Coverage for tests / test_simpleButler.py: 12%

1# This file is part of daf_butler. 

2# 

3# Developed for the LSST Data Management System. 

4# This product includes software developed by the LSST Project 

5# (http://www.lsst.org). 

6# See the COPYRIGHT file at the top-level directory of this distribution 

7# for details of code ownership. 

8# 

9# This software is dual licensed under the GNU General Public License and also 

10# under a 3-clause BSD license. Recipients may choose which of these licenses 

11# to use; please see the files gpl-3.0.txt and/or bsd_license.txt, 

12# respectively. If you choose the GPL option then the following text applies 

13# (but note that there is still no warranty even if you opt for BSD instead): 

14# 

15# This program is free software: you can redistribute it and/or modify 

16# it under the terms of the GNU General Public License as published by 

17# the Free Software Foundation, either version 3 of the License, or 

18# (at your option) any later version. 

19# 

20# This program is distributed in the hope that it will be useful, 

21# but WITHOUT ANY WARRANTY; without even the implied warranty of 

22# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 

23# GNU General Public License for more details. 

24# 

25# You should have received a copy of the GNU General Public License 

26# along with this program. If not, see <http://www.gnu.org/licenses/>. 

27 

28from __future__ import annotations 

29 

30import json 

31import os 

32import re 

33import tempfile 

34import unittest 

35import urllib.parse 

36from typing import Any 

37 

38try: 

39 import numpy as np 

40except ImportError: 

41 np = None 

42 

43import astropy.time 

44 

45from lsst.daf.butler import ( 

46 Butler, 

47 ButlerConfig, 

48 CollectionType, 

49 DataCoordinate, 

50 DatasetId, 

51 DatasetRef, 

52 DatasetType, 

53 LabeledButlerFactory, 

54 StorageClass, 

55 Timespan, 

56) 

57from lsst.daf.butler.datastore.file_templates import FileTemplate 

58from lsst.daf.butler.registry import RegistryConfig, RegistryDefaults, _RegistryFactory 

59from lsst.daf.butler.tests import DatastoreMock 

60from lsst.daf.butler.tests.server_available import butler_server_import_error, butler_server_is_available 

61from lsst.daf.butler.tests.utils import TestCaseMixin, makeTestTempDir, mock_env, removeTestTempDir 

62 

63if butler_server_is_available: 63 ↛ 64line 63 didn't jump to line 64 because the condition on line 63 was never true

64 from lsst.daf.butler.tests.server import create_test_server 

65 

66TESTDIR = os.path.abspath(os.path.dirname(__file__)) 

67 

68 

69class SimpleButlerTests(TestCaseMixin): 

70 """Tests for butler (including import/export functionality) that should not 

71 depend on the Registry Database backend or Datastore implementation, and 

72 can instead utilize an in-memory SQLite Registry and a mocked Datastore. 

73 """ 

74 

75 datasetsImportFile = "datasets.yaml" 

76 

77 supportsCollectionRegex: bool = True 

78 """True if the registry class being tested supports regex searches for 

79 collections.""" 

80 

81 def makeButler(self, writeable: bool = False) -> Butler: 

82 raise NotImplementedError() 

83 

84 def comparableRef(self, ref: DatasetRef) -> DatasetRef: 

85 """Return a DatasetRef that can be compared to a DatasetRef from 

86 other repository. 

87 

88 For repositories that do not support round-trip of ID values this 

89 method returns unresolved DatasetRef, for round-trip-safe repos it 

90 returns unchanged ref. 

91 """ 

92 return ref 

93 

94 def testReadBackwardsCompatibility(self): 

95 """Test that we can read an export file written by a previous version 

96 and commit to the daf_butler git repo. 

97 

98 Notes 

99 ----- 

100 At present this export file includes only dimension data, not datasets, 

101 which greatly limits the usefulness of this test. We should address 

102 this at some point, but I think it's best to wait for the changes to 

103 the export format required for CALIBRATION collections to land. 

104 """ 

105 butler = self.makeButler(writeable=True) 

106 butler.import_(filename=os.path.join(TESTDIR, "data", "registry", "hsc-rc2-subset.yaml")) 

107 # Spot-check a few things, but the most important test is just that 

108 # the above does not raise. 

109 self.assertGreaterEqual( 

110 {record.id for record in butler.registry.queryDimensionRecords("detector", instrument="HSC")}, 

111 set(range(104)), # should have all science CCDs; may have some focus ones. 

112 ) 

113 self.assertGreaterEqual( 

114 { 

115 (record.id, record.physical_filter) 

116 for record in butler.registry.queryDimensionRecords("visit", instrument="HSC") 

117 }, 

118 { 

119 (27136, "HSC-Z"), 

120 (11694, "HSC-G"), 

121 (23910, "HSC-R"), 

122 (11720, "HSC-Y"), 

123 (23900, "HSC-R"), 

124 (22646, "HSC-Y"), 

125 (1248, "HSC-I"), 

126 (19680, "HSC-I"), 

127 (1240, "HSC-I"), 

128 (424, "HSC-Y"), 

129 (19658, "HSC-I"), 

130 (344, "HSC-Y"), 

131 (1218, "HSC-R"), 

132 (1190, "HSC-Z"), 

133 (23718, "HSC-R"), 

134 (11700, "HSC-G"), 

135 (26036, "HSC-G"), 

136 (23872, "HSC-R"), 

137 (1170, "HSC-Z"), 

138 (1876, "HSC-Y"), 

139 }, 

140 ) 

141 

142 def testDatasetTransfers(self): 

143 """Test exporting all datasets from a repo and then importing them all 

144 back in again. 

145 """ 

146 # Import data to play with. 

147 butler1 = self.makeButler(writeable=True) 

148 butler1.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

149 butler1.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

150 with tempfile.NamedTemporaryFile(mode="w", suffix=".yaml") as file: 

151 # Export all datasets. 

152 with butler1.export(filename=file.name) as exporter: 

153 exporter.saveDatasets(butler1.registry.queryDatasets(..., collections=...)) 

154 # Import it all again. 

155 butler2 = self.makeButler(writeable=True) 

156 butler2.import_(filename=file.name) 

157 datasets1 = list(butler1.registry.queryDatasets(..., collections=...)) 

158 datasets2 = list(butler2.registry.queryDatasets(..., collections=...)) 

159 self.assertTrue(all(isinstance(ref.id, DatasetId) for ref in datasets1)) 

160 self.assertTrue(all(isinstance(ref.id, DatasetId) for ref in datasets2)) 

161 self.assertCountEqual( 

162 [self.comparableRef(ref) for ref in datasets1], 

163 [self.comparableRef(ref) for ref in datasets2], 

164 ) 

165 

166 def testImportTwice(self): 

167 """Test exporting dimension records and datasets from a repo and then 

168 importing them all back in again twice. 

169 """ 

170 # Import data to play with. 

171 butler1 = self.makeButler(writeable=True) 

172 butler1.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

173 butler1.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

174 with tempfile.NamedTemporaryFile(mode="w", suffix=".yaml", delete=False) as file: 

175 # Export all datasets. 

176 with butler1.export(filename=file.name) as exporter: 

177 exporter.saveDatasets(butler1.registry.queryDatasets(..., collections=...)) 

178 butler2 = self.makeButler(writeable=True) 

179 # Import it once. 

180 butler2.import_(filename=file.name) 

181 # Import it again 

182 butler2.import_(filename=file.name) 

183 datasets1 = list(butler1.registry.queryDatasets(..., collections=...)) 

184 datasets2 = list(butler2.registry.queryDatasets(..., collections=...)) 

185 self.assertTrue(all(isinstance(ref.id, DatasetId) for ref in datasets1)) 

186 self.assertTrue(all(isinstance(ref.id, DatasetId) for ref in datasets2)) 

187 self.assertCountEqual( 

188 [self.comparableRef(ref) for ref in datasets1], 

189 [self.comparableRef(ref) for ref in datasets2], 

190 ) 

191 

192 def testCollectionTransfers(self): 

193 """Test exporting and then importing collections of various types.""" 

194 # Populate a registry with some datasets. 

195 butler1 = self.makeButler(writeable=True) 

196 butler1.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

197 butler1.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

198 registry1 = butler1.registry 

199 # Add some more collections. 

200 registry1.registerRun("run1") 

201 registry1.registerCollection("tag1", CollectionType.TAGGED) 

202 registry1.registerCollection("calibration1", CollectionType.CALIBRATION) 

203 registry1.registerCollection("chain1", CollectionType.CHAINED) 

204 registry1.registerCollection("chain2", CollectionType.CHAINED) 

205 registry1.setCollectionChain("chain1", ["tag1", "run1", "chain2"]) 

206 registry1.setCollectionChain("chain2", ["calibration1", "run1"]) 

207 # Associate some datasets into the TAGGED and CALIBRATION collections. 

208 flats1 = list(registry1.queryDatasets("flat", collections=...)) 

209 registry1.associate("tag1", flats1) 

210 t1 = astropy.time.Time("2020-01-01T01:00:00", format="isot", scale="tai") 

211 t2 = astropy.time.Time("2020-01-01T02:00:00", format="isot", scale="tai") 

212 t3 = astropy.time.Time("2020-01-01T03:00:00", format="isot", scale="tai") 

213 bias1a = registry1.findDataset("bias", instrument="Cam1", detector=1, collections="imported_g") 

214 bias2a = registry1.findDataset("bias", instrument="Cam1", detector=2, collections="imported_g") 

215 bias3a = registry1.findDataset("bias", instrument="Cam1", detector=3, collections="imported_g") 

216 bias2b = registry1.findDataset("bias", instrument="Cam1", detector=2, collections="imported_r") 

217 bias3b = registry1.findDataset("bias", instrument="Cam1", detector=3, collections="imported_r") 

218 registry1.certify("calibration1", [bias2a, bias3a], Timespan(t1, t2)) 

219 registry1.certify("calibration1", [bias2b], Timespan(t2, None)) 

220 registry1.certify("calibration1", [bias3b], Timespan(t2, t3)) 

221 registry1.certify("calibration1", [bias1a], Timespan.makeEmpty()) 

222 

223 with tempfile.NamedTemporaryFile(mode="w", suffix=".yaml") as file: 

224 # Export all collections, and some datasets. 

225 with butler1.export(filename=file.name) as exporter: 

226 # Sort results to put chain1 before chain2, which is 

227 # intentionally not topological order. 

228 for collection in sorted(registry1.queryCollections()): 

229 exporter.saveCollection(collection) 

230 exporter.saveDatasets(flats1) 

231 exporter.saveDatasets([bias1a, bias2a, bias2b, bias3a, bias3b]) 

232 # Import them into a new registry. 

233 butler2 = self.makeButler(writeable=True) 

234 butler2.import_(filename=file.name) 

235 registry2 = butler2.registry 

236 # Check that it all round-tripped, starting with the collections 

237 # themselves. 

238 self.assertIs(registry2.getCollectionType("run1"), CollectionType.RUN) 

239 self.assertIs(registry2.getCollectionType("tag1"), CollectionType.TAGGED) 

240 self.assertIs(registry2.getCollectionType("calibration1"), CollectionType.CALIBRATION) 

241 self.assertIs(registry2.getCollectionType("chain1"), CollectionType.CHAINED) 

242 self.assertIs(registry2.getCollectionType("chain2"), CollectionType.CHAINED) 

243 self.assertEqual( 

244 list(registry2.getCollectionChain("chain1")), 

245 ["tag1", "run1", "chain2"], 

246 ) 

247 self.assertEqual( 

248 list(registry2.getCollectionChain("chain2")), 

249 ["calibration1", "run1"], 

250 ) 

251 # Check that tag collection contents are the same. 

252 self.maxDiff = None 

253 self.assertCountEqual( 

254 [self.comparableRef(ref) for ref in registry1.queryDatasets(..., collections="tag1")], 

255 [self.comparableRef(ref) for ref in registry2.queryDatasets(..., collections="tag1")], 

256 ) 

257 # Check that calibration collection contents are the same. 

258 self.assertCountEqual( 

259 [ 

260 (self.comparableRef(assoc.ref), assoc.timespan) 

261 for assoc in registry1.queryDatasetAssociations("bias", collections="calibration1") 

262 ], 

263 [ 

264 (self.comparableRef(assoc.ref), assoc.timespan) 

265 for assoc in registry2.queryDatasetAssociations("bias", collections="calibration1") 

266 ], 

267 ) 

268 

269 def testButlerGet(self): 

270 """Test that butler.get can work with different variants.""" 

271 # Import data to play with. 

272 butler = self.makeButler(writeable=True) 

273 butler.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

274 butler.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

275 

276 # Find the DatasetRef for a flat 

277 coll = "imported_g" 

278 flat2g = butler.find_dataset( 

279 "flat", instrument="Cam1", full_name="Ab", physical_filter="Cam1-G", collections=coll 

280 ) 

281 

282 # Create a numpy integer to check that works fine 

283 detector_np = np.int64(2) if np else 2 

284 

285 # Try to get it using different variations of dataId + keyword 

286 # arguments 

287 # Note that instrument.class_name does not work 

288 variants = ( 

289 (None, {"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G"}), 

290 (None, {"instrument": "Cam1", "detector": detector_np, "physical_filter": "Cam1-G"}), 

291 ({"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G"}, {}), 

292 ({"instrument": "Cam1", "detector": detector_np, "physical_filter": "Cam1-G"}, {}), 

293 ({"instrument": "Cam1", "detector": 2}, {"physical_filter": "Cam1-G"}), 

294 ({"detector.full_name": "Ab"}, {"instrument": "Cam1", "physical_filter": "Cam1-G"}), 

295 ({"full_name": "Ab"}, {"instrument": "Cam1", "physical_filter": "Cam1-G"}), 

296 (None, {"full_name": "Ab", "instrument": "Cam1", "physical_filter": "Cam1-G"}), 

297 (None, {"detector": "Ab", "instrument": "Cam1", "physical_filter": "Cam1-G"}), 

298 ({"name_in_raft": "b", "raft": "A"}, {"instrument": "Cam1", "physical_filter": "Cam1-G"}), 

299 ({"name_in_raft": "b"}, {"raft": "A", "instrument": "Cam1", "physical_filter": "Cam1-G"}), 

300 (None, {"name_in_raft": "b", "raft": "A", "instrument": "Cam1", "physical_filter": "Cam1-G"}), 

301 ( 

302 {"detector.name_in_raft": "b", "detector.raft": "A"}, 

303 {"instrument": "Cam1", "physical_filter": "Cam1-G"}, 

304 ), 

305 ( 

306 { 

307 "detector.name_in_raft": "b", 

308 "detector.raft": "A", 

309 "instrument": "Cam1", 

310 "physical_filter": "Cam1-G", 

311 }, 

312 {}, 

313 ), 

314 # Duplicate (but valid) information. 

315 (None, {"instrument": "Cam1", "detector": 2, "raft": "A", "physical_filter": "Cam1-G"}), 

316 ({"detector": 2}, {"instrument": "Cam1", "raft": "A", "physical_filter": "Cam1-G"}), 

317 ({"raft": "A"}, {"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G"}), 

318 ({"raft": "A"}, {"instrument": "Cam1", "detector": "Ab", "physical_filter": "Cam1-G"}), 

319 ) 

320 butler._metrics.reset() 

321 n_got = 0 

322 for dataId, kwds in variants: 

323 try: 

324 flat_id, _ = butler.get("flat", dataId=dataId, collections=coll, **kwds) 

325 n_got += 1 

326 except Exception as e: 

327 e.add_note(f"dataId={dataId}, kwds={kwds}") 

328 raise 

329 self.assertEqual(flat_id, flat2g.id, msg=f"DataId: {dataId}, kwds: {kwds}") 

330 self.assertEqual(butler._metrics.n_get, n_got) 

331 

332 # Check that bad combinations raise. 

333 variants = ( 

334 # Inconsistent detector information. 

335 (None, {"instrument": "Cam1", "detector": 2, "raft": "B", "physical_filter": "Cam1-G"}), 

336 ({"detector": 2}, {"instrument": "Cam1", "raft": "B", "physical_filter": "Cam1-G"}), 

337 ({"detector": 12}, {"instrument": "Cam1", "raft": "B", "physical_filter": "Cam1-G"}), 

338 ({"raft": "B"}, {"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G"}), 

339 ({"raft": "B"}, {"instrument": "Cam1", "detector": "Ab", "physical_filter": "Cam1-G"}), 

340 # Under-specified. 

341 ({"raft": "B"}, {"instrument": "Cam1", "physical_filter": "Cam1-G"}), 

342 # Spurious kwargs. 

343 (None, {"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G", "x": "y"}), 

344 ({"x": "y"}, {"instrument": "Cam1", "detector": 2, "physical_filter": "Cam1-G"}), 

345 ) 

346 for dataId, kwds in variants: 

347 with self.assertRaises((ValueError, LookupError)): 

348 butler.get("flat", dataId=dataId, collections=coll, **kwds) 

349 

350 def testGetCalibration(self): 

351 """Test that `Butler.get` can be used to fetch from 

352 `~CollectionType.CALIBRATION` collections if the data ID includes 

353 extra dimensions with temporal information. 

354 """ 

355 # Import data to play with. 

356 butler = self.makeButler(writeable=True) 

357 butler.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

358 butler.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

359 # Certify some biases into a CALIBRATION collection. 

360 registry = butler.registry 

361 registry.registerCollection("calibs", CollectionType.CALIBRATION) 

362 t1 = astropy.time.Time("2020-01-01T01:00:00", format="isot", scale="tai") 

363 t2 = astropy.time.Time("2020-01-01T02:00:00", format="isot", scale="tai") 

364 t3 = astropy.time.Time("2020-01-01T03:00:00", format="isot", scale="tai") 

365 bias1a = registry.findDataset("bias", instrument="Cam1", detector=1, collections="imported_g") 

366 bias2a = registry.findDataset("bias", instrument="Cam1", detector=2, collections="imported_g") 

367 bias3a = registry.findDataset("bias", instrument="Cam1", detector=3, collections="imported_g") 

368 bias2b = registry.findDataset("bias", instrument="Cam1", detector=2, collections="imported_r") 

369 bias3b = registry.findDataset("bias", instrument="Cam1", detector=3, collections="imported_r") 

370 registry.certify("calibs", [bias1a], Timespan(t1, t2)) 

371 registry.certify("calibs", [bias2a, bias3a], Timespan(t1, t2)) 

372 registry.certify("calibs", [bias2b], Timespan(t2, None)) 

373 registry.certify("calibs", [bias3b], Timespan(t2, t3)) 

374 # Insert some exposure dimension data. 

375 registry.insertDimensionData( 

376 "group", 

377 {"instrument": "Cam1", "group": "three"}, 

378 {"instrument": "Cam1", "group": "four"}, 

379 ) 

380 registry.insertDimensionData( 

381 "day_obs", 

382 {"instrument": "Cam1", "id": 20211114}, 

383 ) 

384 # Choose timespans for exposures within the above calibration ranges 

385 # but make sure they are not identical to the full range. 

386 exp_time = astropy.time.TimeDelta(15.0, format="sec", scale="tai") 

387 span_delta = t2 - t1 

388 exp3_begin = t1 + (span_delta / 2.0) 

389 exp3_end = exp3_begin + exp_time 

390 span_delta = t3 - t2 

391 exp4_begin = t2 + (span_delta / 2.0) 

392 exp4_end = exp4_begin + exp_time 

393 registry.insertDimensionData( 

394 "exposure", 

395 { 

396 "instrument": "Cam1", 

397 "id": 3, 

398 "obs_id": "three", 

399 "timespan": Timespan(exp3_begin, exp3_end), 

400 "physical_filter": "Cam1-G", 

401 "group": "three", 

402 "day_obs": 20211114, 

403 "seq_num": 55, 

404 }, 

405 { 

406 "instrument": "Cam1", 

407 "id": 4, 

408 "obs_id": "four", 

409 "timespan": Timespan(exp4_begin, exp4_end), 

410 "physical_filter": "Cam1-G", 

411 "group": "four", 

412 "day_obs": 20211114, 

413 "seq_num": 42, 

414 }, 

415 ) 

416 # Get some biases from raw-like data IDs. 

417 bias2a_id, _ = butler.get( 

418 "bias", {"instrument": "Cam1", "exposure": 3, "detector": 2}, collections="calibs" 

419 ) 

420 self.assertEqual(bias2a_id, bias2a.id) 

421 bias3b_id, _ = butler.get( 

422 "bias", {"instrument": "Cam1", "exposure": 4, "detector": 3}, collections="calibs" 

423 ) 

424 self.assertEqual(bias3b_id, bias3b.id) 

425 

426 # Use explicit timespan and no exposure record. 

427 bias3b_id, _ = butler.get( 

428 "bias", 

429 {"instrument": "Cam1", "detector": 3}, 

430 collections="calibs", 

431 timespan=Timespan(exp4_begin, exp4_end), 

432 ) 

433 self.assertEqual(bias3b_id, bias3b.id) 

434 

435 # No timespan at all. 

436 # Only one matching dataset in calibs collection so this works with 

437 # a defaulted timespan. 

438 bias1a_id, _ = butler.get("bias", {"instrument": "Cam1", "detector": 1}, collections="calibs") 

439 self.assertEqual(bias1a_id, bias1a.id) 

440 

441 # Multiple datasets match in calibs collection with infinite timespan 

442 # so this fails. 

443 with self.assertRaises(LookupError): 

444 bias3b_id, _ = butler.get("bias", {"instrument": "Cam1", "detector": 3}, collections="calibs") 

445 

446 # Get using the kwarg form 

447 bias3b_id, _ = butler.get("bias", instrument="Cam1", exposure=4, detector=3, collections="calibs") 

448 self.assertEqual(bias3b_id, bias3b.id) 

449 

450 # Do it again but using the record information 

451 bias2a_id, _ = butler.get( 

452 "bias", 

453 {"instrument": "Cam1", "exposure.obs_id": "three", "detector.full_name": "Ab"}, 

454 collections="calibs", 

455 ) 

456 self.assertEqual(bias2a_id, bias2a.id) 

457 bias3b_id, _ = butler.get( 

458 "bias", 

459 {"exposure.obs_id": "four", "detector.full_name": "Ba"}, 

460 collections="calibs", 

461 instrument="Cam1", 

462 ) 

463 self.assertEqual(bias3b_id, bias3b.id) 

464 

465 # And again but this time using the alternate value rather than 

466 # the primary. 

467 bias3b_id, _ = butler.get( 

468 "bias", {"exposure": "four", "detector": "Ba"}, collections="calibs", instrument="Cam1" 

469 ) 

470 self.assertEqual(bias3b_id, bias3b.id) 

471 

472 # And again but this time using the alternate value rather than 

473 # the primary and do it in the keyword arguments. 

474 bias3b_id, _ = butler.get( 

475 "bias", exposure="four", detector="Ba", collections="calibs", instrument="Cam1" 

476 ) 

477 self.assertEqual(bias3b_id, bias3b.id) 

478 

479 # Now with implied record columns 

480 bias3b_id, _ = butler.get( 

481 "bias", 

482 day_obs=20211114, 

483 seq_num=42, 

484 raft="B", 

485 name_in_raft="a", 

486 collections="calibs", 

487 instrument="Cam1", 

488 ) 

489 self.assertEqual(bias3b_id, bias3b.id) 

490 

491 # Allow a fully-specified dataId and unnecessary extra information 

492 # that comes from the record. 

493 bias3b_id, _ = butler.get( 

494 "bias", 

495 dataId=dict( 

496 exposure=4, 

497 day_obs=20211114, 

498 seq_num=42, 

499 detector=3, 

500 instrument="Cam1", 

501 ), 

502 collections="calibs", 

503 ) 

504 self.assertEqual(bias3b_id, bias3b.id) 

505 

506 # Query for a calibration in a RUN and CALIBRATION collection to 

507 # ensure we do not get duplicate results. 

508 results = butler.query_datasets("bias", collections=["calibs", "imported_g"], find_first=False) 

509 self.assertEqual(len(set(results)), len(results)) 

510 

511 # Extra but inconsistent record values are a problem. 

512 with self.assertRaises(ValueError): 

513 bias3b_id, _ = butler.get( 

514 "bias", 

515 exposure=3, 

516 day_obs=20211114, 

517 seq_num=42, 

518 detector=3, 

519 collections="calibs", 

520 instrument="Cam1", 

521 ) 

522 

523 # Ensure that spurious kwargs cause an exception. 

524 with self.assertRaises(ValueError): 

525 butler.get( 

526 "bias", 

527 {"exposure.obs_id": "four", "immediate": True, "detector.full_name": "Ba"}, 

528 collections="calibs", 

529 instrument="Cam1", 

530 ) 

531 

532 with self.assertRaises(ValueError): 

533 butler.get( 

534 "bias", 

535 day_obs=20211114, 

536 seq_num=42, 

537 raft="B", 

538 name_in_raft="a", 

539 collections="calibs", 

540 instrument="Cam1", 

541 immediate=True, 

542 ) 

543 

544 def testRegistryDefaults(self): 

545 """Test that we can default the collections and some data ID keys when 

546 constructing a butler. 

547 

548 Many tests that use default run already exist in ``test_butler.py``, so 

549 that isn't tested here. And while most of this functionality is 

550 implemented in `Registry`, we test it here instead of 

551 ``daf/butler/tests/registry.py`` because it shouldn't depend on the 

552 database backend at all. 

553 """ 

554 butler = self.makeButler(writeable=True) 

555 butler.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

556 butler.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

557 # Need to actually set defaults later, not at construction, because 

558 # we need to import the instrument before we can use it as a default. 

559 # Don't set a default instrument value for data IDs, because 'Cam1' 

560 # should be inferred by virtue of that being the only value in the 

561 # input collections. 

562 butler.registry.defaults = RegistryDefaults(collections=["imported_g"]) 

563 # Use findDataset without collections or instrument. 

564 ref = butler.find_dataset("flat", detector=2, physical_filter="Cam1-G") 

565 # Do the same with Butler.get; this should ultimately invoke a lot of 

566 # the same code, so it's a bit circular, but mostly we're checking that 

567 # it works at all. 

568 dataset_id, _ = butler.get("flat", detector=2, physical_filter="Cam1-G") 

569 self.assertEqual(ref.id, dataset_id) 

570 # Query for datasets. Test defaulting the data ID in both kwargs and 

571 # in the WHERE expression. 

572 queried_refs_1 = set(butler.registry.queryDatasets("flat", detector=2, physical_filter="Cam1-G")) 

573 self.assertEqual({ref}, queried_refs_1) 

574 queried_refs_2 = set( 

575 butler.registry.queryDatasets("flat", where="detector=2 AND physical_filter='Cam1-G'") 

576 ) 

577 self.assertEqual({ref}, queried_refs_2) 

578 # Query for data IDs with a dataset constraint. 

579 queried_data_ids = set( 

580 butler.registry.queryDataIds( 

581 {"instrument", "detector", "physical_filter"}, 

582 datasets={"flat"}, 

583 detector=2, 

584 physical_filter="Cam1-G", 

585 ) 

586 ) 

587 self.assertEqual({ref.dataId}, queried_data_ids) 

588 # Add another instrument to the repo, and a dataset that uses it to 

589 # the `imported_g` collection. 

590 butler.registry.insertDimensionData("instrument", {"name": "Cam2"}) 

591 camera = DatasetType( 

592 "camera", 

593 dimensions=butler.dimensions["instrument"].minimal_group, 

594 storageClass="Camera", 

595 ) 

596 butler.registry.registerDatasetType(camera) 

597 butler.registry.insertDatasets(camera, [{"instrument": "Cam2"}], run="imported_g") 

598 # Initialize a new butler with `imported_g` as its default run. 

599 # This should not have a default instrument, because there are two. 

600 # Pass run instead of collections; this should set both. 

601 butler2 = Butler.from_config(butler=butler, run="imported_g") 

602 self.enterContext(butler2) 

603 self.assertEqual(list(butler2.registry.defaults.collections), ["imported_g"]) 

604 self.assertEqual(butler2.registry.defaults.run, "imported_g") 

605 self.assertFalse(butler2.registry.defaults.dataId) 

606 # Initialize a new butler with an instrument default explicitly given. 

607 # Set collections instead of run, which should then be None. 

608 butler3 = Butler.from_config(butler=butler, collections=["imported_g"], instrument="Cam2") 

609 self.enterContext(butler3) 

610 self.assertEqual(list(butler3.registry.defaults.collections), ["imported_g"]) 

611 self.assertIsNone(butler3.registry.defaults.run, None) 

612 self.assertEqual(butler3.registry.defaults.dataId.required, {"instrument": "Cam2"}) 

613 

614 # Check that repr() does not fail. 

615 defaults = RegistryDefaults(collections=["imported_g"], run="test") 

616 r = repr(defaults) 

617 self.assertIn("collections=('imported_g',)", r) 

618 self.assertIn("run='test'", r) 

619 

620 defaults = RegistryDefaults(run="test", instrument="DummyCam", skypix="pix") 

621 r = repr(defaults) 

622 self.assertIn("skypix='pix'", r) 

623 self.assertIn("instrument='DummyCam'", r) 

624 

625 def testJson(self): 

626 """Test JSON serialization mediated by registry.""" 

627 butler = self.makeButler(writeable=True) 

628 butler.import_(filename="resource://lsst.daf.butler/tests/registry_data/base.yaml") 

629 butler.import_(filename=f"resource://lsst.daf.butler/tests/registry_data/{self.datasetsImportFile}") 

630 # Need to actually set defaults later, not at construction, because 

631 # we need to import the instrument before we can use it as a default. 

632 # Don't set a default instrument value for data IDs, because 'Cam1' 

633 # should be inferred by virtue of that being the only value in the 

634 # input collections. 

635 butler.registry.defaults = RegistryDefaults(collections=["imported_g"]) 

636 # Use findDataset without collections or instrument. 

637 ref = butler.find_dataset("flat", detector=2, physical_filter="Cam1-G") 

638 

639 # Transform the ref and dataset type to and from JSON 

640 # and check that it can be reconstructed properly 

641 

642 # Do it with the ref and a component ref in minimal and standard form 

643 compRef = ref.makeComponentRef("wcs") 

644 

645 for test_item in (ref, ref.datasetType, compRef, compRef.datasetType): 

646 for minimal in (False, True): 

647 json_str = test_item.to_json(minimal=minimal) 

648 from_json = type(test_item).from_json(json_str, registry=butler.registry) 

649 self.assertEqual(from_json, test_item, msg=f"From JSON '{json_str}' using registry") 

650 

651 # for minimal=False case also do a test without registry 

652 if not minimal: 

653 from_json = type(test_item).from_json(json_str, universe=butler.dimensions) 

654 self.assertEqual(from_json, test_item, msg=f"From JSON '{json_str}' using universe") 

655 

656 def test_populated_by(self): 

657 """Test that dimension records can find other records.""" 

658 butler = self.makeButler(writeable=True) 

659 butler.import_(filename=os.path.join(TESTDIR, "data", "registry", "hsc-rc2-subset.yaml")) 

660 

661 elements = frozenset(element for element in butler.dimensions.elements if element.has_own_table) 

662 

663 # Get a visit-based dataId. 

664 data_ids = set(butler.registry.queryDataIds("visit", visit=1232, instrument="HSC")) 

665 

666 # Request all the records related to it. 

667 records = butler._extract_all_dimension_records_from_data_ids(butler, data_ids, elements) 

668 

669 self.assertIn(butler.dimensions["visit_detector_region"], records, f"Keys: {records.keys()}") 

670 self.assertIn(butler.dimensions["visit_system_membership"], records) 

671 self.assertIn(butler.dimensions["visit_system"], records) 

672 

673 def testJsonDimensionRecordsAndHtmlRepresentation(self): 

674 # Dimension Records 

675 butler = self.makeButler(writeable=True) 

676 butler.import_(filename=os.path.join(TESTDIR, "data", "registry", "hsc-rc2-subset.yaml")) 

677 

678 for dimension in ("detector", "visit", "exposure", "day_obs", "group"): 

679 records = butler.registry.queryDimensionRecords(dimension, instrument="HSC") 

680 for r in records: 

681 for minimal in (True, False): 

682 json_str = r.to_json(minimal=minimal) 

683 r_json = type(r).from_json(json_str, registry=butler.registry) 

684 self.assertEqual(r_json, r) 

685 # check with direct method 

686 simple = r.to_simple() 

687 fromDirect = type(simple).direct(**json.loads(json_str)) 

688 self.assertEqual(simple, fromDirect) 

689 # Also check equality of each of the components as dicts 

690 self.assertEqual(r_json.toDict(), r.toDict())