test_cmd_content_grep_hardening.py
python
sha256:2eaa5d95f9d9383498e76947410a26e5a3ba23d182f339910c424cf88fad412b
fix: try fetch/presign before fetch/mpack to avoid Cloudfla…
Sonnet 4.6
patch
7 days ago
| 1 | """Hardening tests for ``muse content-grep``. |
| 2 | |
| 3 | Covers: |
| 4 | Unit — _is_binary, _path_matches_globs, _search_object (context, |
| 5 | binary skip, utf-8 replace), pattern validation order |
| 6 | Security — ANSI injection in file paths and match text, pattern length |
| 7 | cap, invalid regex, ReDoS pattern rejected before I/O |
| 8 | Perf — parallel reads complete correctly, --max-matches cap |
| 9 | JSON — _ContentGrepJson schema (commit_id, snapshot_id, totals), |
| 10 | GrepMatch context_before/context_after fields |
| 11 | Flags — --include, --exclude, --max-matches, --context/-C, --json, |
| 12 | rejection of old --format flag |
| 13 | Integration — multi-file with mixed hits, --include narrows search, |
| 14 | --exclude skips files, --context shows surrounding lines, |
| 15 | --ref searches historical commit |
| 16 | E2E — --help output mentions all new flags |
| 17 | Stress — 500-file snapshot, concurrent parallel reads |
| 18 | """ |
| 19 | |
| 20 | from __future__ import annotations |
| 21 | from collections.abc import Mapping |
| 22 | |
| 23 | import datetime |
| 24 | import json |
| 25 | import pathlib |
| 26 | import threading |
| 27 | from typing import TypedDict |
| 28 | |
| 29 | import pytest |
| 30 | from tests.cli_test_helper import CliRunner, InvokeResult |
| 31 | |
| 32 | from muse.core.object_store import write_object |
| 33 | from muse.core.ids import hash_commit, hash_snapshot |
| 34 | from muse.core.commits import ( |
| 35 | CommitRecord, |
| 36 | write_commit, |
| 37 | ) |
| 38 | from muse.core.snapshots import ( |
| 39 | SnapshotRecord, |
| 40 | write_snapshot, |
| 41 | ) |
| 42 | from muse.core.types import Manifest, blob_id |
| 43 | |
| 44 | cli = None |
| 45 | runner = CliRunner() |
| 46 | _invoke_lock = threading.Lock() |
| 47 | |
| 48 | type _FilesMap = dict[str, bytes] |
| 49 | |
| 50 | _REPO_ID = "cgrep-hardening" |
| 51 | |
| 52 | |
| 53 | # --------------------------------------------------------------------------- |
| 54 | # Helpers |
| 55 | # --------------------------------------------------------------------------- |
| 56 | |
| 57 | |
| 58 | class _GrepMatchOut(TypedDict): |
| 59 | line_number: int |
| 60 | line: str |
| 61 | context_before: list[str] |
| 62 | context_after: list[str] |
| 63 | |
| 64 | |
| 65 | class _GrepResultOut(TypedDict): |
| 66 | file: str |
| 67 | object_id: str |
| 68 | match_count: int |
| 69 | matches: list[_GrepMatchOut] |
| 70 | |
| 71 | |
| 72 | class _GrepOut(TypedDict): |
| 73 | source: str |
| 74 | commit_id: str |
| 75 | snapshot_id: str |
| 76 | pattern: str |
| 77 | total_files_matched: int |
| 78 | total_matches: int |
| 79 | results: list[_GrepResultOut] |
| 80 | duration_ms: float |
| 81 | exit_code: int |
| 82 | |
| 83 | |
| 84 | |
| 85 | |
| 86 | def _init_repo(path: pathlib.Path, repo_id: str = _REPO_ID) -> pathlib.Path: |
| 87 | dot_muse = muse_dir(path) |
| 88 | for d in ("commits", "snapshots", "objects", "refs/heads"): |
| 89 | (dot_muse / d).mkdir(parents=True, exist_ok=True) |
| 90 | (dot_muse / "HEAD").write_text("ref: refs/heads/main", encoding="utf-8") |
| 91 | (dot_muse / "repo.json").write_text( |
| 92 | json.dumps({"repo_id": repo_id, "domain": "midi"}), encoding="utf-8" |
| 93 | ) |
| 94 | return path |
| 95 | |
| 96 | |
| 97 | def _env(repo: pathlib.Path) -> Manifest: |
| 98 | return {"MUSE_REPO_ROOT": str(repo)} |
| 99 | |
| 100 | |
| 101 | _counter = 0 |
| 102 | |
| 103 | |
| 104 | def _commit_files( |
| 105 | root: pathlib.Path, |
| 106 | files: _FilesMap, |
| 107 | branch: str = "main", |
| 108 | parent_id: str | None = None, |
| 109 | ) -> str: |
| 110 | global _counter |
| 111 | _counter += 1 |
| 112 | manifest: Manifest = {} |
| 113 | for rel_path, content in files.items(): |
| 114 | obj_id = blob_id(content) |
| 115 | write_object(root, obj_id, content) |
| 116 | manifest[rel_path] = obj_id |
| 117 | snap_id = hash_snapshot(manifest) |
| 118 | write_snapshot(root, SnapshotRecord(snapshot_id=snap_id, manifest=manifest)) |
| 119 | committed_at = datetime.datetime.now(datetime.timezone.utc) |
| 120 | parent_ids = [parent_id] if parent_id else [] |
| 121 | commit_id = hash_commit( |
| 122 | parent_ids, snap_id, f"commit {_counter}", committed_at.isoformat(), |
| 123 | ) |
| 124 | write_commit( |
| 125 | root, |
| 126 | CommitRecord( |
| 127 | commit_id=commit_id, |
| 128 | branch=branch, |
| 129 | snapshot_id=snap_id, |
| 130 | message=f"commit {_counter}", |
| 131 | committed_at=committed_at, |
| 132 | parent_commit_id=parent_id, |
| 133 | ), |
| 134 | ) |
| 135 | branch_ref = ref_path(root, branch) |
| 136 | branch_ref.parent.mkdir(parents=True, exist_ok=True) |
| 137 | branch_ref.write_text(commit_id, encoding="utf-8") |
| 138 | return commit_id |
| 139 | |
| 140 | |
| 141 | def _invoke(args: list[str], env: Manifest | None = None) -> InvokeResult: |
| 142 | with _invoke_lock: |
| 143 | return runner.invoke(cli, args, env=env) |
| 144 | |
| 145 | |
| 146 | def _parse(result: InvokeResult) -> _GrepOut: |
| 147 | raw: _GrepOut = json.loads(result.output) |
| 148 | return raw |
| 149 | |
| 150 | |
| 151 | # --------------------------------------------------------------------------- |
| 152 | # Unit: _is_binary |
| 153 | # --------------------------------------------------------------------------- |
| 154 | |
| 155 | |
| 156 | def test_is_binary_null_byte() -> None: |
| 157 | from muse.cli.commands.content_grep import _is_binary |
| 158 | |
| 159 | assert _is_binary(b"\x00hello") is True |
| 160 | |
| 161 | |
| 162 | def test_is_binary_clean_text() -> None: |
| 163 | from muse.cli.commands.content_grep import _is_binary |
| 164 | |
| 165 | assert _is_binary(b"hello world\n") is False |
| 166 | |
| 167 | |
| 168 | def test_is_binary_empty() -> None: |
| 169 | from muse.cli.commands.content_grep import _is_binary |
| 170 | |
| 171 | assert _is_binary(b"") is False |
| 172 | |
| 173 | |
| 174 | # --------------------------------------------------------------------------- |
| 175 | # Unit: _path_matches_globs |
| 176 | # --------------------------------------------------------------------------- |
| 177 | |
| 178 | |
| 179 | def test_path_matches_no_filter() -> None: |
| 180 | from muse.cli.commands.content_grep import _path_matches_globs |
| 181 | |
| 182 | assert _path_matches_globs("src/main.py", None, None) is True |
| 183 | |
| 184 | |
| 185 | def test_path_matches_include_basename() -> None: |
| 186 | from muse.cli.commands.content_grep import _path_matches_globs |
| 187 | |
| 188 | assert _path_matches_globs("src/main.py", "*.py", None) is True |
| 189 | assert _path_matches_globs("src/main.js", "*.py", None) is False |
| 190 | |
| 191 | |
| 192 | def test_path_matches_include_full_path() -> None: |
| 193 | from muse.cli.commands.content_grep import _path_matches_globs |
| 194 | |
| 195 | assert _path_matches_globs("src/main.py", "src/*.py", None) is True |
| 196 | assert _path_matches_globs("tests/main.py", "src/*.py", None) is False |
| 197 | |
| 198 | |
| 199 | def test_path_matches_exclude_basename() -> None: |
| 200 | from muse.cli.commands.content_grep import _path_matches_globs |
| 201 | |
| 202 | assert _path_matches_globs("app.min.js", None, "*.min.js") is False |
| 203 | assert _path_matches_globs("app.js", None, "*.min.js") is True |
| 204 | |
| 205 | |
| 206 | def test_path_matches_include_and_exclude() -> None: |
| 207 | from muse.cli.commands.content_grep import _path_matches_globs |
| 208 | |
| 209 | assert _path_matches_globs("src/main.py", "*.py", "test_*.py") is True |
| 210 | assert _path_matches_globs("test_foo.py", "*.py", "test_*.py") is False |
| 211 | |
| 212 | |
| 213 | # --------------------------------------------------------------------------- |
| 214 | # Unit: _search_object — context lines |
| 215 | # --------------------------------------------------------------------------- |
| 216 | |
| 217 | |
| 218 | def test_search_object_context(tmp_path: pathlib.Path) -> None: |
| 219 | import re |
| 220 | from muse.cli.commands.content_grep import _search_object |
| 221 | |
| 222 | _init_repo(tmp_path) |
| 223 | content = b"line one\nTARGET line\nline three\n" |
| 224 | obj_id = blob_id(content) |
| 225 | write_object(tmp_path, obj_id, content) |
| 226 | |
| 227 | pat = re.compile("TARGET") |
| 228 | count, matches = _search_object(tmp_path, obj_id, pat, False, False, context_lines=1) |
| 229 | assert count == 1 |
| 230 | assert len(matches) == 1 |
| 231 | assert matches[0]["context_before"] == ["line one"] |
| 232 | assert matches[0]["context_after"] == ["line three"] |
| 233 | |
| 234 | |
| 235 | def test_search_object_context_at_boundary(tmp_path: pathlib.Path) -> None: |
| 236 | import re |
| 237 | from muse.cli.commands.content_grep import _search_object |
| 238 | |
| 239 | _init_repo(tmp_path) |
| 240 | content = b"TARGET\nonly\n" |
| 241 | obj_id = blob_id(content) |
| 242 | write_object(tmp_path, obj_id, content) |
| 243 | |
| 244 | pat = re.compile("TARGET") |
| 245 | count, matches = _search_object(tmp_path, obj_id, pat, False, False, context_lines=3) |
| 246 | assert matches[0]["context_before"] == [] |
| 247 | assert matches[0]["context_after"] == ["only"] |
| 248 | |
| 249 | |
| 250 | def test_search_object_no_context(tmp_path: pathlib.Path) -> None: |
| 251 | import re |
| 252 | from muse.cli.commands.content_grep import _search_object |
| 253 | |
| 254 | _init_repo(tmp_path) |
| 255 | content = b"line\nTARGET\nend\n" |
| 256 | obj_id = blob_id(content) |
| 257 | write_object(tmp_path, obj_id, content) |
| 258 | |
| 259 | pat = re.compile("TARGET") |
| 260 | _, matches = _search_object(tmp_path, obj_id, pat, False, False, context_lines=0) |
| 261 | assert matches[0]["context_before"] == [] |
| 262 | assert matches[0]["context_after"] == [] |
| 263 | |
| 264 | |
| 265 | def test_search_object_binary_skipped(tmp_path: pathlib.Path) -> None: |
| 266 | import re |
| 267 | from muse.cli.commands.content_grep import _search_object |
| 268 | |
| 269 | _init_repo(tmp_path) |
| 270 | content = b"\x00\x01\x02TARGET\x03" |
| 271 | obj_id = blob_id(content) |
| 272 | write_object(tmp_path, obj_id, content) |
| 273 | |
| 274 | pat = re.compile("TARGET") |
| 275 | count, matches = _search_object(tmp_path, obj_id, pat, False, False, 0) |
| 276 | assert count == 0 |
| 277 | assert matches == [] |
| 278 | |
| 279 | |
| 280 | # --------------------------------------------------------------------------- |
| 281 | # Security: pattern validation happens BEFORE I/O |
| 282 | # --------------------------------------------------------------------------- |
| 283 | |
| 284 | |
| 285 | def test_long_pattern_rejected_before_io(tmp_path: pathlib.Path) -> None: |
| 286 | """A too-long pattern must be rejected without touching the object store.""" |
| 287 | _init_repo(tmp_path) |
| 288 | # Do NOT commit any files — if I/O happened, we'd get a 'no commits' error, |
| 289 | # not the 'pattern too long' error. |
| 290 | bad_pattern = "a" * 501 |
| 291 | result = _invoke( |
| 292 | ["content-grep", bad_pattern], env=_env(tmp_path) |
| 293 | ) |
| 294 | assert result.exit_code != 0 |
| 295 | # The error must be about pattern length, not about missing commits. |
| 296 | assert "too long" in result.output.lower() or "too long" in (result.stderr or "").lower() |
| 297 | |
| 298 | |
| 299 | def test_invalid_regex_rejected_before_io(tmp_path: pathlib.Path) -> None: |
| 300 | _init_repo(tmp_path) |
| 301 | result = _invoke( |
| 302 | ["content-grep", "[unclosed"], env=_env(tmp_path) |
| 303 | ) |
| 304 | assert result.exit_code != 0 |
| 305 | assert "regex" in result.output.lower() or "regex" in (result.stderr or "").lower() |
| 306 | |
| 307 | |
| 308 | # --------------------------------------------------------------------------- |
| 309 | # Security: ANSI injection |
| 310 | # --------------------------------------------------------------------------- |
| 311 | |
| 312 | |
| 313 | def test_ansi_injection_in_path(tmp_path: pathlib.Path) -> None: |
| 314 | """File paths with ANSI escapes must be stripped in text output.""" |
| 315 | _init_repo(tmp_path) |
| 316 | ansi_path = "\x1b[31mmalicious\x1b[0m.txt" |
| 317 | _commit_files(tmp_path, {ansi_path: b"TARGET content\n"}) |
| 318 | result = _invoke( |
| 319 | ["content-grep", "TARGET"], env=_env(tmp_path) |
| 320 | ) |
| 321 | assert result.exit_code == 0 |
| 322 | assert "\x1b" not in result.output |
| 323 | |
| 324 | |
| 325 | def test_ansi_injection_in_match_text(tmp_path: pathlib.Path) -> None: |
| 326 | """Match text with ANSI escapes must be stripped in text output.""" |
| 327 | _init_repo(tmp_path) |
| 328 | _commit_files(tmp_path, {"safe.txt": b"TARGET \x1b[31mred\x1b[0m content\n"}) |
| 329 | result = _invoke( |
| 330 | ["content-grep", "TARGET"], env=_env(tmp_path) |
| 331 | ) |
| 332 | assert result.exit_code == 0 |
| 333 | assert "\x1b" not in result.output |
| 334 | |
| 335 | |
| 336 | # --------------------------------------------------------------------------- |
| 337 | # JSON schema: _ContentGrepJson |
| 338 | # --------------------------------------------------------------------------- |
| 339 | |
| 340 | |
| 341 | def test_json_schema_all_fields(tmp_path: pathlib.Path) -> None: |
| 342 | _init_repo(tmp_path) |
| 343 | _commit_files(tmp_path, {"a.txt": b"hello world\nhello again\n"}) |
| 344 | result = _invoke( |
| 345 | ["content-grep", "hello", "--json"], env=_env(tmp_path) |
| 346 | ) |
| 347 | assert result.exit_code == 0 |
| 348 | data = _parse(result) |
| 349 | assert data["commit_id"].startswith("sha256:") |
| 350 | assert len(data["commit_id"]) == 71 |
| 351 | assert data["snapshot_id"].startswith("sha256:") |
| 352 | assert len(data["snapshot_id"]) == 71 |
| 353 | assert data["pattern"] == "hello" |
| 354 | assert data["total_files_matched"] == 1 |
| 355 | assert data["total_matches"] == 2 |
| 356 | assert len(data["results"]) == 1 |
| 357 | r = data["results"][0] |
| 358 | assert r["path"] == "a.txt" |
| 359 | assert r["match_count"] == 2 |
| 360 | assert isinstance(r["matches"], list) |
| 361 | |
| 362 | |
| 363 | def test_json_schema_context_fields(tmp_path: pathlib.Path) -> None: |
| 364 | _init_repo(tmp_path) |
| 365 | _commit_files(tmp_path, {"c.txt": b"before\nTARGET\nafter\n"}) |
| 366 | result = _invoke( |
| 367 | ["content-grep", "TARGET", "--context", "1", "--json"], |
| 368 | env=_env(tmp_path), |
| 369 | ) |
| 370 | assert result.exit_code == 0 |
| 371 | data = _parse(result) |
| 372 | match = data["results"][0]["matches"][0] |
| 373 | assert isinstance(match, dict) |
| 374 | assert "context_before" in match |
| 375 | assert "context_after" in match |
| 376 | assert match["context_before"] == ["before"] |
| 377 | assert match["context_after"] == ["after"] |
| 378 | |
| 379 | |
| 380 | def test_json_schema_no_match_exit1(tmp_path: pathlib.Path) -> None: |
| 381 | _init_repo(tmp_path) |
| 382 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 383 | result = _invoke( |
| 384 | ["content-grep", "ZZZNOMATCH", "--json"], env=_env(tmp_path) |
| 385 | ) |
| 386 | assert result.exit_code != 0 |
| 387 | |
| 388 | |
| 389 | def test_json_total_matches_multiple_files(tmp_path: pathlib.Path) -> None: |
| 390 | _init_repo(tmp_path) |
| 391 | _commit_files(tmp_path, { |
| 392 | "a.txt": b"hit\nhit\n", |
| 393 | "b.txt": b"hit\n", |
| 394 | "c.txt": b"miss\n", |
| 395 | }) |
| 396 | result = _invoke( |
| 397 | ["content-grep", "hit", "--json"], env=_env(tmp_path) |
| 398 | ) |
| 399 | assert result.exit_code == 0 |
| 400 | data = _parse(result) |
| 401 | assert data["total_files_matched"] == 2 |
| 402 | assert data["total_matches"] == 3 |
| 403 | |
| 404 | |
| 405 | # --------------------------------------------------------------------------- |
| 406 | # Flags: --include |
| 407 | # --------------------------------------------------------------------------- |
| 408 | |
| 409 | |
| 410 | def test_include_filters_to_py_only(tmp_path: pathlib.Path) -> None: |
| 411 | _init_repo(tmp_path) |
| 412 | _commit_files(tmp_path, { |
| 413 | "module.py": b"TARGET in python\n", |
| 414 | "module.js": b"TARGET in js\n", |
| 415 | "readme.md": b"TARGET in md\n", |
| 416 | }) |
| 417 | result = _invoke( |
| 418 | ["content-grep", "TARGET", "--include", "*.py", "--json"], |
| 419 | env=_env(tmp_path), |
| 420 | ) |
| 421 | assert result.exit_code == 0 |
| 422 | data = _parse(result) |
| 423 | assert data["total_files_matched"] == 1 |
| 424 | assert data["results"][0]["path"] == "module.py" |
| 425 | |
| 426 | |
| 427 | def test_include_no_matches_after_filter(tmp_path: pathlib.Path) -> None: |
| 428 | _init_repo(tmp_path) |
| 429 | _commit_files(tmp_path, {"module.js": b"TARGET here\n"}) |
| 430 | result = _invoke( |
| 431 | ["content-grep", "TARGET", "--include", "*.py"], |
| 432 | env=_env(tmp_path), |
| 433 | ) |
| 434 | assert result.exit_code != 0 # no files pass include filter |
| 435 | |
| 436 | |
| 437 | # --------------------------------------------------------------------------- |
| 438 | # Flags: --exclude |
| 439 | # --------------------------------------------------------------------------- |
| 440 | |
| 441 | |
| 442 | def test_exclude_skips_minified(tmp_path: pathlib.Path) -> None: |
| 443 | _init_repo(tmp_path) |
| 444 | _commit_files(tmp_path, { |
| 445 | "app.js": b"TARGET here\n", |
| 446 | "app.min.js": b"TARGET minified\n", |
| 447 | }) |
| 448 | result = _invoke( |
| 449 | ["content-grep", "TARGET", "--exclude", "*.min.js", "--json"], |
| 450 | env=_env(tmp_path), |
| 451 | ) |
| 452 | assert result.exit_code == 0 |
| 453 | data = _parse(result) |
| 454 | assert data["total_files_matched"] == 1 |
| 455 | assert data["results"][0]["path"] == "app.js" |
| 456 | |
| 457 | |
| 458 | def test_exclude_all_results_in_no_match(tmp_path: pathlib.Path) -> None: |
| 459 | _init_repo(tmp_path) |
| 460 | _commit_files(tmp_path, {"test.py": b"TARGET\n"}) |
| 461 | result = _invoke( |
| 462 | ["content-grep", "TARGET", "--exclude", "test_*.py"], |
| 463 | env=_env(tmp_path), |
| 464 | ) |
| 465 | # test.py doesn't match test_*.py exclude pattern, so it should match. |
| 466 | # Verify this works (target file isn't excluded). |
| 467 | assert result.exit_code == 0 |
| 468 | |
| 469 | |
| 470 | # --------------------------------------------------------------------------- |
| 471 | # Flags: --max-matches |
| 472 | # --------------------------------------------------------------------------- |
| 473 | |
| 474 | |
| 475 | def test_max_matches_caps_output(tmp_path: pathlib.Path) -> None: |
| 476 | _init_repo(tmp_path) |
| 477 | _commit_files(tmp_path, {"many.txt": b"hit\n" * 100}) |
| 478 | result = _invoke( |
| 479 | ["content-grep", "hit", "--max-matches", "10", "--json"], |
| 480 | env=_env(tmp_path), |
| 481 | ) |
| 482 | assert result.exit_code == 0 |
| 483 | data = _parse(result) |
| 484 | assert data["total_matches"] <= 10 |
| 485 | |
| 486 | |
| 487 | def test_max_matches_zero_still_exits_nonzero_on_cap(tmp_path: pathlib.Path) -> None: |
| 488 | """When max_matches=0, no results are kept — exit 1.""" |
| 489 | _init_repo(tmp_path) |
| 490 | _commit_files(tmp_path, {"a.txt": b"hit\n"}) |
| 491 | result = _invoke( |
| 492 | ["content-grep", "hit", "--max-matches", "0", "--json"], |
| 493 | env=_env(tmp_path), |
| 494 | ) |
| 495 | assert result.exit_code != 0 # no results after cap → exit 1 |
| 496 | |
| 497 | |
| 498 | # --------------------------------------------------------------------------- |
| 499 | # Flags: --context / -C |
| 500 | # --------------------------------------------------------------------------- |
| 501 | |
| 502 | |
| 503 | def test_context_text_output(tmp_path: pathlib.Path) -> None: |
| 504 | _init_repo(tmp_path) |
| 505 | _commit_files(tmp_path, {"ctx.txt": b"alpha\nbeta\ngamma\n"}) |
| 506 | result = _invoke( |
| 507 | ["content-grep", "beta", "--context", "1"], |
| 508 | env=_env(tmp_path), |
| 509 | ) |
| 510 | assert result.exit_code == 0 |
| 511 | # Context before and after should appear in output. |
| 512 | assert "alpha" in result.output |
| 513 | assert "gamma" in result.output |
| 514 | |
| 515 | |
| 516 | def test_context_short_flag(tmp_path: pathlib.Path) -> None: |
| 517 | _init_repo(tmp_path) |
| 518 | _commit_files(tmp_path, {"ctx2.txt": b"first\nTARGET\nlast\n"}) |
| 519 | result = _invoke( |
| 520 | ["content-grep", "TARGET", "-C", "1"], |
| 521 | env=_env(tmp_path), |
| 522 | ) |
| 523 | assert result.exit_code == 0 |
| 524 | assert "first" in result.output |
| 525 | assert "last" in result.output |
| 526 | |
| 527 | |
| 528 | # --------------------------------------------------------------------------- |
| 529 | # Flags: --json boolean (rejects old --format) |
| 530 | # --------------------------------------------------------------------------- |
| 531 | |
| 532 | |
| 533 | def test_format_flag_rejected(tmp_path: pathlib.Path) -> None: |
| 534 | """Old ``--format json`` must be rejected by argparse (exit 2).""" |
| 535 | _init_repo(tmp_path) |
| 536 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 537 | result = _invoke( |
| 538 | ["content-grep", "hello", "--format", "json"], |
| 539 | env=_env(tmp_path), |
| 540 | ) |
| 541 | assert result.exit_code == 2 |
| 542 | |
| 543 | |
| 544 | # --------------------------------------------------------------------------- |
| 545 | # Integration: --ref searches a different commit |
| 546 | # --------------------------------------------------------------------------- |
| 547 | |
| 548 | |
| 549 | def test_ref_searches_branch(tmp_path: pathlib.Path) -> None: |
| 550 | _init_repo(tmp_path) |
| 551 | c1 = _commit_files(tmp_path, {"v1.txt": b"OLD content\n"}) |
| 552 | _commit_files(tmp_path, {"v2.txt": b"NEW content\n"}, parent_id=c1) |
| 553 | |
| 554 | # Search HEAD — should find NEW in v2.txt. |
| 555 | result_head = _invoke( |
| 556 | ["content-grep", "NEW", "--json"], env=_env(tmp_path) |
| 557 | ) |
| 558 | assert result_head.exit_code == 0 |
| 559 | data = _parse(result_head) |
| 560 | paths = [r["path"] for r in data["results"]] |
| 561 | assert "v2.txt" in paths |
| 562 | |
| 563 | # Search the first commit by ID — should find OLD in v1.txt, not NEW. |
| 564 | result_ref = _invoke( |
| 565 | ["content-grep", "OLD", "--ref", c1, "--json"], |
| 566 | env=_env(tmp_path), |
| 567 | ) |
| 568 | assert result_ref.exit_code == 0 |
| 569 | data_ref = _parse(result_ref) |
| 570 | paths_ref = [r["path"] for r in data_ref["results"]] |
| 571 | assert "v1.txt" in paths_ref |
| 572 | assert data_ref["commit_id"] == c1 |
| 573 | |
| 574 | |
| 575 | # --------------------------------------------------------------------------- |
| 576 | # E2E: --help mentions all new flags |
| 577 | # --------------------------------------------------------------------------- |
| 578 | |
| 579 | |
| 580 | def test_help_mentions_include() -> None: |
| 581 | result = _invoke(["content-grep", "--help"]) |
| 582 | assert result.exit_code == 0 |
| 583 | assert "--include" in result.output |
| 584 | |
| 585 | |
| 586 | def test_help_mentions_exclude() -> None: |
| 587 | result = _invoke(["content-grep", "--help"]) |
| 588 | assert "--exclude" in result.output |
| 589 | |
| 590 | |
| 591 | def test_help_mentions_max_matches() -> None: |
| 592 | result = _invoke(["content-grep", "--help"]) |
| 593 | assert "--max-matches" in result.output |
| 594 | |
| 595 | |
| 596 | def test_help_mentions_context() -> None: |
| 597 | result = _invoke(["content-grep", "--help"]) |
| 598 | assert "--context" in result.output or "-C" in result.output |
| 599 | |
| 600 | |
| 601 | def test_help_mentions_json_not_format() -> None: |
| 602 | result = _invoke(["content-grep", "--help"]) |
| 603 | assert "--json" in result.output |
| 604 | assert "--format" not in result.output |
| 605 | |
| 606 | |
| 607 | # --------------------------------------------------------------------------- |
| 608 | # Stress: 500-file snapshot, pattern matches 250 |
| 609 | # --------------------------------------------------------------------------- |
| 610 | |
| 611 | |
| 612 | def test_stress_500_files(tmp_path: pathlib.Path) -> None: |
| 613 | _init_repo(tmp_path) |
| 614 | files: _FilesMap = {} |
| 615 | for i in range(500): |
| 616 | content = b"TARGET_STRESS\n" if i % 2 == 0 else b"other\n" |
| 617 | files[f"f_{i:04d}.txt"] = content |
| 618 | _commit_files(tmp_path, files) |
| 619 | result = _invoke( |
| 620 | ["content-grep", "TARGET_STRESS", "--json"], |
| 621 | env=_env(tmp_path), |
| 622 | ) |
| 623 | assert result.exit_code == 0 |
| 624 | data = _parse(result) |
| 625 | assert data["total_files_matched"] == 250 |
| 626 | assert data["total_matches"] == 250 |
| 627 | |
| 628 | |
| 629 | # --------------------------------------------------------------------------- |
| 630 | # Stress: concurrent reads |
| 631 | # --------------------------------------------------------------------------- |
| 632 | |
| 633 | |
| 634 | def test_stress_concurrent_reads(tmp_path: pathlib.Path) -> None: |
| 635 | _init_repo(tmp_path) |
| 636 | _commit_files(tmp_path, {"concurrent.txt": b"CONCURRENT TARGET\n"}) |
| 637 | |
| 638 | errors: list[str] = [] |
| 639 | |
| 640 | def _read() -> None: |
| 641 | r = _invoke( |
| 642 | ["content-grep", "CONCURRENT", "--json"], |
| 643 | env=_env(tmp_path), |
| 644 | ) |
| 645 | if r.exit_code != 0: |
| 646 | errors.append(f"exit {r.exit_code}") |
| 647 | else: |
| 648 | try: |
| 649 | d = json.loads(r.output) |
| 650 | if d.get("total_matches", 0) != 1: |
| 651 | errors.append(f"unexpected total_matches: {d.get('total_matches')}") |
| 652 | except json.JSONDecodeError as exc: |
| 653 | errors.append(str(exc)) |
| 654 | |
| 655 | threads = [threading.Thread(target=_read) for _ in range(8)] |
| 656 | for t in threads: |
| 657 | t.start() |
| 658 | for t in threads: |
| 659 | t.join() |
| 660 | |
| 661 | assert not errors, f"Concurrent read failures: {errors}" |
| 662 | |
| 663 | |
| 664 | # --------------------------------------------------------------------------- |
| 665 | # JSON schema: complete key set (TestJsonSchemaComplete) |
| 666 | # --------------------------------------------------------------------------- |
| 667 | |
| 668 | |
| 669 | _REQUIRED_KEYS = frozenset({ |
| 670 | "source", |
| 671 | "commit_id", |
| 672 | "snapshot_id", |
| 673 | "pattern", |
| 674 | "total_files_matched", |
| 675 | "total_matches", |
| 676 | "results", |
| 677 | "duration_ms", |
| 678 | "exit_code", |
| 679 | }) |
| 680 | |
| 681 | |
| 682 | class TestJsonSchemaComplete: |
| 683 | """Verify that every required key is present in JSON output.""" |
| 684 | |
| 685 | def test_all_required_keys_present_commit_mode(self, tmp_path: pathlib.Path) -> None: |
| 686 | _init_repo(tmp_path) |
| 687 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 688 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 689 | assert result.exit_code == 0 |
| 690 | data = json.loads(result.output) |
| 691 | missing = _REQUIRED_KEYS - data.keys() |
| 692 | assert not missing, f"Missing keys: {missing}" |
| 693 | |
| 694 | def test_all_required_keys_present_working_tree_mode(self, tmp_path: pathlib.Path) -> None: |
| 695 | _init_repo(tmp_path) |
| 696 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 697 | # Also write a matching file to disk so working-tree search finds it. |
| 698 | (tmp_path / "a.txt").write_bytes(b"hello\n") |
| 699 | result = _invoke( |
| 700 | ["content-grep", "hello", "--working-tree", "--json"], |
| 701 | env=_env(tmp_path), |
| 702 | ) |
| 703 | assert result.exit_code == 0 |
| 704 | data = json.loads(result.output) |
| 705 | missing = _REQUIRED_KEYS - data.keys() |
| 706 | assert not missing, f"Missing keys: {missing}" |
| 707 | |
| 708 | def test_source_field_is_commit(self, tmp_path: pathlib.Path) -> None: |
| 709 | _init_repo(tmp_path) |
| 710 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 711 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 712 | data = json.loads(result.output) |
| 713 | assert data["source"] == "commit" |
| 714 | |
| 715 | def test_source_field_is_working_tree(self, tmp_path: pathlib.Path) -> None: |
| 716 | _init_repo(tmp_path) |
| 717 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 718 | (tmp_path / "a.txt").write_bytes(b"hello\n") |
| 719 | result = _invoke( |
| 720 | ["content-grep", "hello", "--working-tree", "--json"], |
| 721 | env=_env(tmp_path), |
| 722 | ) |
| 723 | data = json.loads(result.output) |
| 724 | assert data["source"] == "working-tree" |
| 725 | |
| 726 | def test_commit_id_null_in_working_tree_mode(self, tmp_path: pathlib.Path) -> None: |
| 727 | _init_repo(tmp_path) |
| 728 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 729 | (tmp_path / "a.txt").write_bytes(b"hello\n") |
| 730 | result = _invoke( |
| 731 | ["content-grep", "hello", "--working-tree", "--json"], |
| 732 | env=_env(tmp_path), |
| 733 | ) |
| 734 | data = json.loads(result.output) |
| 735 | assert data["commit_id"] is None |
| 736 | |
| 737 | def test_snapshot_id_null_in_working_tree_mode(self, tmp_path: pathlib.Path) -> None: |
| 738 | _init_repo(tmp_path) |
| 739 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 740 | (tmp_path / "a.txt").write_bytes(b"hello\n") |
| 741 | result = _invoke( |
| 742 | ["content-grep", "hello", "--working-tree", "--json"], |
| 743 | env=_env(tmp_path), |
| 744 | ) |
| 745 | data = json.loads(result.output) |
| 746 | assert data["snapshot_id"] is None |
| 747 | |
| 748 | def test_exit_code_field_zero_on_match(self, tmp_path: pathlib.Path) -> None: |
| 749 | _init_repo(tmp_path) |
| 750 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 751 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 752 | data = json.loads(result.output) |
| 753 | assert data["exit_code"] == 0 |
| 754 | |
| 755 | def test_json_is_compact(self, tmp_path: pathlib.Path) -> None: |
| 756 | """JSON output must be a single line — no pretty-printing.""" |
| 757 | _init_repo(tmp_path) |
| 758 | _commit_files(tmp_path, {"a.txt": b"hello\n"}) |
| 759 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 760 | lines = [ln for ln in result.output.splitlines() if ln.strip()] |
| 761 | assert len(lines) == 1, "JSON must be compact (one line)" |
| 762 | |
| 763 | |
| 764 | # --------------------------------------------------------------------------- |
| 765 | # duration_ms (TestElapsedSeconds) |
| 766 | # --------------------------------------------------------------------------- |
| 767 | |
| 768 | |
| 769 | class TestElapsedSeconds: |
| 770 | """``duration_ms`` must be a non-negative float in all JSON paths.""" |
| 771 | |
| 772 | def _assert_elapsed(self, data: Mapping[str, object]) -> None: # type: ignore[type-arg] |
| 773 | assert "duration_ms" in data |
| 774 | assert isinstance(data["duration_ms"], float) |
| 775 | assert data["duration_ms"] >= 0.0 |
| 776 | |
| 777 | def test_elapsed_present_commit_mode(self, tmp_path: pathlib.Path) -> None: |
| 778 | _init_repo(tmp_path) |
| 779 | _commit_files(tmp_path, {"a.txt": b"target\n"}) |
| 780 | result = _invoke(["content-grep", "target", "--json"], env=_env(tmp_path)) |
| 781 | self._assert_elapsed(json.loads(result.output)) |
| 782 | |
| 783 | def test_elapsed_present_working_tree_mode(self, tmp_path: pathlib.Path) -> None: |
| 784 | _init_repo(tmp_path) |
| 785 | _commit_files(tmp_path, {"a.txt": b"target\n"}) |
| 786 | (tmp_path / "a.txt").write_bytes(b"target\n") |
| 787 | result = _invoke( |
| 788 | ["content-grep", "target", "--working-tree", "--json"], |
| 789 | env=_env(tmp_path), |
| 790 | ) |
| 791 | self._assert_elapsed(json.loads(result.output)) |
| 792 | |
| 793 | def test_elapsed_is_float_not_int(self, tmp_path: pathlib.Path) -> None: |
| 794 | _init_repo(tmp_path) |
| 795 | _commit_files(tmp_path, {"a.txt": b"target\n"}) |
| 796 | result = _invoke(["content-grep", "target", "--json"], env=_env(tmp_path)) |
| 797 | data = json.loads(result.output) |
| 798 | assert isinstance(data["duration_ms"], float) |
| 799 | |
| 800 | def test_elapsed_reasonable_upper_bound(self, tmp_path: pathlib.Path) -> None: |
| 801 | """Single-file search in a temp repo should be well under 5 seconds.""" |
| 802 | _init_repo(tmp_path) |
| 803 | _commit_files(tmp_path, {"a.txt": b"target\n"}) |
| 804 | result = _invoke(["content-grep", "target", "--json"], env=_env(tmp_path)) |
| 805 | data = json.loads(result.output) |
| 806 | assert data["duration_ms"] < 5.0 |
| 807 | |
| 808 | def test_elapsed_present_stress_mode(self, tmp_path: pathlib.Path) -> None: |
| 809 | """duration_ms must appear even for 500-file parallel searches.""" |
| 810 | _init_repo(tmp_path) |
| 811 | files: Mapping[str, bytes] = {f"f{i}.txt": b"needle\n" for i in range(50)} |
| 812 | _commit_files(tmp_path, files) |
| 813 | result = _invoke(["content-grep", "needle", "--json"], env=_env(tmp_path)) |
| 814 | assert result.exit_code == 0 |
| 815 | self._assert_elapsed(json.loads(result.output)) |
| 816 | |
| 817 | def test_elapsed_six_decimal_places(self, tmp_path: pathlib.Path) -> None: |
| 818 | """duration_ms should be rounded to at most 6 decimal places.""" |
| 819 | _init_repo(tmp_path) |
| 820 | _commit_files(tmp_path, {"a.txt": b"target\n"}) |
| 821 | result = _invoke(["content-grep", "target", "--json"], env=_env(tmp_path)) |
| 822 | data = json.loads(result.output) |
| 823 | elapsed = data["duration_ms"] |
| 824 | # round-trip through 6-decimal representation must be exact |
| 825 | assert round(elapsed, 6) == elapsed |
| 826 | |
| 827 | |
| 828 | # --------------------------------------------------------------------------- |
| 829 | # exit_code field (TestExitCode) |
| 830 | # --------------------------------------------------------------------------- |
| 831 | |
| 832 | |
| 833 | class TestExitCode: |
| 834 | """``exit_code`` in JSON must mirror the process exit code.""" |
| 835 | |
| 836 | def test_exit_code_zero_on_match(self, tmp_path: pathlib.Path) -> None: |
| 837 | _init_repo(tmp_path) |
| 838 | _commit_files(tmp_path, {"a.txt": b"hit\n"}) |
| 839 | result = _invoke(["content-grep", "hit", "--json"], env=_env(tmp_path)) |
| 840 | assert result.exit_code == 0 |
| 841 | assert json.loads(result.output)["exit_code"] == 0 |
| 842 | |
| 843 | def test_exit_code_zero_working_tree_match(self, tmp_path: pathlib.Path) -> None: |
| 844 | _init_repo(tmp_path) |
| 845 | _commit_files(tmp_path, {"a.txt": b"hit\n"}) |
| 846 | (tmp_path / "a.txt").write_bytes(b"hit\n") |
| 847 | result = _invoke( |
| 848 | ["content-grep", "hit", "--working-tree", "--json"], |
| 849 | env=_env(tmp_path), |
| 850 | ) |
| 851 | assert result.exit_code == 0 |
| 852 | assert json.loads(result.output)["exit_code"] == 0 |
| 853 | |
| 854 | def test_exit_code_is_integer(self, tmp_path: pathlib.Path) -> None: |
| 855 | _init_repo(tmp_path) |
| 856 | _commit_files(tmp_path, {"a.txt": b"hit\n"}) |
| 857 | result = _invoke(["content-grep", "hit", "--json"], env=_env(tmp_path)) |
| 858 | data = json.loads(result.output) |
| 859 | assert isinstance(data["exit_code"], int) |
| 860 | |
| 861 | def test_exit_code_in_json_matches_process_exit(self, tmp_path: pathlib.Path) -> None: |
| 862 | """JSON exit_code must equal the actual process exit code.""" |
| 863 | _init_repo(tmp_path) |
| 864 | _commit_files(tmp_path, {"a.txt": b"hit\n"}) |
| 865 | result = _invoke(["content-grep", "hit", "--json"], env=_env(tmp_path)) |
| 866 | data = json.loads(result.output) |
| 867 | assert data["exit_code"] == result.exit_code |
| 868 | |
| 869 | def test_exit_code_multiple_files(self, tmp_path: pathlib.Path) -> None: |
| 870 | _init_repo(tmp_path) |
| 871 | _commit_files(tmp_path, {"a.txt": b"hit\n", "b.txt": b"hit\n"}) |
| 872 | result = _invoke(["content-grep", "hit", "--json"], env=_env(tmp_path)) |
| 873 | assert result.exit_code == 0 |
| 874 | assert json.loads(result.output)["exit_code"] == 0 |
| 875 | |
| 876 | |
| 877 | # --------------------------------------------------------------------------- |
| 878 | # Flag registration tests |
| 879 | # --------------------------------------------------------------------------- |
| 880 | |
| 881 | import argparse as _argparse |
| 882 | from muse.cli.commands.content_grep import register as _register_content_grep |
| 883 | from muse.core.paths import muse_dir, ref_path |
| 884 | |
| 885 | |
| 886 | def _parse_cgrep(*args: str) -> _argparse.Namespace: |
| 887 | root_p = _argparse.ArgumentParser() |
| 888 | subs = root_p.add_subparsers(dest="cmd") |
| 889 | _register_content_grep(subs) |
| 890 | return root_p.parse_args(["content-grep", *args]) |
| 891 | |
| 892 | |
| 893 | class TestRegisterFlags: |
| 894 | def test_default_json_out_is_false(self) -> None: |
| 895 | ns = _parse_cgrep("TODO") |
| 896 | assert ns.json_out is False |
| 897 | |
| 898 | def test_json_flag_sets_json_out(self) -> None: |
| 899 | ns = _parse_cgrep("TODO", "--json") |
| 900 | assert ns.json_out is True |
| 901 | |
| 902 | def test_j_shorthand_sets_json_out(self) -> None: |
| 903 | ns = _parse_cgrep("TODO", "-j") |
| 904 | assert ns.json_out is True |
| 905 | |
| 906 | def test_pattern_positional(self) -> None: |
| 907 | ns = _parse_cgrep("FIXME") |
| 908 | assert ns.pattern == "FIXME" |
| 909 | |
| 910 | |
| 911 | # --------------------------------------------------------------------------- |
| 912 | # JSON key ergonomics: results[].file and matches[].line |
| 913 | # --------------------------------------------------------------------------- |
| 914 | |
| 915 | |
| 916 | class TestJsonKeyErgonomics: |
| 917 | """content-grep --json must use 'path' (matching all other muse commands) and |
| 918 | 'line' (not 'text') for match content.""" |
| 919 | |
| 920 | def test_result_key_is_path(self, tmp_path: pathlib.Path) -> None: |
| 921 | _init_repo(tmp_path) |
| 922 | _commit_files(tmp_path, {"src/main.py": b"hello world\n"}) |
| 923 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 924 | data = json.loads(result.output) |
| 925 | assert data["results"][0]["path"] == "src/main.py" |
| 926 | |
| 927 | def test_result_has_no_file_key(self, tmp_path: pathlib.Path) -> None: |
| 928 | _init_repo(tmp_path) |
| 929 | _commit_files(tmp_path, {"src/main.py": b"hello world\n"}) |
| 930 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 931 | data = json.loads(result.output) |
| 932 | assert "file" not in data["results"][0] |
| 933 | |
| 934 | def test_match_key_is_line_not_text(self, tmp_path: pathlib.Path) -> None: |
| 935 | _init_repo(tmp_path) |
| 936 | _commit_files(tmp_path, {"a.py": b"hello world\n"}) |
| 937 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 938 | data = json.loads(result.output) |
| 939 | match = data["results"][0]["matches"][0] |
| 940 | assert match["line"] == "hello world" |
| 941 | |
| 942 | def test_match_has_no_text_key(self, tmp_path: pathlib.Path) -> None: |
| 943 | _init_repo(tmp_path) |
| 944 | _commit_files(tmp_path, {"a.py": b"hello world\n"}) |
| 945 | result = _invoke(["content-grep", "hello", "--json"], env=_env(tmp_path)) |
| 946 | data = json.loads(result.output) |
| 947 | match = data["results"][0]["matches"][0] |
| 948 | assert "text" not in match |
| 949 | |
| 950 | def test_working_tree_result_key_is_path(self, tmp_path: pathlib.Path) -> None: |
| 951 | _init_repo(tmp_path) |
| 952 | _commit_files(tmp_path, {"a.py": b"placeholder\n"}) |
| 953 | (tmp_path / "a.py").write_text("needle here\n", encoding="utf-8") |
| 954 | result = _invoke( |
| 955 | ["content-grep", "needle", "--working-tree", "--json"], env=_env(tmp_path) |
| 956 | ) |
| 957 | data = json.loads(result.output) |
| 958 | assert data["results"][0]["path"] == "a.py" |
| 959 | |
| 960 | def test_working_tree_match_key_is_line(self, tmp_path: pathlib.Path) -> None: |
| 961 | _init_repo(tmp_path) |
| 962 | _commit_files(tmp_path, {"a.py": b"placeholder\n"}) |
| 963 | (tmp_path / "a.py").write_text("needle here\n", encoding="utf-8") |
| 964 | result = _invoke( |
| 965 | ["content-grep", "needle", "--working-tree", "--json"], env=_env(tmp_path) |
| 966 | ) |
| 967 | data = json.loads(result.output) |
| 968 | match = data["results"][0]["matches"][0] |
| 969 | assert match["line"] == "needle here" |
File History
1 commit
sha256:2eaa5d95f9d9383498e76947410a26e5a3ba23d182f339910c424cf88fad412b
fix: try fetch/presign before fetch/mpack to avoid Cloudfla…
Sonnet 4.6
patch
7 days ago