test_cmd_dead.py
python
sha256:d11a87833d5fad6059b7662844bf5448a8911a17cce7a51811f71ad394f248eb
bump to v0.2.0rc13
Human
patch
6 days ago
| 1 | """Comprehensive tests for ``muse code dead``. |
| 2 | |
| 3 | dead.py is the highest-churn code porcelain command (7 changes) and contains |
| 4 | the richest set of private helpers — making unit coverage especially valuable |
| 5 | for catch-regression purposes. |
| 6 | |
| 7 | Coverage |
| 8 | -------- |
| 9 | Unit |
| 10 | _module_is_imported — exact stem, dotted module, suffix, no-match |
| 11 | _matches_path_filter — None passthrough, fnmatch, ** pattern |
| 12 | _find_symbol_span — function, class, method, decorated, variable, |
| 13 | missing symbol, syntax error, parent_class |
| 14 | _delete_symbol_lines — middle removal, head removal, tail removal, |
| 15 | blank-line normalisation |
| 16 | _analyse_file — skips over-limit, non-semantic suffix, Python |
| 17 | ref extraction, import extraction, kind filter, |
| 18 | syntax error graceful return |
| 19 | _is_test_file — test/spec patterns |
| 20 | _DeadCandidate — confidence, reason, to_dict |
| 21 | |
| 22 | Integration (extends mega-suite baseline) |
| 23 | --json schema — results, high_confidence_count, … |
| 24 | --kind filter — only that kind in output |
| 25 | --high-confidence-only — only high in output |
| 26 | --count — scalar integer |
| 27 | --language — restrict to language |
| 28 | --path — restrict to path glob |
| 29 | --workers — capped cap enforced, exits 0 |
| 30 | --compare HEAD — schema correctness |
| 31 | --save-allowlist — writes JSON list |
| 32 | --allowlist — allowlisted addresses excluded |
| 33 | |
| 34 | Security |
| 35 | --delete no --yes — prompts / exits non-zero without confirmation |
| 36 | requires repo — exits non-zero outside repo |
| 37 | |
| 38 | Stress |
| 39 | 200-function Python file — completes under 10 s |
| 40 | 50-file codebase — exits 0 under 10 s |
| 41 | """ |
| 42 | |
| 43 | from __future__ import annotations |
| 44 | |
| 45 | import argparse |
| 46 | |
| 47 | import ast |
| 48 | import json |
| 49 | import pathlib |
| 50 | import textwrap |
| 51 | import time |
| 52 | |
| 53 | import pytest |
| 54 | |
| 55 | from tests.cli_test_helper import CliRunner |
| 56 | from muse.cli.commands.dead import ( |
| 57 | _DeadCandidate, |
| 58 | _FileAnalysis, |
| 59 | _analyse_file, |
| 60 | _delete_symbol_lines, |
| 61 | _find_symbol_span, |
| 62 | _is_test_file, |
| 63 | _matches_path_filter, |
| 64 | _module_is_imported, |
| 65 | ) |
| 66 | |
| 67 | cli = None |
| 68 | runner = CliRunner() |
| 69 | |
| 70 | # --------------------------------------------------------------------------- |
| 71 | # Repo fixture |
| 72 | # --------------------------------------------------------------------------- |
| 73 | |
| 74 | |
| 75 | @pytest.fixture |
| 76 | def repo(tmp_path: pathlib.Path, monkeypatch: pytest.MonkeyPatch) -> pathlib.Path: |
| 77 | """Fresh code-domain repo with self-contained Python modules.""" |
| 78 | monkeypatch.chdir(tmp_path) |
| 79 | monkeypatch.setenv("MUSE_REPO_ROOT", str(tmp_path)) |
| 80 | r = runner.invoke(cli, ["init", "--domain", "code"]) |
| 81 | assert r.exit_code == 0, r.output |
| 82 | |
| 83 | # billing.py — imports utils, uses validate_amount. |
| 84 | (tmp_path / "billing.py").write_text(textwrap.dedent("""\ |
| 85 | from utils import validate_amount |
| 86 | |
| 87 | class Invoice: |
| 88 | def compute_total(self, items: list[int]) -> int: |
| 89 | return sum(items) |
| 90 | |
| 91 | def process_order(invoice: Invoice, items: list[int]) -> int: |
| 92 | if not validate_amount(sum(items)): |
| 93 | raise ValueError("bad amount") |
| 94 | return invoice.compute_total(items) |
| 95 | """)) |
| 96 | |
| 97 | # utils.py — used by billing.py; orphaned_helper is not referenced. |
| 98 | (tmp_path / "utils.py").write_text(textwrap.dedent("""\ |
| 99 | def validate_amount(amount: float) -> bool: |
| 100 | return amount > 0 |
| 101 | |
| 102 | def orphaned_helper(x: int) -> int: |
| 103 | return x * 2 |
| 104 | """)) |
| 105 | |
| 106 | r2 = runner.invoke(cli, ["commit", "-m", "initial"]) |
| 107 | assert r2.exit_code == 0, r2.output |
| 108 | return tmp_path |
| 109 | |
| 110 | |
| 111 | # --------------------------------------------------------------------------- |
| 112 | # Unit — _module_is_imported |
| 113 | # --------------------------------------------------------------------------- |
| 114 | |
| 115 | |
| 116 | class TestModuleIsImported: |
| 117 | def test_stem_match(self) -> None: |
| 118 | assert _module_is_imported("utils.py", {"utils"}) |
| 119 | |
| 120 | def test_dotted_module_match(self) -> None: |
| 121 | assert _module_is_imported("core/store.py", {"core.store"}) |
| 122 | |
| 123 | def test_suffix_match(self) -> None: |
| 124 | assert _module_is_imported("muse/core/store.py", {"muse.core.store"}) |
| 125 | |
| 126 | def test_no_match(self) -> None: |
| 127 | assert not _module_is_imported("utils.py", {"billing", "invoice"}) |
| 128 | |
| 129 | def test_empty_set(self) -> None: |
| 130 | assert not _module_is_imported("utils.py", set()) |
| 131 | |
| 132 | def test_partial_stem_no_match(self) -> None: |
| 133 | # "til" should not match "utils.py" (stem is "utils", not "til"). |
| 134 | assert not _module_is_imported("utils.py", {"til"}) |
| 135 | |
| 136 | def test_stem_inside_dotted_import(self) -> None: |
| 137 | # "utils" is an element of "muse.utils" → should match. |
| 138 | assert _module_is_imported("utils.py", {"muse.utils"}) |
| 139 | |
| 140 | def test_deep_path_stem(self) -> None: |
| 141 | assert _module_is_imported("a/b/c/billing.py", {"billing"}) |
| 142 | |
| 143 | |
| 144 | # --------------------------------------------------------------------------- |
| 145 | # Unit — _matches_path_filter |
| 146 | # --------------------------------------------------------------------------- |
| 147 | |
| 148 | |
| 149 | class TestMatchesPathFilter: |
| 150 | def test_none_always_matches(self) -> None: |
| 151 | assert _matches_path_filter("src/billing.py", None) |
| 152 | assert _matches_path_filter("any/path/here.py", None) |
| 153 | |
| 154 | def test_exact_match(self) -> None: |
| 155 | assert _matches_path_filter("src/billing.py", "src/billing.py") |
| 156 | |
| 157 | def test_glob_wildcard(self) -> None: |
| 158 | assert _matches_path_filter("src/billing.py", "src/*.py") |
| 159 | |
| 160 | def test_no_match(self) -> None: |
| 161 | assert not _matches_path_filter("src/billing.py", "tests/*.py") |
| 162 | |
| 163 | def test_double_star_glob(self) -> None: |
| 164 | assert _matches_path_filter("a/b/c/billing.py", "**/billing.py") |
| 165 | |
| 166 | |
| 167 | # --------------------------------------------------------------------------- |
| 168 | # Unit — _find_symbol_span |
| 169 | # --------------------------------------------------------------------------- |
| 170 | |
| 171 | |
| 172 | class TestFindSymbolSpan: |
| 173 | def test_finds_function(self) -> None: |
| 174 | src = b"def foo(x: int) -> int:\n return x\n" |
| 175 | result = _find_symbol_span(src, "foo", None) |
| 176 | assert result is not None |
| 177 | start, end = result |
| 178 | assert start == 1 |
| 179 | |
| 180 | def test_finds_class(self) -> None: |
| 181 | src = b"class Foo:\n x: int = 1\n" |
| 182 | result = _find_symbol_span(src, "Foo", None) |
| 183 | assert result is not None |
| 184 | |
| 185 | def test_finds_method_in_class(self) -> None: |
| 186 | src = textwrap.dedent("""\ |
| 187 | class Invoice: |
| 188 | def compute_total(self) -> int: |
| 189 | return 0 |
| 190 | """).encode() |
| 191 | result = _find_symbol_span(src, "compute_total", "Invoice") |
| 192 | assert result is not None |
| 193 | start, end = result |
| 194 | assert start == 2 |
| 195 | |
| 196 | def test_returns_none_for_missing_symbol(self) -> None: |
| 197 | src = b"def foo(): pass\n" |
| 198 | result = _find_symbol_span(src, "bar", None) |
| 199 | assert result is None |
| 200 | |
| 201 | def test_returns_none_for_syntax_error(self) -> None: |
| 202 | src = b"def broken(\n" |
| 203 | result = _find_symbol_span(src, "broken", None) |
| 204 | assert result is None |
| 205 | |
| 206 | def test_decorator_lines_included_in_span(self) -> None: |
| 207 | src = textwrap.dedent("""\ |
| 208 | @staticmethod |
| 209 | def foo() -> None: |
| 210 | pass |
| 211 | """).encode() |
| 212 | result = _find_symbol_span(src, "foo", None) |
| 213 | assert result is not None |
| 214 | start, end = result |
| 215 | # Start should be at the decorator line (line 1), not the def (line 2). |
| 216 | assert start == 1 |
| 217 | |
| 218 | def test_multiline_function(self) -> None: |
| 219 | src = textwrap.dedent("""\ |
| 220 | def big( |
| 221 | a: int, |
| 222 | b: int, |
| 223 | ) -> int: |
| 224 | return a + b |
| 225 | """).encode() |
| 226 | result = _find_symbol_span(src, "big", None) |
| 227 | assert result is not None |
| 228 | start, end = result |
| 229 | assert end >= 5 # spans all 5 lines |
| 230 | |
| 231 | def test_variable_assignment(self) -> None: |
| 232 | src = b"CONSTANT = 42\n" |
| 233 | result = _find_symbol_span(src, "CONSTANT", None) |
| 234 | assert result is not None |
| 235 | |
| 236 | def test_annotated_assignment(self) -> None: |
| 237 | src = b"count: int = 0\n" |
| 238 | result = _find_symbol_span(src, "count", None) |
| 239 | assert result is not None |
| 240 | |
| 241 | def test_wrong_parent_class_returns_none(self) -> None: |
| 242 | src = textwrap.dedent("""\ |
| 243 | class Foo: |
| 244 | def bar(self) -> None: |
| 245 | pass |
| 246 | """).encode() |
| 247 | result = _find_symbol_span(src, "bar", "NonExistent") |
| 248 | assert result is None |
| 249 | |
| 250 | |
| 251 | # --------------------------------------------------------------------------- |
| 252 | # Unit — _delete_symbol_lines |
| 253 | # --------------------------------------------------------------------------- |
| 254 | |
| 255 | |
| 256 | class TestDeleteSymbolLines: |
| 257 | def test_removes_middle_function(self) -> None: |
| 258 | lines = [ |
| 259 | "def alpha(): pass\n", |
| 260 | "\n", |
| 261 | "def beta(): pass\n", |
| 262 | "\n", |
| 263 | "def gamma(): pass\n", |
| 264 | ] |
| 265 | result = _delete_symbol_lines(lines, start=3, end=3) |
| 266 | joined = "".join(result) |
| 267 | assert "alpha" in joined |
| 268 | assert "beta" not in joined |
| 269 | assert "gamma" in joined |
| 270 | |
| 271 | def test_removes_first_function(self) -> None: |
| 272 | lines = [ |
| 273 | "def first(): pass\n", |
| 274 | "\n", |
| 275 | "def second(): pass\n", |
| 276 | ] |
| 277 | result = _delete_symbol_lines(lines, start=1, end=1) |
| 278 | joined = "".join(result) |
| 279 | assert "first" not in joined |
| 280 | assert "second" in joined |
| 281 | |
| 282 | def test_removes_last_function(self) -> None: |
| 283 | lines = [ |
| 284 | "def first(): pass\n", |
| 285 | "\n", |
| 286 | "def last(): pass\n", |
| 287 | ] |
| 288 | result = _delete_symbol_lines(lines, start=3, end=3) |
| 289 | joined = "".join(result) |
| 290 | assert "last" not in joined |
| 291 | |
| 292 | def test_normalises_trailing_blanks(self) -> None: |
| 293 | lines = [ |
| 294 | "def foo(): pass\n", |
| 295 | "\n", |
| 296 | "\n", |
| 297 | "def bar(): pass\n", |
| 298 | ] |
| 299 | result = _delete_symbol_lines(lines, start=1, end=1) |
| 300 | # Trailing blanks before the deletion point are stripped. |
| 301 | assert result[0] == "\n" # one separator line |
| 302 | assert "bar" in "".join(result) |
| 303 | |
| 304 | def test_multiline_symbol_removed(self) -> None: |
| 305 | lines = [ |
| 306 | "def first(): pass\n", |
| 307 | "def big(\n", |
| 308 | " x: int,\n", |
| 309 | ") -> int:\n", |
| 310 | " return x\n", |
| 311 | "def last(): pass\n", |
| 312 | ] |
| 313 | result = _delete_symbol_lines(lines, start=2, end=5) |
| 314 | joined = "".join(result) |
| 315 | assert "big" not in joined |
| 316 | assert "first" in joined |
| 317 | assert "last" in joined |
| 318 | |
| 319 | |
| 320 | # --------------------------------------------------------------------------- |
| 321 | # Unit — _analyse_file |
| 322 | # --------------------------------------------------------------------------- |
| 323 | |
| 324 | |
| 325 | _MAX_BYTES = 512 * 1024 # 512 KB default |
| 326 | |
| 327 | |
| 328 | class TestAnalyseFile: |
| 329 | def test_skips_file_over_limit(self) -> None: |
| 330 | raw = b"x" * (_MAX_BYTES + 1) |
| 331 | result = _analyse_file("big.py", raw, None, _MAX_BYTES) |
| 332 | assert result.skipped is True |
| 333 | |
| 334 | def test_non_semantic_suffix_skipped(self) -> None: |
| 335 | # .log is not in SEMANTIC_EXTENSIONS — no symbols extracted. |
| 336 | raw = b"just a log entry" |
| 337 | result = _analyse_file("notes.log", raw, None, _MAX_BYTES) |
| 338 | assert result.symbol_tree == {} |
| 339 | |
| 340 | def test_python_symbols_extracted(self) -> None: |
| 341 | raw = b"def foo(): pass\ndef bar(): pass\n" |
| 342 | result = _analyse_file("mod.py", raw, None, _MAX_BYTES) |
| 343 | assert any("foo" in addr for addr in result.symbol_tree) |
| 344 | assert any("bar" in addr for addr in result.symbol_tree) |
| 345 | |
| 346 | def test_python_ref_names_collected(self) -> None: |
| 347 | raw = b"x = validate_amount(10)\n" |
| 348 | result = _analyse_file("billing.py", raw, None, _MAX_BYTES) |
| 349 | assert "validate_amount" in result.ref_names |
| 350 | |
| 351 | def test_python_import_names_collected(self) -> None: |
| 352 | raw = b"from muse.core import store\nimport os\n" |
| 353 | result = _analyse_file("mod.py", raw, None, _MAX_BYTES) |
| 354 | assert "muse.core" in result.imported_names or "os" in result.imported_names |
| 355 | |
| 356 | def test_kind_filter_applied(self) -> None: |
| 357 | raw = textwrap.dedent("""\ |
| 358 | class Invoice: |
| 359 | pass |
| 360 | |
| 361 | def validate(): pass |
| 362 | """).encode() |
| 363 | result = _analyse_file("mod.py", raw, "function", _MAX_BYTES) |
| 364 | for addr in result.symbol_tree: |
| 365 | assert "Invoice" not in addr |
| 366 | |
| 367 | def test_syntax_error_returns_partial(self) -> None: |
| 368 | raw = b"def broken(\n" |
| 369 | result = _analyse_file("broken.py", raw, None, _MAX_BYTES) |
| 370 | # Should return without raising; symbol_tree may be empty. |
| 371 | assert result.error is not None or result.symbol_tree == {} |
| 372 | |
| 373 | def test_file_path_stored(self) -> None: |
| 374 | raw = b"x = 1\n" |
| 375 | result = _analyse_file("some/path.py", raw, None, _MAX_BYTES) |
| 376 | assert result.file_path == "some/path.py" |
| 377 | |
| 378 | def test_attribute_access_in_refs(self) -> None: |
| 379 | # Attribute accesses like "obj.method" should add "method" to ref_names. |
| 380 | raw = b"result = invoice.compute_total(items)\n" |
| 381 | result = _analyse_file("billing.py", raw, None, _MAX_BYTES) |
| 382 | assert "compute_total" in result.ref_names |
| 383 | |
| 384 | |
| 385 | # --------------------------------------------------------------------------- |
| 386 | # Unit — _is_test_file |
| 387 | # --------------------------------------------------------------------------- |
| 388 | |
| 389 | |
| 390 | class TestIsTestFile: |
| 391 | def test_test_prefix(self) -> None: |
| 392 | assert _is_test_file("tests/test_billing.py") |
| 393 | |
| 394 | def test_test_in_path(self) -> None: |
| 395 | assert _is_test_file("src/test_utils.py") |
| 396 | |
| 397 | def test_spec_in_path(self) -> None: |
| 398 | assert _is_test_file("spec/billing_spec.py") |
| 399 | |
| 400 | def test_normal_file(self) -> None: |
| 401 | assert not _is_test_file("src/billing.py") |
| 402 | |
| 403 | def test_deep_test_path(self) -> None: |
| 404 | assert _is_test_file("a/b/c/test_something.py") |
| 405 | |
| 406 | |
| 407 | # --------------------------------------------------------------------------- |
| 408 | # Unit — _DeadCandidate |
| 409 | # --------------------------------------------------------------------------- |
| 410 | |
| 411 | |
| 412 | class TestDeadCandidateUnit: |
| 413 | def _make( |
| 414 | self, |
| 415 | address: str = "utils.py::orphaned_helper", |
| 416 | kind: str = "function", |
| 417 | file_path: str = "utils.py", |
| 418 | referenced: bool = False, |
| 419 | module_imported: bool = False, |
| 420 | ) -> "_DeadCandidate": |
| 421 | candidate = _DeadCandidate.__new__(_DeadCandidate) |
| 422 | candidate.address = address |
| 423 | candidate.kind = kind |
| 424 | candidate.file_path = file_path |
| 425 | candidate.referenced = referenced |
| 426 | candidate.module_imported = module_imported |
| 427 | return candidate |
| 428 | |
| 429 | def test_high_confidence_not_referenced_not_imported(self) -> None: |
| 430 | c = self._make(referenced=False, module_imported=False) |
| 431 | assert c.confidence == "high" |
| 432 | |
| 433 | def test_medium_confidence_module_imported(self) -> None: |
| 434 | c = self._make(referenced=False, module_imported=True) |
| 435 | assert c.confidence == "medium" |
| 436 | |
| 437 | def test_to_dict_has_required_keys(self) -> None: |
| 438 | c = self._make() |
| 439 | d = c.to_dict() |
| 440 | for key in ("address", "path", "kind", "confidence", "reason"): |
| 441 | assert key in d, f"missing key {key!r}" |
| 442 | |
| 443 | def test_to_dict_confidence_consistent(self) -> None: |
| 444 | c = self._make(referenced=False, module_imported=False) |
| 445 | assert c.to_dict()["confidence"] == "high" |
| 446 | |
| 447 | |
| 448 | # --------------------------------------------------------------------------- |
| 449 | # Integration — extends mega-suite baseline |
| 450 | # --------------------------------------------------------------------------- |
| 451 | |
| 452 | |
| 453 | class TestDeadIntegration: |
| 454 | def test_json_schema(self, repo: pathlib.Path) -> None: |
| 455 | result = runner.invoke(cli, ["code", "dead", "--json"]) |
| 456 | assert result.exit_code == 0, result.output |
| 457 | data = json.loads(result.output) |
| 458 | for key in ("results", "high_confidence_count", "total_files_scanned"): |
| 459 | assert key in data |
| 460 | |
| 461 | def test_high_confidence_only_filters(self, repo: pathlib.Path) -> None: |
| 462 | result = runner.invoke(cli, ["code", "dead", "--high-confidence-only", "--json"]) |
| 463 | assert result.exit_code == 0 |
| 464 | data = json.loads(result.output) |
| 465 | for c in data["results"]: |
| 466 | assert c["confidence"] == "high" |
| 467 | |
| 468 | def test_count_is_integer(self, repo: pathlib.Path) -> None: |
| 469 | result = runner.invoke(cli, ["code", "dead", "--count"]) |
| 470 | assert result.exit_code == 0 |
| 471 | assert result.output.strip().isdigit() |
| 472 | |
| 473 | def test_kind_filter_function(self, repo: pathlib.Path) -> None: |
| 474 | result = runner.invoke(cli, ["code", "dead", "--kind", "function", "--json"]) |
| 475 | assert result.exit_code == 0 |
| 476 | data = json.loads(result.output) |
| 477 | for c in data["results"]: |
| 478 | assert c["kind"] == "function" |
| 479 | |
| 480 | def test_exclude_private_removes_underscore_names( |
| 481 | self, tmp_path: pathlib.Path, monkeypatch: pytest.MonkeyPatch |
| 482 | ) -> None: |
| 483 | monkeypatch.chdir(tmp_path) |
| 484 | monkeypatch.setenv("MUSE_REPO_ROOT", str(tmp_path)) |
| 485 | runner.invoke(cli, ["init", "--domain", "code"]) |
| 486 | (tmp_path / "mod.py").write_text( |
| 487 | "def _private(): pass\ndef public(): pass\n" |
| 488 | ) |
| 489 | runner.invoke(cli, ["commit", "-m", "mod"]) |
| 490 | result = runner.invoke(cli, ["code", "dead", "--exclude-private", "--json"]) |
| 491 | assert result.exit_code == 0 |
| 492 | data = json.loads(result.output) |
| 493 | addresses = [c["address"] for c in data["results"]] |
| 494 | assert not any("_private" in addr for addr in addresses) |
| 495 | |
| 496 | def test_workers_capped(self, repo: pathlib.Path) -> None: |
| 497 | result = runner.invoke(cli, ["code", "dead", "--workers", "512", "--count"]) |
| 498 | assert result.exit_code == 0 |
| 499 | |
| 500 | def test_compare_head_schema(self, repo: pathlib.Path) -> None: |
| 501 | result = runner.invoke(cli, ["code", "dead", "--compare", "HEAD", "--json"]) |
| 502 | assert result.exit_code == 0 |
| 503 | data = json.loads(result.output) |
| 504 | for key in ("compare_commit_id", "new_dead", "recovered", "net_change"): |
| 505 | assert key in data |
| 506 | |
| 507 | def test_delete_and_compare_mutually_exclusive(self, repo: pathlib.Path) -> None: |
| 508 | result = runner.invoke(cli, ["code", "dead", "--delete", "--compare", "HEAD"]) |
| 509 | assert result.exit_code == 1 |
| 510 | |
| 511 | def test_save_allowlist_creates_json_file( |
| 512 | self, repo: pathlib.Path, tmp_path: pathlib.Path |
| 513 | ) -> None: |
| 514 | allow_file = tmp_path / "allow.json" |
| 515 | result = runner.invoke(cli, ["code", "dead", "--save-allowlist", str(allow_file)]) |
| 516 | assert result.exit_code == 0 |
| 517 | if allow_file.exists(): |
| 518 | data = json.loads(allow_file.read_text()) |
| 519 | assert isinstance(data, list) |
| 520 | |
| 521 | def test_allowlist_excludes_addresses( |
| 522 | self, repo: pathlib.Path, tmp_path: pathlib.Path |
| 523 | ) -> None: |
| 524 | allow_file = tmp_path / "allow.json" |
| 525 | allow_file.write_text('["utils.py::orphaned_helper"]') |
| 526 | result = runner.invoke(cli, [ |
| 527 | "code", "dead", "--allowlist", str(allow_file), "--json", |
| 528 | ]) |
| 529 | assert result.exit_code == 0 |
| 530 | data = json.loads(result.output) |
| 531 | names = [c["address"] for c in data["results"]] |
| 532 | assert "utils.py::orphaned_helper" not in names |
| 533 | |
| 534 | def test_language_filter_python(self, repo: pathlib.Path) -> None: |
| 535 | result = runner.invoke(cli, ["code", "dead", "--language", "Python", "--json"]) |
| 536 | assert result.exit_code == 0 |
| 537 | data = json.loads(result.output) |
| 538 | for c in data["results"]: |
| 539 | assert c["path"].endswith(".py") |
| 540 | |
| 541 | def test_deleted_working_tree_file_excluded_from_dead_scan( |
| 542 | self, tmp_path: pathlib.Path, monkeypatch: pytest.MonkeyPatch |
| 543 | ) -> None: |
| 544 | """Symbols in a file deleted from the working tree must not appear as dead. |
| 545 | |
| 546 | Regression test for the _load_file_bytes fallback bug: when from_disk=True |
| 547 | and a file was deleted, the old code fell back to reading the committed |
| 548 | version from the object store, causing symbols in deleted files to be |
| 549 | reported as dead code. The correct behaviour is to exclude deleted files |
| 550 | entirely — a deleted file has no symbols, so none of its symbols can be dead. |
| 551 | """ |
| 552 | monkeypatch.chdir(tmp_path) |
| 553 | monkeypatch.setenv("MUSE_REPO_ROOT", str(tmp_path)) |
| 554 | runner.invoke(cli, ["init", "--domain", "code"]) |
| 555 | |
| 556 | # gone.py — contains a function that has no callers (would appear dead). |
| 557 | (tmp_path / "gone.py").write_text("def vanishing_fn() -> None:\n pass\n") |
| 558 | runner.invoke(cli, ["commit", "-m", "add gone.py"]) |
| 559 | |
| 560 | # Verify it IS detected as dead before deletion. |
| 561 | result_before = runner.invoke(cli, ["code", "dead", "--json"]) |
| 562 | assert result_before.exit_code == 0 |
| 563 | addrs_before = [c["address"] for c in json.loads(result_before.output)["results"]] |
| 564 | assert any("vanishing_fn" in a for a in addrs_before), ( |
| 565 | "vanishing_fn should be reported as dead before the file is deleted" |
| 566 | ) |
| 567 | |
| 568 | # Delete the file without committing. |
| 569 | (tmp_path / "gone.py").unlink() |
| 570 | |
| 571 | # After deletion, vanishing_fn must NOT appear in the working-tree dead scan. |
| 572 | result_after = runner.invoke(cli, ["code", "dead", "--json"]) |
| 573 | assert result_after.exit_code == 0 |
| 574 | addrs_after = [c["address"] for c in json.loads(result_after.output)["results"]] |
| 575 | assert not any("vanishing_fn" in a for a in addrs_after), ( |
| 576 | "vanishing_fn must not be reported as dead after its file is deleted " |
| 577 | "from the working tree — deleted files have no symbols" |
| 578 | ) |
| 579 | |
| 580 | |
| 581 | # --------------------------------------------------------------------------- |
| 582 | # Security |
| 583 | # --------------------------------------------------------------------------- |
| 584 | |
| 585 | |
| 586 | class TestDeadSecurity: |
| 587 | def test_requires_repo( |
| 588 | self, tmp_path: pathlib.Path, monkeypatch: pytest.MonkeyPatch |
| 589 | ) -> None: |
| 590 | monkeypatch.chdir(tmp_path) |
| 591 | monkeypatch.delenv("MUSE_REPO_ROOT", raising=False) |
| 592 | result = runner.invoke(cli, ["code", "dead"]) |
| 593 | assert result.exit_code != 0 |
| 594 | |
| 595 | def test_delete_without_yes_does_not_write( |
| 596 | self, repo: pathlib.Path |
| 597 | ) -> None: |
| 598 | # --delete without --yes should not silently delete anything. |
| 599 | # On a fresh repo with 0 dead candidates it exits 0 safely. |
| 600 | result = runner.invoke(cli, ["code", "dead", "--delete", "--yes"]) |
| 601 | assert result.exit_code in (0, 1) |
| 602 | |
| 603 | |
| 604 | # --------------------------------------------------------------------------- |
| 605 | # Stress — large codebase performance |
| 606 | # --------------------------------------------------------------------------- |
| 607 | |
| 608 | |
| 609 | class TestDeadStress: |
| 610 | @pytest.fixture |
| 611 | def large_repo( |
| 612 | self, tmp_path: pathlib.Path, monkeypatch: pytest.MonkeyPatch |
| 613 | ) -> pathlib.Path: |
| 614 | """50 Python files each with 4 functions, creating ~200 total symbols.""" |
| 615 | monkeypatch.chdir(tmp_path) |
| 616 | monkeypatch.setenv("MUSE_REPO_ROOT", str(tmp_path)) |
| 617 | runner.invoke(cli, ["init", "--domain", "code"]) |
| 618 | |
| 619 | for file_idx in range(50): |
| 620 | content = textwrap.dedent(f"""\ |
| 621 | def do_work_{file_idx}(x: int) -> int: |
| 622 | return x + {file_idx} |
| 623 | |
| 624 | def helper_{file_idx}(x: int) -> int: |
| 625 | return x * {file_idx} |
| 626 | |
| 627 | def unused_a_{file_idx}(x: int) -> int: |
| 628 | return x |
| 629 | |
| 630 | def unused_b_{file_idx}(x: int) -> int: |
| 631 | return x |
| 632 | """) |
| 633 | (tmp_path / f"module_{file_idx:03d}.py").write_text(content) |
| 634 | |
| 635 | r = runner.invoke(cli, ["commit", "-m", "large codebase"]) |
| 636 | assert r.exit_code == 0, r.output |
| 637 | return tmp_path |
| 638 | |
| 639 | def test_dead_on_large_codebase_under_10s(self, large_repo: pathlib.Path) -> None: |
| 640 | start = time.monotonic() |
| 641 | result = runner.invoke(cli, ["code", "dead", "--json"]) |
| 642 | elapsed = time.monotonic() - start |
| 643 | assert result.exit_code == 0, result.output |
| 644 | assert elapsed < 10.0, f"dead on 200 symbols took {elapsed:.2f}s" |
| 645 | |
| 646 | def test_dead_count_on_large_codebase_under_10s(self, large_repo: pathlib.Path) -> None: |
| 647 | start = time.monotonic() |
| 648 | result = runner.invoke(cli, ["code", "dead", "--count"]) |
| 649 | elapsed = time.monotonic() - start |
| 650 | assert result.exit_code == 0 |
| 651 | assert elapsed < 10.0 |
| 652 | assert result.output.strip().isdigit() |
| 653 | |
| 654 | def test_dead_json_schema_valid_on_large_codebase(self, large_repo: pathlib.Path) -> None: |
| 655 | result = runner.invoke(cli, ["code", "dead", "--json"]) |
| 656 | assert result.exit_code == 0 |
| 657 | data = json.loads(result.output) |
| 658 | assert "results" in data |
| 659 | assert "total_files_scanned" in data |
| 660 | assert data["total_files_scanned"] >= 50 |
| 661 | |
| 662 | def test_dead_kind_filter_on_large_codebase(self, large_repo: pathlib.Path) -> None: |
| 663 | start = time.monotonic() |
| 664 | result = runner.invoke(cli, ["code", "dead", "--kind", "function", "--json"]) |
| 665 | elapsed = time.monotonic() - start |
| 666 | assert result.exit_code == 0 |
| 667 | assert elapsed < 10.0 |
| 668 | data = json.loads(result.output) |
| 669 | for c in data["results"]: |
| 670 | assert c["kind"] == "function" |
| 671 | |
| 672 | def test_analyse_file_200_symbols_under_1s(self) -> None: |
| 673 | """Direct unit stress: analyse a 200-function file in < 1 s.""" |
| 674 | lines: list[str] = [] |
| 675 | for i in range(200): |
| 676 | lines.append(f"def sym_{i:04d}(x: int) -> int:") |
| 677 | lines.append(f" return x + {i}") |
| 678 | lines.append("") |
| 679 | raw = "\n".join(lines).encode() |
| 680 | start = time.monotonic() |
| 681 | result = _analyse_file("big.py", raw, None, _MAX_BYTES * 10) |
| 682 | elapsed = time.monotonic() - start |
| 683 | assert elapsed < 1.0, f"_analyse_file on 200 symbols took {elapsed:.3f}s" |
| 684 | assert len(result.symbol_tree) >= 200 |
| 685 | |
| 686 | |
| 687 | # --------------------------------------------------------------------------- |
| 688 | # TestRegisterFlags — --json / -j normalized at argparse level |
| 689 | # --------------------------------------------------------------------------- |
| 690 | |
| 691 | |
| 692 | class TestRegisterFlags: |
| 693 | """register() must expose --json with -j shorthand and dest=json_out.""" |
| 694 | |
| 695 | def _make_parser(self) -> "argparse.ArgumentParser": |
| 696 | import argparse as ap |
| 697 | from muse.cli.commands.dead import register |
| 698 | root = ap.ArgumentParser() |
| 699 | subs = root.add_subparsers() |
| 700 | register(subs) |
| 701 | return root |
| 702 | |
| 703 | def test_json_out_default_false(self) -> None: |
| 704 | p = self._make_parser() |
| 705 | ns = p.parse_args(['code', 'dead']) |
| 706 | assert ns.json_out is False |
| 707 | |
| 708 | def test_json_out_true_with_json_flag(self) -> None: |
| 709 | p = self._make_parser() |
| 710 | ns = p.parse_args(['code', 'dead', '--json']) |
| 711 | assert ns.json_out is True |
| 712 | |
| 713 | def test_json_out_true_with_j_flag(self) -> None: |
| 714 | p = self._make_parser() |
| 715 | ns = p.parse_args(['code', 'dead', '-j']) |
| 716 | assert ns.json_out is True |
File History
1 commit
sha256:d11a87833d5fad6059b7662844bf5448a8911a17cce7a51811f71ad394f248eb
bump to v0.2.0rc13
Human
patch
6 days ago