Skip to content

armgit / peach

Public Imported from GitHub
0

Peach: Local-first video metadata scraper and media manager.

fix(catalog): 多个来源名的版次证据取并集,补齐接口夹具的下载任务列

longmeidao93833b47eb29
diff --git a/src/peach/catalog_rules.py b/src/peach/catalog_rules.py
index bdf7b0f47..46ba4a678 100644
--- a/src/peach/catalog_rules.py
+++ b/src/peach/catalog_rules.py
@@ -1139,13 +1139,14 @@ def jav_edition_badges(name: str | None, code: str | None,
- 中字只有「有」这一个方向,任何一边有就算;
- 无码与无码破解以文件侧(文件名、番号形状、标签)为准,文件侧什么都没说时才用来源名。
文件名写 `-U`、种子名写「破解」,结果是「无码」。
+ 几个来源名之间不分先后,各项证据取并集。
"""
subtitled, uncensored, cracked = _edition_evidence(name, code, tags)
- for source in source_names:
- source_subtitled, source_uncensored, source_cracked = _edition_evidence(source, code)
- subtitled = subtitled or source_subtitled
- if not uncensored:
- uncensored, cracked = source_uncensored, source_cracked
+ evidence = [_edition_evidence(source, code) for source in source_names]
+ subtitled = subtitled or any(item[0] for item in evidence)
+ if not uncensored:
+ uncensored = any(item[1] for item in evidence)
+ cracked = any(item[2] for item in evidence)
badges = []
if subtitled:
badges.append("中字")
diff --git a/src/peach/web_catalog.py b/src/peach/web_catalog.py
index 00574f1a6..9c69cdf2e 100644
--- a/src/peach/web_catalog.py
+++ b/src/peach/web_catalog.py
@@ -46,7 +46,7 @@ from .western_artwork import artwork_key
GROUP_TABLES = ('asset', 'asset_entity', 'entity')
-def download_source_names(contract: WebContract) -> tuple[tuple[int, tuple[str, ...]], ...]:
+def _download_source_names(contract: WebContract) -> tuple[tuple[int, tuple[str, ...]], ...]:
"""入库资产对应的云下载种子名与网盘任务名,给版次判据当来源证据。
一个资产绑过几次下载时取最后一次,文件是那一次落地的。整表读一次只要几毫秒,
@@ -71,7 +71,7 @@ def download_source_names(contract: WebContract) -> tuple[tuple[int, tuple[str,
def _group_cached(contract: WebContract, key: str, fn):
"""分组快照及其派生缓存:账本三张表之外,版次还看种子名。"""
- return contract.cached_until_changed(key, fn, download_source_names(contract),
+ return contract.cached_until_changed(key, fn, _download_source_names(contract),
tables=GROUP_TABLES)
#: 卡片随每条记录返回的女优上限。共演作品必须带上全部出镜者,而不是只留第一位;
@@ -497,7 +497,7 @@ def q_items(contract: WebContract, args):
r["creator"] = r["creator_entity"]["name"]
r["performer_total"] = len(refs) or len(all_performers)
r["_entity_kinds"] = tuple(canonical)
- sources = dict(download_source_names(contract)) if rows else {}
+ sources = dict(_download_source_names(contract)) if rows else {}
for r in rows:
r["cost"] = cost_for(r['location'], r.get('path'))
r["has_thumb"] = contract.has_snapshot(r["snapshot_path"])
@@ -739,7 +739,7 @@ def _read_group_rows(contract: WebContract) -> list[dict]:
"JOIN entity e ON e.id=ae.entity_id "
"WHERE e.kind='tag'"):
tags.setdefault(int(row["aid"]), []).append(str(row["name"]))
- sources = dict(download_source_names(contract))
+ sources = dict(_download_source_names(contract))
coded = []
for row in rows:
row['raw_code'] = row['code']
@@ -1108,7 +1108,7 @@ def q_item(contract: WebContract, aid):
attach_jav_display_fields(
d, [tag["k"] for tag in d["tags"]],
[kind for kind, values in d["entities"].items() if values],
- dict(download_source_names(contract)).get(d["id"], ()),
+ dict(_download_source_names(contract)).get(d["id"], ()),
)
d.pop("snapshot_path", None); d.pop("path", None)
attach_multipart_groups(contract, [d])
@@ -1236,7 +1236,7 @@ def q_related(contract: WebContract, aid, limit=24):
for asset_id, tag in con_tags(contract, related_ids, marks):
related_tags.setdefault(asset_id, []).append(tag)
_name_shared_tags(contract, aid, picked, related_tags)
- sources = dict(download_source_names(contract)) if picked else {}
+ sources = dict(_download_source_names(contract)) if picked else {}
for d in picked:
d["cost"] = cost_for(d['location'], d.get('path'))
d["has_thumb"] = contract.has_snapshot(d["snapshot_path"])
diff --git a/tests/test_fastapi_api.py b/tests/test_fastapi_api.py
index d872f7a4c..3f8cfeb4a 100644
--- a/tests/test_fastapi_api.py
+++ b/tests/test_fastapi_api.py
@@ -146,7 +146,8 @@ CREATE TABLE review_decision(
reviewer TEXT NOT NULL DEFAULT 'local-default',note TEXT NOT NULL DEFAULT '',updated_at TEXT NOT NULL,
PRIMARY KEY(category,item_key));
CREATE TABLE download_task(
- id INTEGER PRIMARY KEY, code TEXT, state TEXT NOT NULL, ledger_path TEXT, asset_id INTEGER);
+ id INTEGER PRIMARY KEY, code TEXT, state TEXT NOT NULL, ledger_path TEXT, asset_id INTEGER,
+ display_name TEXT NOT NULL DEFAULT '', remote_name TEXT);
CREATE TABLE genre_decision(
source_genre TEXT PRIMARY KEY, raw_genre TEXT NOT NULL, peach_tag TEXT, decided_at TEXT NOT NULL,
CHECK(peach_tag IS NULL OR length(trim(peach_tag))>0));
diff --git a/tests/test_rm_web.py b/tests/test_rm_web.py
index 7c2ab06c4..6c70b82c3 100644
--- a/tests/test_rm_web.py
+++ b/tests/test_rm_web.py
@@ -3052,6 +3052,9 @@ class JavModeAndCoverTests(unittest.TestCase):
("+++ [FHDC] SONE-804 作品标题", "SONE-804ch")), ["中字"])
self.assertEqual(badges("ABP-123.mp4", "ABP-123", (), ("ABP-123 中文字幕",)), ["中字"])
self.assertEqual(badges("ABP-123.mp4", "ABP-123", (), ("ABP-123-U",)), ["无码"])
+ # 磁力名与任务名各给一半证据时取并集,不看先后。
+ self.assertEqual(badges("ABP-123.mp4", "ABP-123", (), ("ABP-123-U", "ABP-123 无码破解")),
+ ["无码破解"])
# 没有下载任务就没有来源名,结果与只看文件名相同。
self.assertEqual(badges("ABP-123.mp4", "ABP-123", (), ()), [])
# 种子名是另一部片的标记,不能借来:番号对不上就不算这部的版次。