,56 @@ class MusicBatchContext: resolved_contexts: dict[Tuple[str, str], tuple[MetaMusic, MusicInfo]] = field(default_factory=dict) +def restore_music_resource_meta( + history: Optional[DownloadHistorySnapshot], + path: Path, + file_meta: Optional[MetaMusic] = None, +) -> Optional[MetaMusic]: + """从原始种子主副标题补足文件的发行线索，不复用已丢弃的远端身份。 + + 专辑字段只传到下载根及其碟片目录；全集或更深的独立专辑目录只继承 + 可信艺人。文件曲名、曲序和已有标签不被整包标题覆盖。 + """ + title = str(getattr(history, "torrent_name", None) or "") + description = str(getattr(history, "torrent_description", None) or "") + history_type = getattr(history, "type", None) + if history_type in (MediaType.MOVIE, MediaType.TV, MediaType.MOVIE.value, MediaType.TV.value): + return file_meta + if not title and not description: + return file_meta + root_text = getattr(history, "path", None) + root = Path(root_text) if root_text else None + if root and root != path and not path.is_relative_to(root): + return file_meta + resource = MetaMusic.parse_resource(title, description) + meta = deepcopy(file_meta) if file_meta else ( + AudioMetadataHelper.read(path) if path.is_file() else AudioMetadataHelper.read_filename(path) + ) + parent = path.parent.parent if MetaMusic.parse_disc_dir(path.parent.name) else path.parent + direct_release = bool(root and (root == path or root == parent)) + collection = bool(re.search( + r"全集|专辑合集|專輯合集|\bdiscography\b|\bcomplete\s+(?:albums?|collection)\b|\d+\s*张专辑", + f"{title} {description}", re.IGNORECASE, + )) + collective = {music_text_key(value) for value in ("Various Artists", "Various", "VA", "群星", "众艺人", "眾藝人")} + artists = [artist for artist in resource.artists if music_text_key(artist) not in collective] + if not meta.artists and artists: + meta.artists = list(artists) + if not meta.album_artist and len(resource.artists) == 1: + meta.album_artist = resource.artists[0] + if direct_release and not collection: + album = resource.album + if root != path and not resource.track_number: + album = album or resource.title + if not meta.album and album: + meta.album = album + if not meta.year and resource.year: + meta.year = resource.year + if not meta.version and resource.version: + meta.version = resource.version + return meta + + def _music_directory_consensus(values: list[str]) -> Optional[str]: """只在同目录至少两个标签高度一致时返回共识值。""" normalized = [ diff --git a/docs/architecture-overview.md b/docs/architecture-overview.md index 9f8a544ab..c64903b56 100644 --- a/docs/architecture-overview.md +++ b/docs/architecture-overview.md @@ -763,7 +763,7 @@ flowchart LR | 指标 | 当前值 | |---|---:| | Python 模块 | 1054 | -| 内部导入边 | 8,955 | +| 内部导入边 | 8,956 | | 非平凡 SCC | 1（精确 containment 的 TMDB 移植包环） | | Application / Chain 具体 Adapter 直连 | 0 / 0 | | Direct egress | 53（债务已清零，53 条精确 containment） | @@ -773,7 +773,7 @@ flowchart LR | Model/Oper 自动事务与自建 Session | 0 | | 组合根外 `SystemConfigOper()` | 0 | -整理失败反馈由 `app.application.transfer.feedback` 集中投影；已整理下载的站点字幕复用 `app.chain.transfer` 自动整理队列；网络连通性检测按 `app.application.nettest` 归档，并覆盖 HTTP 与 WebSocket 探测。未启用智能助手的消息音乐交互由 `app.chain.music_interaction` 包中的解析展示与订阅协调模块处理；音乐标签准入与批次共识由 `app.chain.transfer.music` 负责；宿主依赖边现为 8,955，音乐交互包未新增 SCC，插件一次性任务 owner `app.scheduler.oncejob` 带来 1 个模块和 9 条依赖边。Agent 持久回执新增 Application 端口及 DB Model/Oper/Adapter 四个冷导入模块。缓存组合根只在启动时选用 Redis 缓存才导入 Redis 适配器，依赖 langchain_core 的对话记忆模型归属 `app.agent.memory`，文件缓存且未启用智能体的实例不再加载 redis 与 langchain_core。当前 `app.startup.lifecycle` 为 574、`app.factory` 为 586、`app.main` 为 588。站点图片域名快照新增 `app.application.security.image`，各启动入口增加一个模块；性能基线只同步模块数量，原有耗时预算、历史采样和生命周期资源约束保持有效。 +整理失败反馈由 `app.application.transfer.feedback` 集中投影；已整理下载的站点字幕复用 `app.chain.transfer` 自动整理队列；网络连通性检测按 `app.application.nettest` 归档，并覆盖 HTTP 与 WebSocket 探测。未启用智能助手的消息音乐交互由 `app.chain.music_interaction` 包中的解析展示与订阅协调模块处理；音乐标签准入与批次共识由 `app.chain.transfer.music` 负责；宿主依赖边现为 8,956，音乐交互包未新增 SCC，插件一次性任务 owner `app.scheduler.oncejob` 带来 1 个模块和 9 条依赖边。Agent 持久回执新增 Application 端口及 DB Model/Oper/Adapter 四个冷导入模块。缓存组合根只在启动时选用 Redis 缓存才导入 Redis 适配器，依赖 langchain_core 的对话记忆模型归属 `app.agent.memory`，文件缓存且未启用智能体的实例不再加载 redis 与 langchain_core。当前 `app.startup.lifecycle` 为 574、`app.factory` 为 586、`app.main` 为 588。站点图片域名快照新增 `app.application.security.image`，各启动入口增加一个模块；性能基线只同步模块数量，原有耗时预算、历史采样和生命周期资源约束保持有效。 架构专项验证分为两个 CI 投影：`Check event semantic policy` 先运行依赖、Adapter、出口和 Event 语义门禁，`Check host architecture snapshot` 再执行快照测试及一次 diff --git a/docs/refactor/optimization-checklist.md b/docs/refactor/optimization-checklist.md index fb22fa9d9..4e5987403 100644 --- a/docs/refactor/optimization-checklist.md +++ b/docs/refactor/optimization-checklist.md @@ -94,7 +94,7 @@ ARCH-201 至 ARCH-204 均达到实现、验证、提交、推送和远端门禁 | 指标 | 当前值 | 解释 | |---|---:|---| -| 宿主 Python 模块 / 内部依赖边 | 1054 / 8,955 | `dependency-baseline.json` 当前快照；系统设置合同按 `app.application.settings` 包组织，系统设置 API 端点独立按 `app.api.endpoints.settings` 归档；消息音乐交互由 `app.chain.music_interaction` 包中的音乐解析/展示与订阅协调模块处理；已整理下载的站点字幕经 `app.chain.transfer` 自动整理队列完成命名与归档；网络连通性检测按 `app.application.nettest` 归档 | +| 宿主 Python 模块 / 内部依赖边 | 1054 / 8,956 | `dependency-baseline.json` 当前快照；系统设置合同按 `app.application.settings` 包组织，系统设置 API 端点独立按 `app.api.endpoints.settings` 归档；消息音乐交互由 `app.chain.music_interaction` 包中的音乐解析/展示与订阅协调模块处理；已整理下载的站点字幕经 `app.chain.transfer` 自动整理队列完成命名与归档；网络连通性检测按 `app.application.nettest` 归档 | | 非平凡 SCC | 1 | 仅保留精确 containment 的 29 模块 TMDB 移植包环 | | 跨层 DB 边界债务 | 0 | Application、Chain、API、Agent、Runtime、Workflow 到 DB 的受控债务均为零 | | Model/Oper 事务债务 | 0 | 自建 Session、自动事务装饰器、直接 commit/rollback 等基线均为零 | diff --git a/tests/fixtures/architecture/dependency-baseline.json b/tests/fixtures/architecture/dependency-baseline.json index 12b058104..a801848eb 100644 --- a/tests/fixtures/architecture/dependency-baseline.json +++ b/tests/fixtures/architecture/dependency-baseline.json @@ -1097,8 +1097,8 @@ "runtime_only": true } }, - "edge_count": 8955, - "edge_sha256": "b56dbef23e86e017c08b09b354d227a0c286cae52c224f797c254efa956de368", + "edge_count": 8956, + "edge_sha256": "b6547bb8bb60c5adc0dae7afaf2dddb11a782c3b9644b550e28945bfadbe2daf", "edges": [ "app -> app.foundation", "app -> app.foundation.environment", @@ -5582,6 +5582,7 @@ "app.chain.transfer.history -> app.schemas.workflow", "app.chain.transfer.music -> app.application", "app.chain.transfer.music -> app.application.audio", + "app.chain.transfer.music -> app.application.history", "app.chain.transfer.music -> app.chain", "app.chain.transfer.music -> app.chain.transfer", "app.chain.transfer.music -> app.chain.transfer.contract", diff --git a/tests/test_music_resource_context.py b/tests/test_music_resource_context.py new file mode 100644 index 000000000..54563daf4 --- /dev/null +++ b/tests/test_music_resource_context.py @@ -0,0 +1,197 @@ +"""下载种子原始证据在丢弃旧身份后仍参与整专查询，并保持作用范围。""" + +import asyncio +import shutil +from pathlib import Path +from unittest.mock import AsyncMock, Mock + +import pytest +from mutagen.flac import FLAC + +from app.application.history import DownloadHistorySnapshot +from app.chain.media import MediaChain +from app.chain.transfer.facade import TransferChain +from app.domain.context import MusicAlbumInfo, MusicInfo +from app.domain.meta.metamusic import MetaMusic +from app.schemas.types import MediaType +from tests.test_transfer_sync_extra_files import make_fileitem + + +def _history(root: Path, **kwargs) -> DownloadHistorySnapshot: + """构造只有站点原始名称、没有已选媒体身份的下载历史。""" + values = dict( + id=1, path=str(root), type=MediaType.MUSIC.value, title="Unrecognized", + torrent_name="Artist - Actual Album (2003) [FLAC]", + torrent_description="音乐专辑", note=None, + ) + values.update(kwargs) + return DownloadHistorySnapshot(**values) + + +def _audio_files(root: Path) -> list[Path]: + """生成无标签但具有真实时长和曲名的整专输入。""" + root.mkdir(parents=True) + files = [root / "01 - First Song.flac", root / "02 - Second Song.flac"] + for path in files: + shutil.copyfile(Path(__file__).parent / "fixtures/audio/silence.flac", path) + return files + + +@pytest.mark.parametrize("note", [ + None, {"music": {"version": 99}}, {"music": {"version": 1}}, + {"music": {"version": 1, "meta": {"type": "电影"}}}, +]) +def test_missing_or_invalid_music_snapshot_keeps_torrent_evidence(tmp_path, note): + """旧历史没有可用音乐快照时，原始主副标题仍可补艺人、专辑和年份。""" + root = tmp_path / "opaque" + paths = _audio_files(root) + history = _history(root, note=note) + + meta, info = TransferChain._restore_music_download_context(history, paths[0]) + + assert info is None + assert meta.title == "First Song" + assert meta.track_number == 1 + assert meta.duration == 3 + assert meta.artists == ["Artist"] + assert meta.album == "Actual Album" + assert meta.year == 2003 + assert meta.media_id is None + + +def test_discarded_wrong_identity_does_not_discard_original_resource(tmp_path): + """选错的媒体快照可以丢弃，原始种子不是错误身份的一部分。""" + root = tmp_path / "opaque" + paths = _audio_files(root) + wrong = MusicInfo(media_source="musicbrainz", media_id="wrong", title="Wrong Song", album="Wrong Album") + history = _history(root, note={"music": { + "version": 1, "meta": MetaMusic.from_music_info(wrong).to_dict(), "media": wrong.to_dict(), + }}) + + meta, info = TransferChain._restore_music_download_context(history, paths[0], discard_saved_identity=True) + + assert info is None + assert (meta.title, meta.album) == ("First Song", "Actual Album") + assert meta.artists == ["Artist"] + assert meta.media_id is None + + +@pytest.mark.parametrize("subdir,torrent", [ + ("2001 - Child Album", "Artist - Actual Album (2003) [FLAC]"), + ("", "Artist - Discography (2003) [FLAC]"), +]) +def test_package_title_does_not_override_independent_albums(tmp_path, subdir, torrent): + """全集和独立子专辑只接收艺人，不能被顶层下载标题改成同一个专辑。""" + root = tmp_path / "opaque" + paths = _audio_files(root / subdir if subdir else root) + + meta, _ = TransferChain._restore_music_download_context(_history(root, torrent_name=torrent), paths[0]) + + assert meta.artists == ["Artist"] + assert meta.album == ("Child Album" if subdir else None) + assert meta.year == (2001 if subdir else None) + + +def test_resource_album_reaches_disc_directory_and_uses_subtitle(tmp_path): + """CD 子目录继承同一发行信息，中文副标题补齐主标题没有的作品与署名。""" + root = tmp_path / "opaque" + paths = _audio_files(root / "CD2") + history = _history(root, torrent_name="FLAC", torrent_description="歌手：周杰伦；专辑：叶惠美；[2003]") + + meta, _ = TransferChain._restore_music_download_context(history, paths[0]) + + assert meta.artists == ["周杰伦"] + assert meta.album == "叶惠美" + assert meta.year == 2003 + assert meta.disc_number == 2 + assert meta.title == "First Song" + + +@pytest.mark.parametrize("outside,media_type", [(True, "音乐"), (False, "电影"), (False, "电视剧")]) +def test_unrelated_history_cannot_supply_music_identity(tmp_path, outside, media_type): + """原始资源必须属于当前音频，影视任务的附加音轨不能被改为音乐。""" + root = tmp_path / "opaque" + paths = _audio_files(root) + history = _history(tmp_path / "other" if outside else root, type=media_type) + + assert TransferChain._restore_music_download_context(history, paths[0]) == (None, None) + + +@pytest.fixture +def media_chain(monkeypatch): + """使用隔离目录缓存，避免同步、异步测试共享其它用例的识别结果。""" + from app.chain.media.cache import AlbumDirectoryCache + + chain = MediaChain() + monkeypatch.setattr(chain, "_album_dir_cache", AlbumDirectoryCache(16)) + return chain + + +@pytest.mark.parametrize("asynchronous", [False, True]) +def test_original_resource_is_used_by_album_query(tmp_path, monkeypatch, media_chain, asynchronous): + """资源艺人、专辑、年份进入真实目录查询，并保留各文件曲名和时长。""" + root = tmp_path / "opaque" + paths = _audio_files(root) + meta, _ = TransferChain._restore_music_download_context(_history(root), paths[0]) + album = MusicAlbumInfo(media_source="musicbrainz", media_id="group", title="Actual Album", tracks=[ + MusicInfo(media_source="musicbrainz", media_id=f"recording-{index}", title=title, + album="Actual Album", artists=["Artist"], year=2003, track_number=index, duration=3) + for index, title in enumerate(("First Song", "Second Song"), 1) + ]) + source = Mock(match_music_album=Mock(return_value=album), async_match_music_album=AsyncMock(return_value=album)) + monkeypatch.setattr("app.chain.media.album.MusicBrainzChain", Mock(return_value=source)) + + if asynchronous: + matched = asyncio.run(media_chain.async_recognize_music_album_directory(root, contextual_meta=meta)) + query = source.async_match_music_album.call_args + else: + matched_meta, info = TransferChain._match_music_album_context(make_fileitem(str(paths[0])), paths[0], meta) + assert info.media_id == "recording-1" + assert matched_meta.title == "First Song" + matched = media_chain.recognize_music_album_directory(root, contextual_meta=meta) + query = source.match_music_album.call_args + + assert query.args[0].title == query.args[0].album == "Actual Album" + assert query.args[0].artists == ["Artist"] + assert query.args[0].year == 2003 + assert [track.duration for track in query.args[1]] == [3, 3] + assert [info.title for info in matched.values()] == ["First Song", "Second Song"] + + +@pytest.mark.parametrize("asynchronous", [False, True]) +def test_album_cache_changes_with_resource_context(tmp_path, monkeypatch, media_chain, asynchronous): + """同一目录和文件签名也必须区分先后提供的不同专辑证据。""" + root = tmp_path / "opaque" + _audio_files(root) + source = Mock(match_music_album=Mock(return_value=None), async_match_music_album=AsyncMock(return_value=None)) + monkeypatch.setattr("app.chain.media.album.MusicBrainzChain", Mock(return_value=source)) + for album in ("First Album", "First Album", "Second Album"): + context = MetaMusic(album=album, artists=["Artist"]) + if asynchronous: + asyncio.run(media_chain.async_recognize_music_album_directory(root, contextual_meta=context)) + else: + media_chain.recognize_music_album_directory(root, contextual_meta=context) + operation = source.async_match_music_album if asynchronous else source.match_music_album + + assert [call.args[0].album for call in operation.call_args_list] == ["First Album", "Second Album"] + + +def test_local_album_tags_override_conflicting_resource(tmp_path, monkeypatch, media_chain): + """目录里已明确的专辑标签和年份不能被种子候选覆盖。""" + root = tmp_path / "opaque" + paths = _audio_files(root) + for path in paths: + audio = FLAC(path) + audio.update({"album": ["Tagged Album"], "artist": ["Tagged Artist"], "date": ["2001"]}) + audio.save() + source = Mock(match_music_album=Mock(return_value=None)) + monkeypatch.setattr("app.chain.media.album.MusicBrainzChain", Mock(return_value=source)) + + media_chain.recognize_music_album_directory(root, contextual_meta=MetaMusic( + album="Wrong Album", artists=["Wrong Artist"], year=2020, + )) + + query = source.match_music_album.call_args.args[0] + assert query.album == "Tagged Album" + assert query.artists == ["Tagged Artist"] + assert query.year == 2001 diff --git a/tests/test_transfer_music_classification.py b/tests/test_transfer_music_classification.py index 6b88e361d..5c651a8ea 100644 --- a/tests/test_transfer_music_classification.py +++ b/tests/test_transfer_music_classification.py @@ -83,7 +83,7 @@ def test_album_context_preserves_classification_and_remote_music_facts( monkeypatch.setattr( MediaChain, "recognize_music_album_directory", - lambda _self, _directory: {str(file_path.resolve()): matched_info}, + lambda _self, _directory, **_kwargs: {str(file_path.resolve()): matched_info}, ) merged_meta, merged_info = FileFilterMixin._match_music_album_context( @@ -160,7 +160,7 @@ def test_album_context_rejects_candidate_conflicting_with_file_tags( monkeypatch.setattr( MediaChain, "recognize_music_album_directory", - lambda _self, _directory: {str(file_path.resolve()): wrong_release}, + lambda _self, _directory, **_kwargs: {str(file_path.resolve()): wrong_release}, ) returned_meta, returned_info = FileFilterMixin._match_music_album_context( @@ -212,7 +212,7 @@ def test_album_context_preserves_local_release_track_position( monkeypatch.setattr( MediaChain, "recognize_music_album_directory", - lambda _self, _directory: {str(file_path.resolve()): matched_info}, + lambda _self, _directory, **_kwargs: {str(file_path.resolve()): matched_info}, ) merged_meta, merged_info = FileFilterMixin._match_music_album_context(