Initial curated RAGcore source from c16667f0c9397e2bfc8dd0255b114fba23c8d05e

tests/integration/connectors/git/test_git_connector.py

@@ -0,0 +1,99 @@
+"""Integration tests against real local Git object databases."""
+
+from __future__ import annotations
+
+import shutil
+import subprocess
+from pathlib import Path
+from uuid import uuid4
+
+import pytest
+
+from ragcore.domain.sources import GitSourceConfig
+from ragcore.infrastructure.connectors.git import GitConnector
+
+
+def git(root: Path, *arguments: str) -> str:
+    git_executable = shutil.which("git")
+    if git_executable is None:
+        raise RuntimeError("git is required for this integration test")
+    result = subprocess.run(  # noqa: S603 - test arguments are passed without a shell
+        [git_executable, *arguments],
+        cwd=root,
+        check=True,
+        capture_output=True,
+        text=True,
+    )
+    return result.stdout.strip()
+
+
+def repository(tmp_path: Path) -> tuple[Path, str]:
+    root = tmp_path / "repository"
+    root.mkdir()
+    git(root, "init", "--quiet")
+    git(root, "config", "user.email", "tests@example.invalid")
+    git(root, "config", "user.name", "RAGCore Tests")
+    (root / "README.md").write_text("read me", encoding="utf-8")
+    (root / "src").mkdir()
+    (root / "src" / "main.py").write_text("print('safe')\n", encoding="utf-8")
+    git(root, "add", "README.md", "src/main.py")
+    git(root, "commit", "--quiet", "-m", "initial")
+    return root, git(root, "rev-parse", "HEAD")
+
+
+def make_connector(root: Path, commit: str, *, max_items: int = 256) -> GitConnector:
+    return GitConnector(
+        uuid4(),
+        GitSourceConfig("local-checkout", commit, root.resolve()),
+        allowed_roots=(root.resolve(),),
+        max_items=max_items,
+    )
+
+
+@pytest.mark.integration
+@pytest.mark.anyio
+async def test_scan_and_acquire_store_commit_path_and_language_provenance(
+    tmp_path: Path,
+) -> None:
+    root, commit = repository(tmp_path)
+    current = make_connector(root, commit)
+
+    result = await current.scan()
+    item = next(item for item in result.items if item.locator == "src/main.py")
+    acquisition = await current.acquire(item)
+
+    assert result.complete is True
+    assert item.external_id == f"{commit}:src/main.py"
+    assert item.metadata["commit"] == commit
+    assert item.metadata["path"] == "src/main.py"
+    assert item.metadata["language"] == "python"
+    assert b"".join([chunk async for chunk in acquisition.stream]) == b"print('safe')\n"
+    assert acquisition.fingerprint == current.fingerprint(item)
+
+
+@pytest.mark.integration
+@pytest.mark.anyio
+async def test_same_pinned_commit_scan_is_idempotent_and_cursorable(tmp_path: Path) -> None:
+    root, commit = repository(tmp_path)
+    current = make_connector(root, commit, max_items=1)
+
+    first = await current.scan()
+    second = await current.scan()
+    assert first == second
+    assert first.complete is False
+    assert first.next_cursor is not None
+
+    remainder = await current.scan(first.next_cursor)
+    assert remainder.complete is False
+    assert remainder.next_cursor is None
+    assert all(item.external_id.startswith(f"{commit}:") for item in remainder.items)
+
+    (root / "src" / "second.py").write_text("value = 2\n", encoding="utf-8")
+    git(root, "add", "src/second.py")
+    git(root, "commit", "--quiet", "-m", "second")
+    latest = git(root, "rev-parse", "HEAD")
+    old_scan = await current.scan()
+    latest_scan = await make_connector(root, latest).scan()
+    assert all(item.metadata["commit"] == commit for item in old_scan.items)
+    assert all(item.metadata["commit"] == latest for item in latest_scan.items)
+    assert any(item.locator == "src/second.py" for item in latest_scan.items)