CodeRabbit Generated Unit Tests: Add pytest tests for discover_tests_candidates: globs, unicode

tests/test_discover_tests_candidates.py

@@ -0,0 +1,146 @@
+# Tests for discover_tests_candidates
+# Framework: pytest (preferred) with standard assertions and parametrization.
+# If your project uses unittest, these tests can be adapted to TestCase easily.
+
+import os
+import sys
+from pathlib import Path
+
+import pytest
+
+try:
+    # Try common import paths; adjust as needed if your module path differs.
+    from discover import discover_tests_candidates
+except ImportError:
+    try:
+        from discovery import discover_tests_candidates
+    except ImportError as err:
+        # Fallback to project-local packages; this keeps tests resilient across refactors.
+        from importlib import import_module as _imp
+        _candidates = None
+        for mod in (
+            "src.discover",
+            "src.discovery",
+            "project.discover",
+            "project.discovery",
+            "tools.discover",
+        ):
+            try:
+                _m = _imp(mod)
+                _candidates = getattr(_m, "discover_tests_candidates", None)
+                if _candidates:
+                    discover_tests_candidates = _candidates  # type: ignore
+                    break
+            except ImportError:
+                continue
+        if not _candidates:
+            raise ImportError() from err
+
+
+def _touch(p: Path, text: str = "") -> None:
+    p.parent.mkdir(parents=True, exist_ok=True)
+    p.write_text(text, encoding="utf-8")
+
+
+@pytest.fixture
+def tmp_tree(tmp_path: Path):
+    # Create a representative test tree with various files and patterns
+    files = {
+        "tests/test_alpha.py": "def test_a(): pass\n",
+        "tests/test_beta.py": "def test_b(): pass\n",
+        "tests/helpers/util.py": "def helper(): pass\n",
+        "package/module_test.py": "def test_c(): pass\n",
+        "package/not_a_test.txt": "content\n",
+        "nested/deep/test_gamma.py": "def test_g(): pass\n",
+        "nested/deep/_private_test.py": "def test_p(): pass\n",
+        "nested/deep/test_🌟.py": "def test_unicode(): pass\n",
+        "nested/deep/__init__.py": "",
+        "README.md": "# docs\n",
+    }
+    for rel, content in files.items():
+        _touch(tmp_path / rel, content)
+    return tmp_path
+
+
+def norm(p: Path) -> str:
+    return str(p).replace("\\", "/")
+
+
+@pytest.mark.parametrize(
+    "inputs,expected_contains,expected_absent",
+    [
+        # Happy path: standard tests/ directory with test_*.py
+        (["tests/"], ["tests/test_alpha.py", "tests/test_beta.py"], ["tests/helpers/util.py"]),
+        # Glob patterns
+        (["**/test_*.py"], ["tests/test_alpha.py", "nested/deep/test_gamma.py"], ["package/not_a_test.txt"]),
+        # Mixed files and dirs
+        (["tests/test_beta.py", "nested/"], ["tests/test_beta.py", "nested/deep/test_gamma.py"], ["package/module_test.py"]),
+        # File with unicode name
+        (["nested/deep/"], ["nested/deep/test_🌟.py"], []),
+    ],
+)
+def test_discover_candidates_basic_patterns(tmp_tree: Path, inputs, expected_contains, expected_absent):
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_tree)
+        results = list(discover_tests_candidates(inputs))
+        results_n = sorted(norm(Path(p)) for p in results)
+        for exp in expected_contains:
+            assert any(r.endswith(exp) for r in results_n), f"Expected {exp} in {results_n}"
+        for bad in expected_absent:
+            assert all(not r.endswith(bad) for r in results_n), f"Did not expect {bad} in {results_n}"
+    finally:
+        os.chdir(cwd)
+
+def test_discover_candidates_ignores_non_python_files(tmp_tree: Path):
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_tree)
+        results = list(discover_tests_candidates(["."]))
+        assert all(r.endswith(".py") for r in results), f"Non-Python files leaked into results: {results}"
+    finally:
+        os.chdir(cwd)
+
+@pytest.mark.parametrize(
+    "inputs",
+    [
+        ([]),
+        (["nonexistent_dir/"]),
+        (["README.md"]),
+    ],
+)
+def test_discover_candidates_empty_or_invalid_inputs(tmp_tree: Path, inputs):
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_tree)
+        results = list(discover_tests_candidates(inputs))
+        # Either empty or gracefully handles invalid paths without raising
+        assert isinstance(results, list) or hasattr(results, "__iter__")
+        # Count should be zero for no valid candidates
+        assert len(list(results)) == 0
+    finally:
+        os.chdir(cwd)
+
+def test_discover_candidates_deduplicates_and_sorts(tmp_tree: Path):
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_tree)
+        res = list(discover_tests_candidates(["tests/test_alpha.py", "tests/", "tests/test_alpha.py"]))
+        # Expect no duplicates
+        normed = [norm(Path(p)) for p in res]
+        assert len(normed) == len(set(normed)), f"Duplicates found: {normed}"
+        # Many implementations return deterministic order; if not, we at least check that sorting is stable
+        assert sorted(normed) == sorted(set(normed))
+    finally:
+        os.chdir(cwd)
+
+def test_discover_candidates_recurses_directories_respecting_test_naming(tmp_tree: Path):
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_tree)
+        res = list(discover_tests_candidates(["package/"]))
+        normed = [norm(Path(p)) for p in res]
+        assert any(p.endswith("package/module_test.py") for p in normed), f"Expected module_test.py, got {normed}"
+        assert all(not p.endswith("package/not_a_test.txt") for p in normed)
+    finally:
+        os.chdir(cwd)