Coverage for tests / test_templates.py: 11%
187 statements
« prev ^ index » next coverage.py v7.13.5, created at 2026-04-17 08:49 +0000
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/>.
28"""Test file name templating."""
30import os.path
31import unittest
32import uuid
34from lsst.daf.butler import (
35 DataCoordinate,
36 DatasetId,
37 DatasetRef,
38 DatasetType,
39 DimensionUniverse,
40 StorageClass,
41)
42from lsst.daf.butler.datastore.file_templates import (
43 FileTemplate,
44 FileTemplates,
45 FileTemplatesConfig,
46 FileTemplateValidationError,
47)
49TESTDIR = os.path.abspath(os.path.dirname(__file__))
51PlaceHolder = StorageClass("PlaceHolder")
53REFUUID = DatasetId(int=uuid.uuid4().int)
56class TestFileTemplates(unittest.TestCase):
57 """Test creation of paths from templates."""
59 def makeDatasetRef(
60 self, datasetTypeName, dataId=None, storageClassName="DefaultStorageClass", run="run2", conform=True
61 ):
62 """Make a simple DatasetRef"""
63 if dataId is None:
64 dataId = self.dataId
65 if "physical_filter" in dataId and "band" not in dataId:
66 dataId["band"] = "b" # Add fake band.
67 dimensions = self.universe.conform(dataId.keys())
68 dataId = DataCoordinate.standardize(dataId, dimensions=dimensions)
70 # Pretend we have a parent if this looks like a composite
71 compositeName, componentName = DatasetType.splitDatasetTypeName(datasetTypeName)
72 parentStorageClass = PlaceHolder if componentName else None
74 datasetType = DatasetType(
75 datasetTypeName,
76 dimensions,
77 StorageClass(storageClassName),
78 parentStorageClass=parentStorageClass,
79 )
80 return DatasetRef(datasetType, dataId, id=REFUUID, run=run, conform=conform)
82 def setUp(self):
83 self.universe = DimensionUniverse()
84 self.dataId = {
85 "instrument": "dummy",
86 "visit": 52,
87 "physical_filter": "Most Amazing U Filter Ever",
88 "day_obs": 20200101,
89 }
91 def assertTemplate(self, template, answer, ref):
92 fileTmpl = FileTemplate(template)
93 path = fileTmpl.format(ref)
94 self.assertEqual(path, answer)
96 def testBasic(self):
97 tmplstr = "{run}/{datasetType}/{visit:05d}/{physical_filter}"
98 self.assertTemplate(
99 tmplstr,
100 "run2/calexp/00052/Most_Amazing_U_Filter_Ever",
101 self.makeDatasetRef("calexp"),
102 )
103 tmplstr = "{run}/{datasetType}/{visit:05d}/{physical_filter}-trail"
104 self.assertTemplate(
105 tmplstr,
106 "run2/calexp/00052/Most_Amazing_U_Filter_Ever-trail",
107 self.makeDatasetRef("calexp"),
108 )
110 tmplstr = "{run}/{datasetType}/{visit:05d}/{physical_filter}-trail-{run}"
111 self.assertTemplate(
112 tmplstr,
113 "run2/calexp/00052/Most_Amazing_U_Filter_Ever-trail-run2",
114 self.makeDatasetRef("calexp"),
115 )
116 self.assertTemplate(
117 tmplstr,
118 "run_2/calexp/00052/Most_Amazing_U_Filter_Ever-trail-run_2",
119 self.makeDatasetRef("calexp", run="run/2"),
120 )
122 # Check that the id is sufficient without any other information.
123 self.assertTemplate("{id}", str(REFUUID), self.makeDatasetRef("calexp", run="run2"))
125 self.assertTemplate("{run}/{id}", f"run2/{str(REFUUID)}", self.makeDatasetRef("calexp", run="run2"))
127 self.assertTemplate(
128 "fixed/{id}",
129 f"fixed/{str(REFUUID)}",
130 self.makeDatasetRef("calexp", run="run2"),
131 )
133 self.assertTemplate(
134 "fixed/{id}_{physical_filter}",
135 f"fixed/{str(REFUUID)}_Most_Amazing_U_Filter_Ever",
136 self.makeDatasetRef("calexp", run="run2"),
137 )
139 # Retain any "/" in run
140 tmplstr = "{run:/}/{datasetType}/{visit:05d}/{physical_filter}-trail-{run}"
141 self.assertTemplate(
142 tmplstr,
143 "run/2/calexp/00052/Most_Amazing_U_Filter_Ever-trail-run_2",
144 self.makeDatasetRef("calexp", run="run/2"),
145 )
147 # Check that "." are replaced in the file basename, but not directory.
148 dataId = {"instrument": "dummy", "visit": 52, "physical_filter": "g.10", "day_obs": 20250101}
149 self.assertTemplate(
150 tmplstr,
151 "run.2/calexp/00052/g_10-trail-run_2",
152 self.makeDatasetRef("calexp", run="run.2", dataId=dataId),
153 )
155 # Check that type conversion is applied
156 dataId = {"instrument": "dummy", "day_obs": 20250203}
157 tmplstr = "{run}/{datasetType}/{day_obs!s:.4s}"
158 self.assertTemplate(
159 tmplstr,
160 "run3/calexp/2025",
161 self.makeDatasetRef("calexp", run="run3", dataId=dataId),
162 )
164 with self.assertRaises(FileTemplateValidationError):
165 FileTemplate("no fields at all")
167 with self.assertRaises(FileTemplateValidationError):
168 FileTemplate("{visit}")
170 with self.assertRaises(FileTemplateValidationError):
171 FileTemplate("{run}_{datasetType}")
173 with self.assertRaises(FileTemplateValidationError):
174 FileTemplate("{id}/fixed")
176 with self.assertRaises(FileTemplateValidationError):
177 FileTemplate("{run}/../{datasetType}_{visit}")
179 def testAlternates(self):
180 tmplstr = "{run}/{datasetType}/{visit:05d}/{physical_filter|day_obs}_{day_obs|physical_filter}"
181 self.assertTemplate(
182 tmplstr,
183 "run2/calexp/00052/Most_Amazing_U_Filter_Ever_20200101",
184 self.makeDatasetRef("calexp"),
185 )
186 tmplstr = "{run}/{datasetType}/{exposure|visit:05d}/{physical_filter|day_obs}_{group|exposure:?}"
187 self.assertTemplate(
188 tmplstr,
189 "run2/calexp/00052/Most_Amazing_U_Filter_Ever",
190 self.makeDatasetRef("calexp"),
191 )
193 def testRunOrCollectionNeeded(self):
194 tmplstr = "{datasetType}/{visit:05d}/{physical_filter}"
195 with self.assertRaises(FileTemplateValidationError):
196 self.assertTemplate(tmplstr, "run2/calexp/00052/U", self.makeDatasetRef("calexp"))
198 def testNoRecord(self):
199 # Attaching records is not possible in this test code but we can check
200 # that a missing record when a metadata entry has been requested
201 # does fail.
202 tmplstr = "{run}/{datasetType}/{visit.name}/{physical_filter}"
203 with self.assertRaises(RuntimeError) as cm:
204 self.assertTemplate(tmplstr, "", self.makeDatasetRef("calexp"))
205 self.assertIn("No metadata", str(cm.exception))
207 def testOptional(self):
208 """Optional units in templates."""
209 ref = self.makeDatasetRef("calexp")
210 tmplstr = "{run}/{datasetType}/v{visit:05d}_f{physical_filter:?}_{skypix:?}"
211 self.assertTemplate(
212 tmplstr,
213 "run2/calexp/v00052_fMost_Amazing_U_Filter_Ever",
214 self.makeDatasetRef("calexp"),
215 )
217 du = {"visit": 48, "tract": 265, "skymap": "big", "instrument": "dummy", "htm7": 12345}
218 self.assertTemplate(tmplstr, "run2/calexpT/v00048_12345", self.makeDatasetRef("calexpT", du))
220 # Ensure that this returns a relative path even if the first field
221 # is optional
222 tmplstr = "{run}/{tract:?}/{visit:?}/f{physical_filter}"
223 self.assertTemplate(tmplstr, "run2/52/fMost_Amazing_U_Filter_Ever", ref)
225 # Ensure that // from optionals are converted to singles
226 tmplstr = "{run}/{datasetType}/{patch:?}/{tract:?}/f{physical_filter}"
227 self.assertTemplate(tmplstr, "run2/calexp/fMost_Amazing_U_Filter_Ever", ref)
229 # Optionals with some text between fields
230 tmplstr = "{run}/{datasetType}/p{patch:?}_t{tract:?}/f{physical_filter}"
231 self.assertTemplate(tmplstr, "run2/calexp/p/fMost_Amazing_U_Filter_Ever", ref)
232 tmplstr = "{run}/{datasetType}/p{patch:?}_t{visit:04d?}/f{physical_filter}"
233 self.assertTemplate(tmplstr, "run2/calexp/p_t0052/fMost_Amazing_U_Filter_Ever", ref)
235 def testComponent(self):
236 """Test handling of components in templates."""
237 refMetricOutput = self.makeDatasetRef("metric.output")
238 refMetric = self.makeDatasetRef("metric")
239 refMaskedImage = self.makeDatasetRef("calexp.maskedimage.variance")
240 refWcs = self.makeDatasetRef("calexp.wcs")
242 tmplstr = "{run}_c_{component}_v{visit}"
243 self.assertTemplate(tmplstr, "run2_c_output_v52", refMetricOutput)
245 # We want this template to have both a directory and basename, to
246 # test that the right parts of the output are replaced.
247 tmplstr = "{component:?}/{run}_{component:?}_{visit}"
248 self.assertTemplate(tmplstr, "run2_52", refMetric)
249 self.assertTemplate(tmplstr, "output/run2_output_52", refMetricOutput)
250 self.assertTemplate(tmplstr, "maskedimage.variance/run2_maskedimage_variance_52", refMaskedImage)
251 self.assertTemplate(tmplstr, "output/run2_output_52", refMetricOutput)
253 # Providing a component but not using it
254 tmplstr = "{run}/{datasetType}/v{visit:05d}"
255 with self.assertRaises(KeyError):
256 self.assertTemplate(tmplstr, "", refWcs)
258 def testFields(self):
259 # Template, mandatory fields, optional non-special fields,
260 # special fields, optional special fields
261 testData = (
262 (
263 "{run}/{datasetType}/{visit:05d}/{physical_filter}-trail",
264 {"visit", "physical_filter"},
265 set(),
266 {"run", "datasetType"},
267 set(),
268 ),
269 (
270 "{run}/{component:?}_{visit}",
271 {"visit"},
272 set(),
273 {"run"},
274 {"component"},
275 ),
276 (
277 "{run}/{component:?}_{visit:?}_{physical_filter}_{instrument}_{datasetType}",
278 {"physical_filter", "instrument"},
279 {"visit"},
280 {"run", "datasetType"},
281 {"component"},
282 ),
283 )
284 for tmplstr, mandatory, optional, special, optionalSpecial in testData:
285 with self.subTest(template=tmplstr):
286 tmpl = FileTemplate(tmplstr)
287 fields = tmpl.fields()
288 self.assertEqual(fields, mandatory)
289 fields = tmpl.fields(optionals=True)
290 self.assertEqual(fields, mandatory | optional)
291 fields = tmpl.fields(specials=True)
292 self.assertEqual(fields, mandatory | special)
293 fields = tmpl.fields(specials=True, optionals=True)
294 self.assertEqual(fields, mandatory | special | optional | optionalSpecial)
296 def testSimpleConfig(self):
297 """Test reading from config file"""
298 configRoot = os.path.join(TESTDIR, "config", "templates")
299 config1 = FileTemplatesConfig(os.path.join(configRoot, "templates-nodefault.yaml"))
300 templates = FileTemplates(config1, universe=self.universe)
301 ref = self.makeDatasetRef("calexp")
302 tmpl = templates.getTemplate(ref)
303 self.assertIsInstance(tmpl, FileTemplate)
305 # This config file should not allow defaulting
306 ref2 = self.makeDatasetRef("unknown")
307 with self.assertRaises(KeyError):
308 templates.getTemplate(ref2)
310 # This should fall through the datasetTypeName check and use
311 # StorageClass instead
312 ref3 = self.makeDatasetRef("unknown2", storageClassName="StorageClassX")
313 tmplSc = templates.getTemplate(ref3)
314 self.assertIsInstance(tmplSc, FileTemplate)
316 # Try with a component: one with defined formatter and one without
317 refWcs = self.makeDatasetRef("calexp.wcs")
318 refImage = self.makeDatasetRef("calexp.image")
319 tmplCalexp = templates.getTemplate(ref)
320 tmplWcs = templates.getTemplate(refWcs) # Should be special
321 tmpl_image = templates.getTemplate(refImage)
322 self.assertIsInstance(tmplCalexp, FileTemplate)
323 self.assertIsInstance(tmpl_image, FileTemplate)
324 self.assertIsInstance(tmplWcs, FileTemplate)
325 self.assertEqual(tmplCalexp, tmpl_image)
326 self.assertNotEqual(tmplCalexp, tmplWcs)
328 # Check dimensions lookup order.
329 # The order should be: dataset type name, dimension, storage class
330 # This one will not match name but might match storage class.
331 # It should match dimensions
332 refDims = self.makeDatasetRef(
333 "nomatch", dataId={"instrument": "LSST", "physical_filter": "z"}, storageClassName="StorageClassX"
334 )
335 tmplDims = templates.getTemplate(refDims)
336 self.assertIsInstance(tmplDims, FileTemplate)
337 self.assertNotEqual(tmplDims, tmplSc)
339 # Test that instrument overrides retrieve specialist templates
340 refPvi = self.makeDatasetRef("pvi")
341 refPviHsc = self.makeDatasetRef("pvi", dataId={"instrument": "HSC", "physical_filter": "z"})
342 refPviLsst = self.makeDatasetRef("pvi", dataId={"instrument": "LSST", "physical_filter": "z"})
344 tmplPvi = templates.getTemplate(refPvi)
345 tmplPviHsc = templates.getTemplate(refPviHsc)
346 tmplPviLsst = templates.getTemplate(refPviLsst)
347 self.assertEqual(tmplPvi, tmplPviLsst)
348 self.assertNotEqual(tmplPvi, tmplPviHsc)
350 # Have instrument match and dimensions look up with no name match
351 refNoPviHsc = self.makeDatasetRef(
352 "pvix", dataId={"instrument": "HSC", "physical_filter": "z"}, storageClassName="StorageClassX"
353 )
354 tmplNoPviHsc = templates.getTemplate(refNoPviHsc)
355 self.assertNotEqual(tmplNoPviHsc, tmplDims)
356 self.assertNotEqual(tmplNoPviHsc, tmplPviHsc)
358 # Format config file with defaulting
359 config2 = FileTemplatesConfig(os.path.join(configRoot, "templates-withdefault.yaml"))
360 templates = FileTemplates(config2, universe=self.universe)
361 tmpl = templates.getTemplate(ref2)
362 self.assertIsInstance(tmpl, FileTemplate)
364 # Format config file with bad format string
365 with self.assertRaises(FileTemplateValidationError):
366 FileTemplates(os.path.join(configRoot, "templates-bad.yaml"), universe=self.universe)
368 # Config file with no defaulting mentioned
369 config3 = os.path.join(configRoot, "templates-nodefault2.yaml")
370 templates = FileTemplates(config3, universe=self.universe)
371 with self.assertRaises(KeyError):
372 templates.getTemplate(ref2)
374 # Try again but specify a default in the constructor
375 default = "{run}/{datasetType}/{physical_filter}"
376 templates = FileTemplates(config3, default=default, universe=self.universe)
377 tmpl = templates.getTemplate(ref2)
378 self.assertEqual(tmpl.template, default)
380 def testValidation(self):
381 configRoot = os.path.join(TESTDIR, "config", "templates")
382 config1 = FileTemplatesConfig(os.path.join(configRoot, "templates-nodefault.yaml"))
383 templates = FileTemplates(config1, universe=self.universe)
385 entities = {}
386 entities["calexp"] = self.makeDatasetRef(
387 "calexp",
388 storageClassName="StorageClassX",
389 dataId={"instrument": "dummy", "physical_filter": "i", "visit": 52},
390 )
392 with self.assertLogs(level="WARNING") as cm:
393 templates.validateTemplates(entities.values(), logFailures=True)
394 self.assertIn("Unchecked keys", cm.output[0])
395 self.assertIn("StorageClassX", cm.output[0])
397 entities["pvi"] = self.makeDatasetRef(
398 "pvi", storageClassName="StorageClassX", dataId={"instrument": "dummy", "physical_filter": "i"}
399 )
400 entities["StorageClassX"] = self.makeDatasetRef(
401 "storageClass", storageClassName="StorageClassX", dataId={"instrument": "dummy", "visit": 2}
402 )
403 entities["calexp.wcs"] = self.makeDatasetRef(
404 "calexp.wcs",
405 storageClassName="StorageClassX",
406 dataId={"instrument": "dummy", "physical_filter": "i", "visit": 23},
407 conform=False,
408 )
410 entities["instrument+physical_filter"] = self.makeDatasetRef(
411 "filter_inst",
412 storageClassName="StorageClassX",
413 dataId={"physical_filter": "i", "instrument": "SCUBA"},
414 )
415 entities["hsc+pvi"] = self.makeDatasetRef(
416 "pvi", storageClassName="StorageClassX", dataId={"physical_filter": "i", "instrument": "HSC"}
417 )
419 entities["hsc+instrument+physical_filter"] = self.makeDatasetRef(
420 "filter_inst",
421 storageClassName="StorageClassX",
422 dataId={"physical_filter": "i", "instrument": "HSC"},
423 )
425 entities["metric6"] = self.makeDatasetRef(
426 "filter_inst",
427 storageClassName="Integer",
428 dataId={"physical_filter": "i", "instrument": "HSC"},
429 )
431 templates.validateTemplates(entities.values(), logFailures=True)
433 # Rerun but with a failure
434 entities["pvi"] = self.makeDatasetRef("pvi", storageClassName="StorageClassX", dataId={"band": "i"})
435 with self.assertRaises(FileTemplateValidationError):
436 with self.assertLogs(level="FATAL"):
437 templates.validateTemplates(entities.values(), logFailures=True)
440if __name__ == "__main__":
441 unittest.main()