test_musehub_export.py
python
| 1 | """Tests for the Muse Hub export endpoint and musehub_exporter service. |
| 2 | |
| 3 | Covers every acceptance criterion: |
| 4 | - GET /musehub/repos/{repo_id}/export/{ref}?format=midi returns a .mid file |
| 5 | - GET /musehub/repos/{repo_id}/export/{ref}?format=json returns valid JSON |
| 6 | - split_tracks=true bundles artifacts into a ZIP with per-track files |
| 7 | - sections filter restricts artifacts to matching path substrings |
| 8 | - Unknown format string returns 422 Unprocessable Entity |
| 9 | - Unresolvable ref returns 404 |
| 10 | - No matching artifacts for a format returns 404 |
| 11 | |
| 12 | All tests use the shared fixtures from conftest.py. |
| 13 | """ |
| 14 | from __future__ import annotations |
| 15 | |
| 16 | import base64 |
| 17 | import io |
| 18 | import json |
| 19 | import tempfile |
| 20 | import zipfile |
| 21 | from pathlib import Path |
| 22 | from unittest.mock import AsyncMock, patch |
| 23 | |
| 24 | import pytest |
| 25 | from httpx import AsyncClient |
| 26 | from sqlalchemy.ext.asyncio import AsyncSession |
| 27 | |
| 28 | from musehub.services.musehub_exporter import ( |
| 29 | ExportFormat, |
| 30 | ExportResult, |
| 31 | export_repo_at_ref, |
| 32 | ) |
| 33 | |
| 34 | |
| 35 | # --------------------------------------------------------------------------- |
| 36 | # Helpers |
| 37 | # --------------------------------------------------------------------------- |
| 38 | |
| 39 | _MIDI_BYTES = b"MThd\x00\x00\x00\x06\x00\x01\x00\x01\x01\xe0" # minimal valid MIDI header |
| 40 | _MP3_BYTES = b"\xff\xfb\x90\x00" + b"\x00" * 60 # minimal MP3 frame marker |
| 41 | |
| 42 | |
| 43 | def _b64(data: bytes) -> str: |
| 44 | return base64.b64encode(data).decode() |
| 45 | |
| 46 | |
| 47 | async def _create_repo(client: AsyncClient, auth_headers: dict[str, str], name: str = "export-test") -> str: |
| 48 | r = await client.post("/api/v1/musehub/repos", json={"name": name, "owner": "testuser"}, headers=auth_headers) |
| 49 | assert r.status_code == 201 |
| 50 | repo_id: str = r.json()["repoId"] |
| 51 | return repo_id |
| 52 | |
| 53 | |
| 54 | async def _push_with_objects( |
| 55 | client: AsyncClient, |
| 56 | auth_headers: dict[str, str], |
| 57 | repo_id: str, |
| 58 | commit_id: str, |
| 59 | objects: list[dict[str, object]], |
| 60 | tmp_dir: str, |
| 61 | ) -> None: |
| 62 | """Push a commit + objects and patch musehub_sync.settings to use tmp_dir.""" |
| 63 | with patch("musehub.services.musehub_sync.settings") as mock_cfg: |
| 64 | mock_cfg.musehub_objects_dir = tmp_dir |
| 65 | r = await client.post( |
| 66 | f"/api/v1/musehub/repos/{repo_id}/push", |
| 67 | json={ |
| 68 | "branch": "main", |
| 69 | "headCommitId": commit_id, |
| 70 | "commits": [ |
| 71 | { |
| 72 | "commitId": commit_id, |
| 73 | "parentIds": [], |
| 74 | "message": "test commit", |
| 75 | "timestamp": "2024-01-01T00:00:00Z", |
| 76 | } |
| 77 | ], |
| 78 | "objects": objects, |
| 79 | "force": False, |
| 80 | }, |
| 81 | headers=auth_headers, |
| 82 | ) |
| 83 | assert r.status_code == 200, r.text |
| 84 | |
| 85 | |
| 86 | # --------------------------------------------------------------------------- |
| 87 | # test_export_midi — MIDI export returns a .mid file |
| 88 | # --------------------------------------------------------------------------- |
| 89 | |
| 90 | |
| 91 | @pytest.mark.anyio |
| 92 | async def test_export_midi( |
| 93 | client: AsyncClient, |
| 94 | auth_headers: dict[str, str], |
| 95 | db_session: AsyncSession, |
| 96 | ) -> None: |
| 97 | """A single MIDI object → direct .mid download with correct MIME type.""" |
| 98 | with tempfile.TemporaryDirectory() as tmp: |
| 99 | repo_id = await _create_repo(client, auth_headers, "midi-export") |
| 100 | await _push_with_objects( |
| 101 | client, |
| 102 | auth_headers, |
| 103 | repo_id, |
| 104 | commit_id="c-midi-001", |
| 105 | objects=[ |
| 106 | { |
| 107 | "objectId": "sha256:midi001", |
| 108 | "path": "tracks/bass.mid", |
| 109 | "contentB64": _b64(_MIDI_BYTES), |
| 110 | } |
| 111 | ], |
| 112 | tmp_dir=tmp, |
| 113 | ) |
| 114 | |
| 115 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 116 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 117 | from datetime import datetime, timezone |
| 118 | |
| 119 | mock_repo.get_commit = AsyncMock( |
| 120 | return_value=CommitResponse( |
| 121 | commit_id="c-midi-001", |
| 122 | branch="main", |
| 123 | parent_ids=[], |
| 124 | message="test", |
| 125 | author="alice", |
| 126 | timestamp=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 127 | snapshot_id=None, |
| 128 | ) |
| 129 | ) |
| 130 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 131 | meta = ObjectMetaResponse( |
| 132 | object_id="sha256:midi001", |
| 133 | path="tracks/bass.mid", |
| 134 | size_bytes=len(_MIDI_BYTES), |
| 135 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 136 | ) |
| 137 | mock_repo.list_objects = AsyncMock(return_value=[meta]) |
| 138 | |
| 139 | from musehub.db import musehub_models as db_models |
| 140 | |
| 141 | disk_path = str(Path(tmp) / "sha256_midi001.mid") |
| 142 | Path(disk_path).write_bytes(_MIDI_BYTES) |
| 143 | |
| 144 | fake_row = db_models.MusehubObject() |
| 145 | fake_row.object_id = "sha256:midi001" |
| 146 | fake_row.path = "tracks/bass.mid" |
| 147 | fake_row.disk_path = disk_path |
| 148 | fake_row.size_bytes = len(_MIDI_BYTES) |
| 149 | mock_repo.get_object_row = AsyncMock(return_value=fake_row) |
| 150 | |
| 151 | r = await client.get( |
| 152 | f"/api/v1/musehub/repos/{repo_id}/export/c-midi-001?format=midi", |
| 153 | headers=auth_headers, |
| 154 | ) |
| 155 | |
| 156 | assert r.status_code == 200 |
| 157 | assert r.headers["content-type"] == "audio/midi" |
| 158 | assert "bass.mid" in r.headers.get("content-disposition", "") |
| 159 | assert r.content == _MIDI_BYTES |
| 160 | |
| 161 | |
| 162 | # --------------------------------------------------------------------------- |
| 163 | # test_export_json — JSON export returns valid JSON |
| 164 | # --------------------------------------------------------------------------- |
| 165 | |
| 166 | |
| 167 | @pytest.mark.anyio |
| 168 | async def test_export_json( |
| 169 | client: AsyncClient, |
| 170 | auth_headers: dict[str, str], |
| 171 | db_session: AsyncSession, |
| 172 | ) -> None: |
| 173 | """format=json returns a valid JSON document with commit and object metadata.""" |
| 174 | with tempfile.TemporaryDirectory() as tmp: |
| 175 | repo_id = await _create_repo(client, auth_headers, "json-export") |
| 176 | |
| 177 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 178 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 179 | from datetime import datetime, timezone |
| 180 | |
| 181 | mock_repo.get_commit = AsyncMock( |
| 182 | return_value=CommitResponse( |
| 183 | commit_id="c-json-001", |
| 184 | branch="main", |
| 185 | parent_ids=[], |
| 186 | message="json export test", |
| 187 | author="bob", |
| 188 | timestamp=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 189 | snapshot_id=None, |
| 190 | ) |
| 191 | ) |
| 192 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 193 | mock_repo.list_objects = AsyncMock( |
| 194 | return_value=[ |
| 195 | ObjectMetaResponse( |
| 196 | object_id="sha256:obj001", |
| 197 | path="tracks/keys.mid", |
| 198 | size_bytes=100, |
| 199 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 200 | ) |
| 201 | ] |
| 202 | ) |
| 203 | |
| 204 | r = await client.get( |
| 205 | f"/api/v1/musehub/repos/{repo_id}/export/c-json-001?format=json", |
| 206 | headers=auth_headers, |
| 207 | ) |
| 208 | |
| 209 | assert r.status_code == 200 |
| 210 | assert "application/json" in r.headers["content-type"] |
| 211 | payload = json.loads(r.content) |
| 212 | assert payload["commit_id"] == "c-json-001" |
| 213 | assert payload["repo_id"] == repo_id |
| 214 | assert len(payload["objects"]) == 1 |
| 215 | assert payload["objects"][0]["path"] == "tracks/keys.mid" |
| 216 | |
| 217 | |
| 218 | # --------------------------------------------------------------------------- |
| 219 | # test_export_split_tracks_zip — split_tracks produces a ZIP |
| 220 | # --------------------------------------------------------------------------- |
| 221 | |
| 222 | |
| 223 | @pytest.mark.anyio |
| 224 | async def test_export_split_tracks_zip( |
| 225 | client: AsyncClient, |
| 226 | auth_headers: dict[str, str], |
| 227 | db_session: AsyncSession, |
| 228 | ) -> None: |
| 229 | """split_tracks=true with two MIDI objects produces a ZIP with both files.""" |
| 230 | with tempfile.TemporaryDirectory() as tmp: |
| 231 | repo_id = await _create_repo(client, auth_headers, "zip-export") |
| 232 | |
| 233 | bass_path = str(Path(tmp) / "bass.mid") |
| 234 | keys_path = str(Path(tmp) / "keys.mid") |
| 235 | Path(bass_path).write_bytes(_MIDI_BYTES) |
| 236 | Path(keys_path).write_bytes(_MIDI_BYTES + b"\x00") |
| 237 | |
| 238 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 239 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 240 | from datetime import datetime, timezone |
| 241 | from musehub.db import musehub_models as db_models |
| 242 | |
| 243 | mock_repo.get_commit = AsyncMock( |
| 244 | return_value=CommitResponse( |
| 245 | commit_id="c-zip-001", |
| 246 | branch="main", |
| 247 | parent_ids=[], |
| 248 | message="zip test", |
| 249 | author="carol", |
| 250 | timestamp=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 251 | snapshot_id=None, |
| 252 | ) |
| 253 | ) |
| 254 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 255 | mock_repo.list_objects = AsyncMock( |
| 256 | return_value=[ |
| 257 | ObjectMetaResponse( |
| 258 | object_id="sha256:bass", |
| 259 | path="tracks/bass.mid", |
| 260 | size_bytes=len(_MIDI_BYTES), |
| 261 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 262 | ), |
| 263 | ObjectMetaResponse( |
| 264 | object_id="sha256:keys", |
| 265 | path="tracks/keys.mid", |
| 266 | size_bytes=len(_MIDI_BYTES) + 1, |
| 267 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 268 | ), |
| 269 | ] |
| 270 | ) |
| 271 | |
| 272 | def _fake_get_object_row( |
| 273 | session: object, repo_id: str, object_id: str |
| 274 | ) -> db_models.MusehubObject: |
| 275 | row = db_models.MusehubObject() |
| 276 | row.object_id = object_id |
| 277 | if object_id == "sha256:bass": |
| 278 | row.path = "tracks/bass.mid" |
| 279 | row.disk_path = bass_path |
| 280 | else: |
| 281 | row.path = "tracks/keys.mid" |
| 282 | row.disk_path = keys_path |
| 283 | row.size_bytes = 0 |
| 284 | return row |
| 285 | |
| 286 | mock_repo.get_object_row = AsyncMock(side_effect=_fake_get_object_row) |
| 287 | |
| 288 | r = await client.get( |
| 289 | f"/api/v1/musehub/repos/{repo_id}/export/c-zip-001?format=midi&splitTracks=true", |
| 290 | headers=auth_headers, |
| 291 | ) |
| 292 | |
| 293 | assert r.status_code == 200 |
| 294 | assert r.headers["content-type"] == "application/zip" |
| 295 | zf = zipfile.ZipFile(io.BytesIO(r.content)) |
| 296 | names = zf.namelist() |
| 297 | assert "bass.mid" in names |
| 298 | assert "keys.mid" in names |
| 299 | |
| 300 | |
| 301 | # --------------------------------------------------------------------------- |
| 302 | # test_export_section_filter — sections param filters artifacts by path substring |
| 303 | # --------------------------------------------------------------------------- |
| 304 | |
| 305 | |
| 306 | @pytest.mark.anyio |
| 307 | async def test_export_section_filter( |
| 308 | client: AsyncClient, |
| 309 | auth_headers: dict[str, str], |
| 310 | db_session: AsyncSession, |
| 311 | ) -> None: |
| 312 | """sections=verse includes only artifacts whose path contains 'verse'.""" |
| 313 | with tempfile.TemporaryDirectory() as tmp: |
| 314 | repo_id = await _create_repo(client, auth_headers, "section-export") |
| 315 | |
| 316 | verse_path = str(Path(tmp) / "verse_bass.mid") |
| 317 | chorus_path = str(Path(tmp) / "chorus_bass.mid") |
| 318 | Path(verse_path).write_bytes(_MIDI_BYTES) |
| 319 | Path(chorus_path).write_bytes(_MIDI_BYTES + b"\x01") |
| 320 | |
| 321 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 322 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 323 | from datetime import datetime, timezone |
| 324 | from musehub.db import musehub_models as db_models |
| 325 | |
| 326 | mock_repo.get_commit = AsyncMock( |
| 327 | return_value=CommitResponse( |
| 328 | commit_id="c-sec-001", |
| 329 | branch="main", |
| 330 | parent_ids=[], |
| 331 | message="section test", |
| 332 | author="dave", |
| 333 | timestamp=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 334 | snapshot_id=None, |
| 335 | ) |
| 336 | ) |
| 337 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 338 | mock_repo.list_objects = AsyncMock( |
| 339 | return_value=[ |
| 340 | ObjectMetaResponse( |
| 341 | object_id="sha256:verse", |
| 342 | path="tracks/verse_bass.mid", |
| 343 | size_bytes=len(_MIDI_BYTES), |
| 344 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 345 | ), |
| 346 | ObjectMetaResponse( |
| 347 | object_id="sha256:chorus", |
| 348 | path="tracks/chorus_bass.mid", |
| 349 | size_bytes=len(_MIDI_BYTES) + 1, |
| 350 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 351 | ), |
| 352 | ] |
| 353 | ) |
| 354 | |
| 355 | verse_row = db_models.MusehubObject() |
| 356 | verse_row.object_id = "sha256:verse" |
| 357 | verse_row.path = "tracks/verse_bass.mid" |
| 358 | verse_row.disk_path = verse_path |
| 359 | verse_row.size_bytes = len(_MIDI_BYTES) |
| 360 | mock_repo.get_object_row = AsyncMock(return_value=verse_row) |
| 361 | |
| 362 | r = await client.get( |
| 363 | f"/api/v1/musehub/repos/{repo_id}/export/c-sec-001?format=midi§ions=verse", |
| 364 | headers=auth_headers, |
| 365 | ) |
| 366 | |
| 367 | assert r.status_code == 200 |
| 368 | assert r.content == _MIDI_BYTES |
| 369 | assert "verse_bass.mid" in r.headers.get("content-disposition", "") |
| 370 | |
| 371 | |
| 372 | # --------------------------------------------------------------------------- |
| 373 | # test_export_unknown_format_422 — invalid format returns 422 |
| 374 | # --------------------------------------------------------------------------- |
| 375 | |
| 376 | |
| 377 | @pytest.mark.anyio |
| 378 | async def test_export_unknown_format_422( |
| 379 | client: AsyncClient, |
| 380 | auth_headers: dict[str, str], |
| 381 | db_session: AsyncSession, |
| 382 | ) -> None: |
| 383 | """An unrecognised format query param returns HTTP 422.""" |
| 384 | repo_id = await _create_repo(client, auth_headers, "bad-format") |
| 385 | r = await client.get( |
| 386 | f"/api/v1/musehub/repos/{repo_id}/export/main?format=flac", |
| 387 | headers=auth_headers, |
| 388 | ) |
| 389 | assert r.status_code == 422 |
| 390 | |
| 391 | |
| 392 | # --------------------------------------------------------------------------- |
| 393 | # test_export_ref_not_found_404 — unresolvable ref returns 404 |
| 394 | # --------------------------------------------------------------------------- |
| 395 | |
| 396 | |
| 397 | @pytest.mark.anyio |
| 398 | async def test_export_ref_not_found_404( |
| 399 | client: AsyncClient, |
| 400 | auth_headers: dict[str, str], |
| 401 | db_session: AsyncSession, |
| 402 | ) -> None: |
| 403 | """A ref that does not match any commit or branch returns HTTP 404.""" |
| 404 | repo_id = await _create_repo(client, auth_headers, "missing-ref") |
| 405 | |
| 406 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 407 | mock_repo.get_commit = AsyncMock(return_value=None) |
| 408 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 409 | |
| 410 | r = await client.get( |
| 411 | f"/api/v1/musehub/repos/{repo_id}/export/nonexistent-sha?format=midi", |
| 412 | headers=auth_headers, |
| 413 | ) |
| 414 | |
| 415 | assert r.status_code == 404 |
| 416 | assert "nonexistent-sha" in r.json()["detail"] |
| 417 | |
| 418 | |
| 419 | # --------------------------------------------------------------------------- |
| 420 | # Unit tests for export_repo_at_ref service function |
| 421 | # --------------------------------------------------------------------------- |
| 422 | |
| 423 | |
| 424 | @pytest.mark.anyio |
| 425 | async def test_export_service_returns_ref_not_found_sentinel() -> None: |
| 426 | """export_repo_at_ref returns 'ref_not_found' when the ref resolves to nothing.""" |
| 427 | from unittest.mock import MagicMock |
| 428 | |
| 429 | mock_session = MagicMock() |
| 430 | |
| 431 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 432 | mock_repo.get_commit = AsyncMock(return_value=None) |
| 433 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 434 | |
| 435 | result = await export_repo_at_ref( |
| 436 | mock_session, |
| 437 | repo_id="repo-x", |
| 438 | ref="deadbeef", |
| 439 | format=ExportFormat.midi, |
| 440 | ) |
| 441 | |
| 442 | assert result == "ref_not_found" |
| 443 | |
| 444 | |
| 445 | @pytest.mark.anyio |
| 446 | async def test_export_service_returns_no_matching_objects_sentinel() -> None: |
| 447 | """export_repo_at_ref returns 'no_matching_objects' when no objects match the format.""" |
| 448 | from unittest.mock import MagicMock |
| 449 | from datetime import datetime, timezone |
| 450 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 451 | |
| 452 | mock_session = MagicMock() |
| 453 | |
| 454 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 455 | mock_repo.get_commit = AsyncMock( |
| 456 | return_value=CommitResponse( |
| 457 | commit_id="abc123", |
| 458 | branch="main", |
| 459 | parent_ids=[], |
| 460 | message="x", |
| 461 | author="x", |
| 462 | timestamp=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 463 | snapshot_id=None, |
| 464 | ) |
| 465 | ) |
| 466 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 467 | mock_repo.list_objects = AsyncMock( |
| 468 | return_value=[ |
| 469 | ObjectMetaResponse( |
| 470 | object_id="sha256:img", |
| 471 | path="piano_roll.webp", |
| 472 | size_bytes=512, |
| 473 | created_at=datetime(2024, 1, 1, tzinfo=timezone.utc), |
| 474 | ) |
| 475 | ] |
| 476 | ) |
| 477 | |
| 478 | result = await export_repo_at_ref( |
| 479 | mock_session, |
| 480 | repo_id="repo-y", |
| 481 | ref="abc123", |
| 482 | format=ExportFormat.midi, |
| 483 | ) |
| 484 | |
| 485 | assert result == "no_matching_objects" |
| 486 | |
| 487 | |
| 488 | @pytest.mark.anyio |
| 489 | async def test_export_service_json_format_no_disk_access() -> None: |
| 490 | """format=json returns ExportResult with JSON bytes without reading any disk file.""" |
| 491 | from unittest.mock import MagicMock |
| 492 | from datetime import datetime, timezone |
| 493 | from musehub.models.musehub import CommitResponse, ObjectMetaResponse |
| 494 | |
| 495 | mock_session = MagicMock() |
| 496 | |
| 497 | with patch("musehub.services.musehub_exporter.musehub_repository") as mock_repo: |
| 498 | mock_repo.get_commit = AsyncMock( |
| 499 | return_value=CommitResponse( |
| 500 | commit_id="abc999", |
| 501 | branch="dev", |
| 502 | parent_ids=[], |
| 503 | message="json only", |
| 504 | author="eve", |
| 505 | timestamp=datetime(2024, 3, 1, tzinfo=timezone.utc), |
| 506 | snapshot_id=None, |
| 507 | ) |
| 508 | ) |
| 509 | mock_repo.list_branches = AsyncMock(return_value=[]) |
| 510 | mock_repo.list_objects = AsyncMock( |
| 511 | return_value=[ |
| 512 | ObjectMetaResponse( |
| 513 | object_id="sha256:mid", |
| 514 | path="track.mid", |
| 515 | size_bytes=200, |
| 516 | created_at=datetime(2024, 3, 1, tzinfo=timezone.utc), |
| 517 | ) |
| 518 | ] |
| 519 | ) |
| 520 | |
| 521 | result = await export_repo_at_ref( |
| 522 | mock_session, |
| 523 | repo_id="repo-z", |
| 524 | ref="abc999", |
| 525 | format=ExportFormat.json, |
| 526 | ) |
| 527 | |
| 528 | assert isinstance(result, ExportResult) |
| 529 | assert result.content_type == "application/json" |
| 530 | parsed = json.loads(result.content) |
| 531 | assert parsed["commit_id"] == "abc999" |
| 532 | assert parsed["repo_id"] == "repo-z" |
| 533 | assert len(parsed["objects"]) == 1 |