feat: 拆分 wecom_shown_keys 与 movement_baseline 历史层

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
2026-07-14 10:31:49 +08:00
parent 02e97e057f
commit f563239e0b
6 changed files with 551 additions and 5 deletions

View File

@@ -0,0 +1,65 @@
# tests/test_board_history.py
from __future__ import annotations
import json
import os
import tempfile
import unittest
from pathlib import Path
from unittest.mock import patch
from daily.board_history import extract_shown_keys, load_recent_shown_keys, merge_wecom_shown_into_data
from daily.config import board_dedup_days, news_backfill_enabled
class ConfigDiversityTests(unittest.TestCase):
def test_board_dedup_days_default(self):
with patch.dict(os.environ, {}, clear=True):
self.assertEqual(board_dedup_days(), 7)
def test_news_backfill_default_off(self):
with patch.dict(os.environ, {}, clear=True):
self.assertFalse(news_backfill_enabled())
class ShownKeysTests(unittest.TestCase):
def test_extract_github_repo_keys(self):
items = [{"repo": "a/b"}, {"repo": "c/d"}]
self.assertEqual(extract_shown_keys("github_trending", items), ["a/b", "c/d"])
def test_load_recent_reads_wecom_shown_not_baseline(self):
with tempfile.TemporaryDirectory() as tmp:
out = Path(tmp)
# 前日shown 只有 x/ybaseline raw 含 a/b —— 周去重只能看到 x/y
payload = {
"data": {
"date": "2026-07-13",
"movement_baseline": {
"github_trending": [{"repo": "a/b"}, {"repo": "x/y"}],
},
"wecom_shown_keys": {"github_trending": ["x/y"]},
}
}
(out / "2026-07-13.data.json").write_text(
json.dumps(payload, ensure_ascii=False), encoding="utf-8"
)
with patch("daily.board_history.OUTPUT_DIR", out):
keys = load_recent_shown_keys("2026-07-14", lookback_days=7)
self.assertEqual(keys["github_trending"], {"x/y"})
self.assertNotIn("a/b", keys["github_trending"])
def test_merge_shown_does_not_touch_baseline(self):
data = {
"movement_baseline": {"github_trending": [{"repo": "raw/one"}]},
}
merged = merge_wecom_shown_into_data(
data, {"github_trending": ["shown/one"]}
)
self.assertEqual(
merged["movement_baseline"]["github_trending"][0]["repo"], "raw/one"
)
self.assertEqual(merged["wecom_shown_keys"]["github_trending"], ["shown/one"])
if __name__ == "__main__":
unittest.main()

View File

@@ -44,6 +44,12 @@ class ConfigHelpersTests(unittest.TestCase):
with patch.dict(os.environ, {}, clear=True):
self.assertEqual(delta_baseline_fallback(), "full")
def test_wecom_delta_pad_default_true(self):
from daily.config import wecom_delta_pad
with patch.dict(os.environ, {}, clear=True):
self.assertTrue(wecom_delta_pad())
class NewsPushedLinksTests(unittest.TestCase):
def test_filter_and_record_roundtrip(self):
@@ -121,7 +127,7 @@ class DeltaFormatTests(unittest.TestCase):
]
text = build_skills_delta_sections(moves, [])
self.assertIn("Skills Trending 变化", text)
self.assertIn("[新入 #3]", text)
self.assertNotIn("[新入 #", text)
self.assertNotIn("Skills Hot 变化", text)
def test_github_delta_omits_stable_board(self):
@@ -129,14 +135,268 @@ class DeltaFormatTests(unittest.TestCase):
movement = {
"github_trending_moves": [
{"repo": "a/b", "url": "https://github.com/a/b", "rank": 1, "language": "Go"}
{
"repo": "a/b",
"url": "https://github.com/a/b",
"rank": 1,
"language": "Go",
"description": "open-source codebase and curriculum",
}
],
"github_emerging_moves": [],
"github_topic_moves": [],
}
text = build_github_delta_sections(movement, topic_name="llm")
with patch(
"daily.format_wecom.localize_brief_descriptions",
return_value={"github:a/b": "开源代码库与课程体系"},
):
text = build_github_delta_sections(movement, topic_name="llm")
self.assertIn("开源代码库", text)
self.assertIn("GitHub Trending 变化", text)
self.assertNotIn("新兴", text)
self.assertNotIn("[新入 #", text)
self.assertNotIn("\n > ", text)
def test_delta_pad_groups_same_source_moves(self):
from daily.format_wecom import build_skills_delta_sections
moves = [
{
"id": f"lllllllama/rigorpilot-skills/s{i}",
"title": f"s{i}",
"source": "lllllllama/rigorpilot-skills",
"installs": 250 - i,
"installs_fmt": str(250 - i),
"link": f"https://www.skills.sh/lllllllama/rigorpilot-skills/s{i}",
"description": f"skill {i}",
}
for i in range(1, 11)
]
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
with patch("daily.format_wecom.needs_chinese", return_value=False):
text = build_skills_delta_sections(moves, [], trending_limit=10, pad=True)
self.assertIn("10 skills", text)
self.assertNotIn("[**s2**]", text)
def test_delta_pad_fills_skills_to_limit(self):
from daily.format_wecom import build_skills_delta_sections
moves = [
{
"id": "a/b/new",
"rank": 3,
"title": "new",
"source": "a/b",
"installs": 99,
"installs_fmt": "99",
"link": "https://skills.sh/a/b/new",
"description": "new skill",
}
]
full = [
{
"id": "a/b/new",
"title": "new",
"source": "a/b",
"installs": 99,
"installs_fmt": "99",
"link": "https://skills.sh/a/b/new",
"description": "new skill",
},
*[
{
"id": f"src{i}/skill",
"title": "skill",
"source": f"src{i}/pkg",
"installs": 100 - i,
"installs_fmt": str(100 - i),
"link": f"https://skills.sh/src{i}/pkg/skill",
"description": f"skill from src{i}",
}
for i in range(1, 12)
],
]
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
with patch("daily.format_wecom.needs_chinese", return_value=False):
text = build_skills_delta_sections(
moves,
[],
trending_full=full,
hot_full=[],
trending_limit=10,
pad=True,
)
self.assertIn("Skills Trending Top 10", text)
self.assertNotIn("Skills Trending 变化", text)
def test_delta_pad_uses_large_pool_when_recent_excludes_top(self):
from daily.format_wecom import build_skills_delta_sections
full_small = [
{
"id": f"seen/src/s{i}",
"title": f"s{i}",
"source": "seen/src",
"installs": 100 - i,
"installs_fmt": str(100 - i),
"link": f"https://skills.sh/seen/src/s{i}",
"description": f"seen {i}",
}
for i in range(1, 11)
]
full_large = [
{
"id": f"fresh/src{n}/skill",
"title": "skill",
"source": f"fresh/src{n}",
"installs": 50 - n,
"installs_fmt": str(50 - n),
"link": f"https://skills.sh/fresh/src{n}/skill",
"description": f"fresh {n}",
}
for n in range(1, 11)
]
recent = {f"seen/src/s{i}" for i in range(1, 11)}
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
with patch("daily.format_wecom.needs_chinese", return_value=False):
small = build_skills_delta_sections(
[],
[],
trending_full=full_small,
trending_limit=10,
pad=True,
recent_trending=recent,
)
large = build_skills_delta_sections(
[],
[],
trending_full=full_large,
trending_limit=10,
pad=True,
recent_trending=recent,
)
self.assertNotIn("Skills Trending Top 10", small)
self.assertIn("Skills Trending Top 10", large)
self.assertIn("fresh/src1", large)
def test_delta_pad_fills_github_to_limit(self):
from daily.format_wecom import build_github_delta_sections
movement = {"github_trending_moves": [], "github_emerging_moves": [], "github_topic_moves": []}
full = [
{
"repo": f"org/r{i}",
"url": f"https://github.com/org/r{i}",
"language": "Go",
"stars_today_fmt": "100",
"total_stars_fmt": "1K",
"description": f"repo {i}",
"desc_short": f"repo {i}",
}
for i in range(1, 12)
]
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
text = build_github_delta_sections(
movement,
topic_name="llm",
github_trending=full,
trending_limit=10,
pad=True,
)
self.assertIn("GitHub Trending Top 10", text)
self.assertNotIn("GitHub Trending 变化", text)
def test_delta_pad_skips_recent_skills(self):
from daily.format_wecom import build_skills_delta_sections
full = [
{
"id": f"x/y/s{i}",
"title": f"s{i}",
"source": "x/y",
"installs": 100 - i,
"installs_fmt": str(100 - i),
"link": f"https://skills.sh/x/y/s{i}",
"description": f"skill {i}",
}
for i in range(4, 6)
]
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
with patch("daily.format_wecom.needs_chinese", return_value=False):
text = build_skills_delta_sections(
[],
[],
trending_full=full,
hot_full=[],
trending_limit=10,
pad=True,
recent_trending={f"x/y/s{i}" for i in range(1, 4)},
)
self.assertIn("Skills Trending Top 1", text)
self.assertIn("2 skills", text)
def test_delta_pad_skips_recent_github(self):
from daily.format_wecom import build_github_delta_sections
movement = {"github_trending_moves": [], "github_emerging_moves": [], "github_topic_moves": []}
full = [
{
"repo": f"org/r{i}",
"url": f"https://github.com/org/r{i}",
"language": "Go",
"total_stars_fmt": "1K",
"description": f"repo {i}",
"desc_short": f"repo {i}",
}
for i in range(1, 6)
]
with patch("daily.format_wecom.localize_brief_descriptions", return_value={}):
text = build_github_delta_sections(
movement,
topic_name="llm",
github_trending=full,
trending_limit=10,
pad=True,
recent_board_keys={"github_trending": {f"org/r{i}" for i in range(1, 4)}},
)
self.assertIn("GitHub Trending Top 2", text)
self.assertIn("org/r4", text)
self.assertNotIn("org/r1", text)
def test_load_recent_board_keys_from_data_json(self):
import json
import tempfile
from pathlib import Path
from daily.delta import load_recent_board_keys
with tempfile.TemporaryDirectory() as tmp:
out = Path(tmp)
payload = {
"data": {
"date": "2026-07-09",
"movement_baseline": {
"skills_trending": [
{"id": "a/b/raw", "title": "raw", "source": "a/b"},
],
"skills_hot": [],
"github_trending": [{"repo": "org/raw"}],
"github_emerging": [],
"github_topic": [],
},
"wecom_shown_keys": {
"skills_trending": ["a/b/foo"],
"github_trending": ["org/bar"],
},
}
}
(out / "2026-07-09.data.json").write_text(json.dumps(payload), encoding="utf-8")
with patch("daily.board_history.OUTPUT_DIR", out):
recent = load_recent_board_keys("2026-07-10", lookback_days=7)
self.assertIn("a/b/foo", recent["skills_trending"])
self.assertIn("org/bar", recent["github_trending"])
self.assertNotIn("a/b/raw", recent["skills_trending"])
self.assertNotIn("org/raw", recent["github_trending"])
class PushGateTests(unittest.TestCase):
@@ -243,6 +503,25 @@ class E2ESmokeTests(unittest.TestCase):
trending=[],
hot=[],
topic_name="llm",
pad=False,
)
self.assertIn("Skills Trending 变化", out)
self.assertNotIn("Skills Trending Top", out)
class SyncMovementGithubTests(unittest.TestCase):
def test_sync_copies_localized_description(self):
from daily.generate import _sync_movement_github_descriptions
movement = {
"github_trending_moves": [{"repo": "a/b", "description": "english"}],
"github_emerging_moves": [],
"github_topic_moves": [],
}
_sync_movement_github_descriptions(
movement,
github_trending=[{"repo": "a/b", "description": "中文描述"}],
github_emerging=[],
github_topic=[],
)
self.assertEqual(movement["github_trending_moves"][0]["description"], "中文描述")