first commit

backend/tests/test_git_parser.py

@@ -0,0 +1,40 @@
+"""Tests for Git parser and candidate selection."""
+
+import os
+from pathlib import Path
+
+import pytest
+
+from app.dataset.git_parser import score_commit, select_candidates
+
+
+def _make_repo(tmp_path: Path) -> str:
+    repo = tmp_path / "repo"
+    repo.mkdir()
+    os.system(
+        f'cd {repo} && git init -q && git config user.email "test@example.com" && git config user.name "Test"'
+    )
+    (repo / "main.py").write_text("print('hello')\n")
+    os.system(f'cd {repo} && git add . && git commit -q -m "init"')
+    (repo / "main.py").write_text("print('world')\n")
+    os.system(f'cd {repo} && git add . && git commit -q -m "update main"')
+    (repo / "app.java").write_text("class App {}\n")
+    os.system(f'cd {repo} && git add . && git commit -q -m "add java app"')
+    return str(repo)
+
+
+def test_score_commit_prefers_message_and_size(tmp_path):
+    repo = _make_repo(tmp_path)
+    from app.dataset.git_parser import list_commits
+
+    commits = list_commits(repo)
+    assert len(commits) >= 2
+    scores = [score_commit(c) for c in commits]
+    assert all(0 <= s <= 1 for s in scores)
+
+
+def test_select_candidates(tmp_path):
+    repo = _make_repo(tmp_path)
+    candidates = select_candidates(repo, count=2, languages=["python", "java"])
+    assert len(candidates) <= 2
+    assert all(c.repo == "repo" for c in candidates)