prgenius-core 2.2.2__tar.gz → 2.2.3__tar.gz

{prgenius_core-2.2.2 → prgenius_core-2.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prgenius-core
-Version: 2.2.2
+Version: 2.2.3
 Summary: Evidence-backed lookup for big-repo PR contributions. Local-only, zero-runtime, stdlib-first. Renamed from `prgenius` to avoid PyPI name collision with an unrelated 2024 GPT-3 PR-description tool. Distribution renamed again in v0.7.8: `prgenius-kb` → `prgenius-core` (aligned with `misakanet-core` convention).
 Author: pr-genius contributors
 License-Expression: MIT

{prgenius_core-2.2.2 → prgenius_core-2.2.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "prgenius-core"
-version = "2.2.2"
+version = "2.2.3"
 description = "Evidence-backed lookup for big-repo PR contributions. Local-only, zero-runtime, stdlib-first. Renamed from `prgenius` to avoid PyPI name collision with an unrelated 2024 GPT-3 PR-description tool. Distribution renamed again in v0.7.8: `prgenius-kb` → `prgenius-core` (aligned with `misakanet-core` convention)."
 readme = "README.md"
 requires-python = ">=3.9"

{prgenius_core-2.2.2 → prgenius_core-2.2.3}/src/prgenius/__init__.py

@@ -12,7 +12,7 @@ Public surface:
 
 from .parser import load, iter_profiles, iter_case_studies, profile_get, schema_info, clear_profile_cache
 
-__version__ = "2.2.2"
+__version__ = "2.2.3"
 __all__ = [
     "__version__",
     "load",

{prgenius_core-2.2.2 → prgenius_core-2.2.3}/src/prgenius_core.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prgenius-core
-Version: 2.2.2
+Version: 2.2.3
 Summary: Evidence-backed lookup for big-repo PR contributions. Local-only, zero-runtime, stdlib-first. Renamed from `prgenius` to avoid PyPI name collision with an unrelated 2024 GPT-3 PR-description tool. Distribution renamed again in v0.7.8: `prgenius-kb` → `prgenius-core` (aligned with `misakanet-core` convention).
 Author: pr-genius contributors
 License-Expression: MIT

{prgenius_core-2.2.2 → prgenius_core-2.2.3}/src/prgenius_core.egg-info/SOURCES.txt

@@ -29,6 +29,7 @@ tests/test_evaluator.py
 tests/test_evaluator_v163.py
 tests/test_extract_features.py
 tests/test_gh_fallback.py
+tests/test_harvest.py
 tests/test_issue_evaluator.py
 tests/test_maintainer_view.py
 tests/test_mcp.py

prgenius_core-2.2.3/tests/test_harvest.py

@@ -0,0 +1,88 @@
+"""harvest 结果分类 —— 防止把成功 PR 收割成失败反模式.
+
+背景: 2026-10-04 在 anti-patterns/ 里发现 4 条空壳, 其中 3 条是**已合并**的 PR
+("已合并 by @维护者"、"Root Cause: 无 maintainer 评论"), 却被盖上
+`category: pr-failure` 且 Lesson/Solution/Verification 全是 TODO。
+根因是 harvest 的 draft 模板无条件写死 pr-failure。
+
+这里的判据是"有没有真失败信号", 不是"这个 PR 有没有被收割过"。
+"""
+
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+import pytest
+
+_SCRIPTS = Path(__file__).resolve().parents[2] / "scripts"
+if str(_SCRIPTS) not in sys.path:
+    sys.path.insert(0, str(_SCRIPTS))
+
+from harvest import classify_outcome, generate_lesson_draft  # noqa: E402
+
+
+def _pr(merged: bool = True, number: int = 1) -> dict:
+    return {
+        "number": number,
+        "title": "feat: some change",
+        "html_url": f"https://example.com/pull/{number}",
+        "user": {"login": "someone"},
+        "body": "does a thing",
+        "merged_at": "2026-08-13" if merged else None,
+    }
+
+
+BOT = [{"author_association": "NONE", "user": {"login": "dco-bot"},
+        "body": "## DCO Sign-off Required\nPlease add Signed-off-by to your commits"}]
+MAINTAINER = [{"author_association": "MEMBER", "user": {"login": "maintainer"},
+               "body": "Please split this into smaller PRs before we can take it"}]
+
+
+class TestClassifyOutcome:
+    def test_unmerged_pr_is_a_failure(self):
+        assert classify_outcome(_pr(merged=False), "Closed without merge", []) == "rejected"
+
+    def test_merged_with_no_comments_has_no_signal(self):
+        assert classify_outcome(_pr(merged=True), "已合并 by @maintainer", []) == "merged-clean"
+
+    def test_bot_comment_is_not_a_failure_signal(self):
+        # DCO / CI bot 评论不是 maintainer 反馈 —— 与 _extract_key_comments 同口径。
+        # 若这里判成 merged-with-review, Root Cause 段会显示"无 maintainer 评论",
+        # 两段自相矛盾。
+        assert classify_outcome(_pr(merged=True), "已合并 by @maintainer", BOT) == "merged-clean"
+
+    def test_maintainer_comment_on_merged_pr_is_feedback(self):
+        assert classify_outcome(_pr(merged=True), "已合并 by @maintainer", MAINTAINER) == "merged-with-review"
+
+    def test_tiny_comment_is_not_feedback(self):
+        # _extract_key_comments 要求 body 长度 > 10; 分类必须同此门槛。
+        tiny = [{"author_association": "MEMBER", "user": {"login": "m"}, "body": "ok"}]
+        assert classify_outcome(_pr(merged=True), "已合并", tiny) == "merged-clean"
+
+
+class TestLessonDraftIsHonest:
+    def test_merged_clean_pr_is_not_labeled_pr_failure(self):
+        out = generate_lesson_draft("example/repo", _pr(merged=True), "已合并 by @m", [])
+        assert "category: pr-failure" not in out
+        assert "category: merged-reference" in out
+        assert "severity: info" in out
+
+    def test_merged_clean_pr_does_not_prompt_you_to_write_a_lesson(self):
+        # 邀请别人给"没有失败的 PR"写教训 = 邀请编造。
+        out = generate_lesson_draft("example/repo", _pr(merged=True), "已合并 by @m", [])
+        assert "没有失败信号" in out
+        assert "不要往这里填 Lesson" in out
+        assert "提炼以下可复用教训" not in out
+
+    def test_rejected_pr_still_gets_lesson_guidance(self):
+        out = generate_lesson_draft("example/repo", _pr(merged=False), "Closed without merge", MAINTAINER)
+        assert "category: pr-failure" in out
+        assert "提炼以下可复用教训" in out
+        assert "没有失败信号" not in out
+
+    def test_root_cause_matches_the_classification(self):
+        # 两段不能互相矛盾: 说"无 maintainer 评论"就不能同时判成"有反馈可提炼"。
+        out = generate_lesson_draft("example/repo", _pr(merged=True), "已合并 by @m", BOT)
+        assert "无 maintainer 评论" in out
+        assert "category: merged-reference" in out