kcloud/re-xxm
2
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(classify): tighten scoring with reaction co-occurrence and date negatives

Browse Source 
强正向 reaction 需标题同时含 require_title；移除听/看等宽泛词，新增弱正向组；charge 命中>=2 记+3、==1 记+1；负向补充录播/录像并新增标题带日期强负向；时长仅在已命中 reaction/charge 时加权；discover 关键词改为单个咻咻满、pages=3、window_days=7，notify.min_score=4；同步示例配置、project.md/README 与相关测试。
This commit is contained in:
kairee committed 2026-11-01 19:00:48 +08:00
1 parent e04908ccad
commit 742c67a1c3
9 files changed
+203 -100
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

README.md
+1 -1
View File
@@ -19,7 +19,7 @@
| 视频入库 | 可用 | 搜索候选经 Telegram 确认入库；私聊发送视频链接、BV 号、av 号或 `b23.tv` 短链手动入库。只写本地库。 | | 视频入库 | 可用 | 搜索候选经 Telegram 确认入库；私聊发送视频链接、BV 号、av 号或 `b23.tv` 短链手动入库。只写本地库。 |
| 视频日快照 | 可用（P1） | 逐条采集播放、点赞、投币、收藏、转发、评论、弹幕。 | | 视频日快照 | 可用（P1） | 逐条采集播放、点赞、投币、收藏、转发、评论、弹幕。 |
| 粉丝追踪 | 可用（P1） | 每日记录咻咻满等配置 UP 的粉丝数。 | | 粉丝追踪 | 可用（P1） | 每日记录咻咻满等配置 UP 的粉丝数。 |
| 搜索发现与候选池 | 可用（P3） | 每小时搜索前 2 页，分类打分后写入候选池。 | | 搜索发现与候选池 | 可用（P3） | 每小时按 `discover` 配置搜索并分类打分后写入候选池。 |
| Telegram 通知与入库 | 可用（P3） | 入库 / 忽略 / 拉黑按钮；私聊消息手动入库。 | | Telegram 通知与入库 | 可用（P3） | 入库 / 忽略 / 拉黑按钮；私聊消息手动入库。 |
| 本地数据看板 | 可用（P2/P9） | 只读总览、视频榜、UP 主榜、内容效率、视频/UP 详情、粉丝曲线、数据说明、候选池和系统状态。 | | 本地数据看板 | 可用（P2/P9） | 只读总览、视频榜、UP 主榜、内容效率、视频/UP 详情、粉丝曲线、数据说明、候选池和系统状态。 |
| 访问统计（PV/UV） | 可用 | 前端埋点上报，按客户端 IP 与页面去重；先落内存、每分钟批量入库；系统页展示近 30 天。 | | 访问统计（PV/UV） | 可用 | 前端埋点上报，按客户端 IP 与页面去重；先落内存、每分钟批量入库；系统页展示近 30 天。 |
config.docker.example.json
+16 -19
View File
@@ -67,33 +67,30 @@
"candidate_ttl_days": 7 "candidate_ttl_days": 7
}, },
"discover": { "discover": {
"keywords": ["咻咻满 reaction", "咻咻满 充电"], "keywords": ["咻咻满"],
"pages": 2, "pages": 3,
"window_days": 14 "window_days": 7
}, },
"notify": { "notify": {
"min_score": 2, "min_score": 4,
"max_items_per_message": 10 "max_items_per_message": 10
}, },
"classify": { "classify": {
"require_title": "咻咻满", "require_title": "咻咻满",
"self_mid": 37754047, "self_mid": 37754047,
"positive": [ "reaction": ["reaction", "反应"],
"reaction", "charge": ["加急", "特急", "充电", "专场", "转场"],
"re", "weak_positive": ["双人", "路人", "第一次", "首次"],
"反应", "negative": [
"看", "歌切",
"听", "切片",
"初听", "纯享",
"首次", "搬运",
"点评", "完整版",
"锐评", "录播",
"路人", "录像"
"震惊",
"双人"
], ],
"charge": ["充电", "加急", "专场"], "date_negative": true,
"negative": ["歌切", "切片", "纯享", "搬运", "完整版"],
"duration_boost": { "gt_30min": 1, "gt_60min": 2 } "duration_boost": { "gt_30min": 1, "gt_60min": 2 }
}, },
"heat_weights": { "heat_weights": {
config.example.json
+9 -20
View File
@@ -77,33 +77,22 @@
"candidate_ttl_days": 7 "candidate_ttl_days": 7
}, },
"discover": { "discover": {
"keywords": ["咻咻满 reaction", "咻咻满 充电"], "keywords": ["咻咻满"],
"pages": 2, "pages": 3,
"window_days": 14 "window_days": 7
}, },
"notify": { "notify": {
"min_score": 2, "min_score": 4,
"max_items_per_message": 10 "max_items_per_message": 10
}, },
"classify": { "classify": {
"require_title": "咻咻满", "require_title": "咻咻满",
"self_mid": 37754047, "self_mid": 37754047,
"positive": [ "reaction": ["reaction", "反应"],
"reaction", "charge": ["加急", "特急", "充电", "专场", "转场"],
"re", "weak_positive": ["双人", "路人", "第一次", "首次"],
"反应", "negative": ["歌切", "切片", "纯享", "搬运", "完整版", "录播", "录像"],
"看", "date_negative": true,
"听",
"初听",
"首次",
"点评",
"锐评",
"路人",
"震惊",
"双人"
],
"charge": ["充电", "加急", "专场"],
"negative": ["歌切", "切片", "纯享", "搬运", "完整版"],
"duration_boost": { "gt_30min": 1, "gt_60min": 2 } "duration_boost": { "gt_30min": 1, "gt_60min": 2 }
}, },
"heat_weights": { "heat_weights": {
project.md
+24 -14
View File
@@ -118,12 +118,16 @@ Telegram 私聊发链接/BV 号 ───┘
], ],
"schedule": { "discover_minutes": 60, "schedule": { "discover_minutes": 60,
"snapshot_at": "00:01", "candidate_ttl_days": 7 }, "snapshot_at": "00:01", "candidate_ttl_days": 7 },
"notify": { "min_score": 2, "max_items_per_message": 10 }, "discover": { "keywords": ["咻咻满"], "pages": 3, "window_days": 7 },
"notify": { "min_score": 4, "max_items_per_message": 10 },
"classify": { "classify": {
"require_title": "咻咻满", "require_title": "咻咻满",
"positive": ["reaction", "re", "反应", "看", "听", "初听", "首次", "点评", "锐评", "路人", "震惊", "双人"], "self_mid": 37754047,
"charge": ["充电", "加急", "专场"], "reaction": ["reaction", "反应"],
"negative": ["歌切", "切片", "纯享", "搬运", "完整版"], "charge": ["加急", "特急", "充电", "专场", "转场"],
"weak_positive": ["双人", "路人", "第一次", "首次"],
"negative": ["歌切", "切片", "纯享", "搬运", "完整版", "录播", "录像"],
"date_negative": true,
"duration_boost": { "gt_30min": 1, "gt_60min": 2 } "duration_boost": { "gt_30min": 1, "gt_60min": 2 }
}, },
"heat_weights": { "view": 0.30, "like": 0.20, "coin": 0.20, "heat_weights": { "view": 0.30, "like": 0.20, "coin": 0.20,
@@ -257,24 +261,30 @@ title 含 "咻咻满"（全半角/大小写归一化后），否则直接丢弃
up_mid 命中黑名单 → 直接排除（永不通知，连候选池都不进） up_mid 命中黑名单 → 直接排除（永不通知，连候选池都不进）
score = 0 score = 0
+2 标题含 reaction|re|反应|看|听|初听|首次|点评|锐评|路人|震惊|双人 +3 标题同时含 "咻咻满" 与 reaction|反应（强正向；只含 reaction 不成立）
+2 标题含 充电|加急|专场 +3 标题含 加急|特急|充电|专场|转场 中任意 >=2 个
+2 时长 > 60 分钟 +1 标题含其中恰好 1 个
+1 时长 > 30 分钟（与上一条不叠加，取大） +1 标题含 双人|路人|第一次|首次（弱正向）
-5 标题含 歌切|切片|纯享|搬运|完整版 +2 时长 > 60 分钟（仅当已命中 reaction 或 charge 时计入）
+1 时长 > 30 分钟（同上，与上一条不叠加，取大）
-6 标题含 歌切|切片|纯享|搬运|完整版|录播|录像
-6 标题带日期（2026-10-28 / 2026/10/25 / 2026.10.26 / 2024年11月24日 / 10月25日）
-3 UP 为 咻咻满本人（mid=37754047） -3 UP 为 咻咻满本人（mid=37754047）
通知阈值 = 2 通知阈值 = 4
0 <= score < 2 且含"咻咻满" → 进看板"候选池"（7 天后删除），不通知 0 <= score < 4 且含"咻咻满" → 进看板"候选池"（7 天后删除），不通知
score < 0 → 丢弃 score < 0 → 丢弃
``` ```
**几条设计依据**： **几条设计依据**：
- 对现有 664 条收藏回测，**92.6% 的标题含 reaction 类词**，规则召回足够；误判代价只是"多发/少发一条通知"，最终由人点击决定； - **强正向必须"咻咻满 + reaction"同现**：B 站搜索会命中标题根本不含关键词的结果（简介/标签/相似内容），所以"标题含 reaction"本身不足以判定为高置信，必须与 `require_title` 同时成立；
- **"直播回放"不再算负向词**：确实有人先直播看、再发直播回放，这类仍是 reaction；改由"UP 黑名单"处理专发回放/切片的号； - **"听/看"等宽泛词已移除**：几乎所有 reaction 标题都含，无区分度；
- **时长是强正向信号**：实测这批 reaction 视频时长普遍 100~336 分钟（`105:05`、`158:24`、`336:31`），30/60 分钟阈值能把大多数切片（十几分钟）区分开； - **时长只在已命中 reaction/charge 时加权**：避免"完整录播/直播录像"这类既无 reaction 也无 charge 的长视频仅靠时长进池；
- **带日期的标题按强负向处理**：`2026-10-28`、`10月25日` 这类通常对应某场直播的录播/回放/录像/自用收藏；实测对真 reaction 标题零误伤，代价是极少数"真 reaction 恰好写日期"会被漏掉；
- **`录播`/`录像` 计负向**：`完整版` 覆盖不到 `完整录播`；`录像` 覆盖 `直播录像`。裸 `回放` 不列入，因为 `【回放/咻咻满reaction】` 之类确为 reaction；
- 时长直接取**搜索结果自带的 `duration` 字段**（格式是 `分钟:秒`，超过 60 分钟写作 `105:5`），**不需要额外请求**； - 时长直接取**搜索结果自带的 `duration` 字段**（格式是 `分钟:秒`，超过 60 分钟写作 `105:5`），**不需要额外请求**；
- 关键词用单个 `咻咻满`（`discover.pages=3` 覆盖 7 天窗口），因此 `require_title` 是过滤非相关结果的关键兜底；
- UP 黑名单在配置文件里维护，可热加载，后续从"误通知"记录里一键加入。 - UP 黑名单在配置文件里维护，可热加载，后续从"误通知"记录里一键加入。
### 3.4 通知内容 ### 3.4 通知内容
reaction_board/classify.py
+61 -12
View File
@@ -47,11 +47,16 @@ def parse_duration(value: Any) -> int:
return 0 return 0
def contains_any(title: str, words: Any) -> bool: def matched_words(title: str, words: Any) -> list[str]:
"""返回标题命中（归一化子串）的词，保持 ``words`` 顺序、跳过空词。"""
normalized = normalize_title(title) normalized = normalize_title(title)
if not isinstance(words, (list, tuple)): if not isinstance(words, (list, tuple)):
return False return []
return any(normalize_title(str(word)) in normalized for word in words if word) return [str(word) for word in words if word and normalize_title(str(word)) in normalized]
def contains_any(title: str, words: Any) -> bool:
return bool(matched_words(title, words))
def passes_required_title(title: str, require_title: str) -> bool: def passes_required_title(title: str, require_title: str) -> bool:
@@ -60,6 +65,30 @@ def passes_required_title(title: str, require_title: str) -> bool:
return normalize_title(require_title) in normalize_title(title) return normalize_title(require_title) in normalize_title(title)
# 标题带日期通常意味着录播/回放/录像（按具体某场直播命名），作为强负向信号。
# 覆盖 `2026-10-28` / `2026/10/25` / `2026.10.26` / `2024年11月24日` / `10月25日`。
_DATE_PATTERNS = (
r"\d{4}\s*[-/.]\s*\d{1,2}\s*[-/.]\s*\d{1,2}",
r"\d{4}\s*年\s*\d{1,2}\s*月(?:\s*\d{1,2}\s*日)?",
r"\d{1,2}\s*月\s*\d{1,2}\s*日",
)
_DATE_RE = re.compile("|".join(_DATE_PATTERNS))
# 各信号的权重（见 project.md §3.3）。
REACTION_SCORE = 3
CHARGE_MULTI_SCORE = 3
CHARGE_SINGLE_SCORE = 1
WEAK_SCORE = 1
NEGATIVE_SCORE = 6
DATE_SCORE = 6
SELF_SCORE = 3
def date_hits(title: str) -> list[str]:
"""返回标题中的日期片段，用于判定「带日期≈录播/回放」。"""
return _DATE_RE.findall(title or "")
def score_item( def score_item(
*, *,
title: str, title: str,
@@ -67,13 +96,31 @@ def score_item(
duration_seconds: int, duration_seconds: int,
classify: Mapping[str, Any], classify: Mapping[str, Any],
) -> int: ) -> int:
"""按 project.md §3.3 计算候选分数。""" """按 project.md §3.3 计算候选分数。
score = 0
if contains_any(title, classify.get("positive")):
score += 2
if contains_any(title, classify.get("charge")):
score += 2
强正向 ``reaction`` 只在标题同时命中 ``require_title`` 时成立；时长只作为
已存在 reaction / charge 信号时的加权，避免纯长录像仅靠时长进池。
"""
score = 0
require_title = str(classify.get("require_title") or "")
reaction_matched = bool(matched_words(title, classify.get("reaction")))
reaction_hit = reaction_matched and (
not require_title or passes_required_title(title, require_title)
)
if reaction_hit:
score += REACTION_SCORE
charge_hits = matched_words(title, classify.get("charge"))
if len(charge_hits) >= 2:
score += CHARGE_MULTI_SCORE
elif len(charge_hits) == 1:
score += CHARGE_SINGLE_SCORE
if matched_words(title, classify.get("weak_positive")):
score += WEAK_SCORE
if reaction_hit or charge_hits:
boost = classify.get("duration_boost") or {} boost = classify.get("duration_boost") or {}
gt_60 = int(boost.get("gt_60min", 2)) gt_60 = int(boost.get("gt_60min", 2))
gt_30 = int(boost.get("gt_30min", 1)) gt_30 = int(boost.get("gt_30min", 1))
@@ -82,8 +129,10 @@ def score_item(
elif duration_seconds > 30 * 60: elif duration_seconds > 30 * 60:
score += gt_30 score += gt_30
if contains_any(title, classify.get("negative")): if matched_words(title, classify.get("negative")):
score -= 5 score -= NEGATIVE_SCORE
if classify.get("date_negative", True) and date_hits(title):
score -= DATE_SCORE
if up_mid is not None and int(up_mid) == int(classify.get("self_mid", 37754047)): if up_mid is not None and int(up_mid) == int(classify.get("self_mid", 37754047)):
score -= 3 score -= SELF_SCORE
return score return score
reaction_board/config.py
+14 -11
View File
@@ -25,12 +25,11 @@ from .client_ip import (
DEFAULT_CLASSIFY: dict[str, Any] = { DEFAULT_CLASSIFY: dict[str, Any] = {
"require_title": "咻咻满", "require_title": "咻咻满",
"self_mid": 37754047, "self_mid": 37754047,
"positive": [ "reaction": ["reaction", "反应"],
"reaction", "re", "反应", "看", "听", "初听", "首次", "charge": ["加急", "特急", "充电", "专场", "转场"],
"点评", "锐评", "路人", "震惊", "双人", "weak_positive": ["双人", "路人", "第一次", "首次"],
], "negative": ["歌切", "切片", "纯享", "搬运", "完整版", "录播", "录像"],
"charge": ["充电", "加急", "专场"], "date_negative": True,
"negative": ["歌切", "切片", "纯享", "搬运", "完整版"],
"duration_boost": {"gt_30min": 1, "gt_60min": 2}, "duration_boost": {"gt_30min": 1, "gt_60min": 2},
} }
@@ -87,11 +86,11 @@ DEFAULT_CONFIG: dict[str, Any] = {
"candidate_ttl_days": 7, "candidate_ttl_days": 7,
}, },
"discover": { "discover": {
"keywords": ["咻咻满 reaction", "咻咻满 充电"], "keywords": ["咻咻满"],
"pages": 2, "pages": 3,
"window_days": 14, "window_days": 7,
}, },
"notify": {"min_score": 2, "max_items_per_message": 10}, "notify": {"min_score": 4, "max_items_per_message": 10},
"classify": DEFAULT_CLASSIFY, "classify": DEFAULT_CLASSIFY,
"heat_weights": DEFAULT_HEAT_WEIGHTS, "heat_weights": DEFAULT_HEAT_WEIGHTS,
"analytics": DEFAULT_ANALYTICS, "analytics": DEFAULT_ANALYTICS,
@@ -500,8 +499,12 @@ def _build_config(data: Mapping[str, Any], path: Path) -> AppConfig:
classify["self_mid"] = _as_int( classify["self_mid"] = _as_int(
classify.get("self_mid", 37754047), "classify.self_mid", minimum=1 classify.get("self_mid", 37754047), "classify.self_mid", minimum=1
) )
for key in ("positive", "charge", "negative"): for key in ("reaction", "charge", "weak_positive", "negative"):
classify[key] = _as_str_list(classify.get(key, []), f"classify.{key}") classify[key] = _as_str_list(classify.get(key, []), f"classify.{key}")
date_negative = classify.get("date_negative", True)
if not isinstance(date_negative, bool):
raise ConfigError("classify.date_negative 必须是布尔值")
classify["date_negative"] = date_negative
boost = classify.get("duration_boost", {}) boost = classify.get("duration_boost", {})
if not isinstance(boost, Mapping): if not isinstance(boost, Mapping):
raise ConfigError("classify.duration_boost 必须是对象") raise ConfigError("classify.duration_boost 必须是对象")
tests/test_classify.py
+70 -19
View File
@@ -10,6 +10,8 @@ sys.stdout.reconfigure(encoding="utf-8")
from reaction_board.classify import ( # noqa: E402 from reaction_board.classify import ( # noqa: E402
clean_title, clean_title,
contains_any, contains_any,
date_hits,
matched_words,
normalize_title, normalize_title,
parse_duration, parse_duration,
passes_required_title, passes_required_title,
@@ -19,13 +21,22 @@ from reaction_board.classify import ( # noqa: E402
CLASSIFY = { CLASSIFY = {
"require_title": "咻咻满", "require_title": "咻咻满",
"self_mid": 37754047, "self_mid": 37754047,
"positive": ["reaction", "re", "反应", "看", "听", "初听", "首次", "点评", "锐评", "路人", "震惊", "双人"], "reaction": ["reaction", "反应"],
"charge": ["充电", "加急", "专场"], "charge": ["加急", "特急", "充电", "专场", "转场"],
"negative": ["歌切", "切片", "纯享", "搬运", "完整版"], "weak_positive": ["双人", "路人", "第一次", "首次"],
"negative": ["歌切", "切片", "纯享", "搬运", "完整版", "录播", "录像"],
"date_negative": True,
"duration_boost": {"gt_30min": 1, "gt_60min": 2}, "duration_boost": {"gt_30min": 1, "gt_60min": 2},
} }
def score(title, up_mid=100, duration=0, classify=CLASSIFY):
return score_item(
title=title, up_mid=up_mid,
duration_seconds=duration, classify=classify,
)
def test_title_helpers(): def test_title_helpers():
assert normalize_title("咻咻满 REACTION") == "咻咻满 reaction" assert normalize_title("咻咻满 REACTION") == "咻咻满 reaction"
assert clean_title('<em class="keyword">咻咻满</em> reaction') == "咻咻满 reaction" assert clean_title('<em class="keyword">咻咻满</em> reaction') == "咻咻满 reaction"
@@ -34,6 +45,8 @@ def test_title_helpers():
assert not passes_required_title("其他 UP 的 reaction", "咻咻满") assert not passes_required_title("其他 UP 的 reaction", "咻咻满")
assert contains_any("咻咻满 REACTION", ["reaction"]) assert contains_any("咻咻满 REACTION", ["reaction"])
assert not contains_any("咻咻满", []) assert not contains_any("咻咻满", [])
assert matched_words("咻咻满 REACTION 反应", ["reaction", "反应"]) == ["reaction", "反应"]
assert matched_words("咻咻满", []) == []
def test_parse_duration(): def test_parse_duration():
@@ -46,30 +59,68 @@ def test_parse_duration():
assert parse_duration(-5) == 0 assert parse_duration(-5) == 0
def test_score_rules(): def test_date_hits():
def score(title, up_mid=100, duration=0): assert date_hits("咻咻满-2026-10-28-素颜满") == ["2026-10-28"]
return score_item( assert date_hits("【咻咻满】10月25日直播片段") == ["10月25日"]
title=title, up_mid=up_mid, assert date_hits("咻咻满 ☆ 2026.10.26") == ["2026.10.26"]
duration_seconds=duration, classify=CLASSIFY, assert date_hits("【咻咻满】2026/10/25直播回放") == ["2026/10/25"]
) assert date_hits("咻咻满：《发如雪》2024年10月5日直播歌切") == ["2024年10月5日"]
assert date_hits("咻咻满 reaction") == []
assert score("咻咻满 reaction", duration=60) == 2
assert score("咻咻满 充电", duration=60) == 2 def test_score_strong_reaction_requires_title():
# 强正向 reaction 只在标题含 require_title 时成立。
assert score("咻咻满 reaction", duration=60) == 3
assert score("咻咻满 反应", duration=60) == 3
assert score("其他 UP 的 reaction", duration=61 * 60) == 0
def test_score_charge_levels():
assert score("咻咻满 充电", duration=60) == 1
assert score("咻咻满 加急专场", duration=60) == 3
assert score("咻咻满 reaction 充电", duration=60) == 4 assert score("咻咻满 reaction 充电", duration=60) == 4
assert score("咻咻满 reaction", duration=61 * 60) == 4 assert score("咻咻满 reaction 加急充电专场", duration=60) == 6
assert score("咻咻满 reaction", duration=31 * 60) == 3
assert score("咻咻满 reaction", duration=30 * 60) == 2
assert score("咻咻满 切片", duration=60) == -5 def test_score_weak_positive():
assert score("咻咻满 双人", duration=60) == 1
assert score("咻咻满 reaction 双人", duration=60) == 4
def test_duration_only_counts_with_signal():
assert score("咻咻满 reaction", duration=61 * 60) == 5
assert score("咻咻满 reaction", duration=31 * 60) == 4
assert score("咻咻满 reaction", duration=30 * 60) == 3
# 无 reaction / charge 信号时，时长不给分（避免纯长录像进池）。
assert score("咻咻满", duration=61 * 60) == 0
assert score("咻咻满 双人", duration=61 * 60) == 1
def test_score_negative_and_self():
assert score("咻咻满 切片", duration=60) == -6
assert score("咻咻满 reaction 切片", duration=60) == -3 assert score("咻咻满 reaction 切片", duration=60) == -3
assert score("咻咻满 reaction", up_mid=37754047, duration=60) == -1 assert score("咻咻满 reaction 直播录像", duration=60) == -3
assert score("咻咻满", duration=0) == 0 assert score("咻咻满 reaction", up_mid=37754047, duration=60) == 0
assert score("咻咻满", duration=61 * 60) == 2
def test_score_date_negative():
assert score("咻咻满 reaction 2026-10-28 完整录播", duration=60) == -9
assert score("咻咻满 reaction 2026-10-28", duration=60) == -3
assert score("咻咻满 充电 10月25日", duration=60) == -5
disabled = dict(CLASSIFY, date_negative=False)
assert score("咻咻满 reaction 2026-10-28", duration=60, classify=disabled) == 3
def main(): def main():
test_title_helpers() test_title_helpers()
test_parse_duration() test_parse_duration()
test_score_rules() test_date_hits()
test_score_strong_reaction_requires_title()
test_score_charge_levels()
test_score_weak_positive()
test_duration_only_counts_with_signal()
test_score_negative_and_self()
test_score_date_negative()
print("OK: 候选标题归一化、时长解析与分类打分测试全部通过") print("OK: 候选标题归一化、时长解析与分类打分测试全部通过")
tests/test_config.py
+4 -2
View File
@@ -54,9 +54,11 @@ def test_dual_accounts_and_env_override():
assert (cfg.bilibili.write.mid, cfg.bilibili.write.csrf, cfg.bilibili.write.cookie) == (202, "env-write", "env-write-cookie") assert (cfg.bilibili.write.mid, cfg.bilibili.write.csrf, cfg.bilibili.write.cookie) == (202, "env-write", "env-write-cookie")
assert cfg.storage.database_path == Path(tmp) / "custom.db" assert cfg.storage.database_path == Path(tmp) / "custom.db"
assert cfg.up_blacklist_seed == (1, 2) assert cfg.up_blacklist_seed == (1, 2)
assert cfg.discover.keywords == ("咻咻满 reaction", "咻咻满 充电") assert cfg.discover.keywords == ("咻咻满",)
assert cfg.discover.pages == 2 assert cfg.discover.pages == 3
assert cfg.classify["self_mid"] == 37754047 assert cfg.classify["self_mid"] == 37754047
assert cfg.classify["reaction"] == ["reaction", "反应"]
assert cfg.classify["date_negative"] is True
def test_flat_and_missing_account_rejected(): def test_flat_and_missing_account_rejected():
tests/test_discovery.py
+4 -2
View File
@@ -40,12 +40,14 @@ def write_config(tmp, **discover):
"classify": { "classify": {
"require_title": "咻咻满", "require_title": "咻咻满",
"self_mid": 37754047, "self_mid": 37754047,
"positive": ["reaction"], "reaction": ["reaction"],
"charge": ["充电"], "charge": ["充电"],
"weak_positive": [],
"negative": ["切片"], "negative": ["切片"],
"date_negative": True,
"duration_boost": {"gt_30min": 1, "gt_60min": 2}, "duration_boost": {"gt_30min": 1, "gt_60min": 2},
}, },
"notify": {"min_score": 2, "max_items_per_message": 10}, "notify": {"min_score": 3, "max_items_per_message": 10},
} }
path = Path(tmp) / "config.json" path = Path(tmp) / "config.json"
path.write_text(json.dumps(data, ensure_ascii=False), encoding="utf-8") path.write_text(json.dumps(data, ensure_ascii=False), encoding="utf-8")