Coverage for python/lsst/daf/butler/transfers/_yaml.py : 11%

Hot-keys on this page
r m x p toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
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 program is free software: you can redistribute it and/or modify
10# it under the terms of the GNU General Public License as published by
11# the Free Software Foundation, either version 3 of the License, or
12# (at your option) any later version.
13#
14# This program is distributed in the hope that it will be useful,
15# but WITHOUT ANY WARRANTY; without even the implied warranty of
16# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17# GNU General Public License for more details.
18#
19# You should have received a copy of the GNU General Public License
20# along with this program. If not, see <http://www.gnu.org/licenses/>.
22from __future__ import annotations
24__all__ = ["YamlRepoExportBackend", "YamlRepoImportBackend"]
26import os
27from datetime import datetime
28from typing import (
29 Any,
30 Dict,
31 IO,
32 Iterable,
33 List,
34 Mapping,
35 Optional,
36 Set,
37 Tuple,
38 Type,
39)
40import warnings
41from collections import defaultdict
43import yaml
44import astropy.time
46from lsst.utils import doImport
47from ..core import (
48 DatasetAssociation,
49 DatasetRef,
50 DatasetType,
51 DataCoordinate,
52 Datastore,
53 DimensionElement,
54 DimensionRecord,
55 FileDataset,
56 Timespan,
57)
58from ..core.utils import iterable
59from ..core.named import NamedValueSet
60from ..registry import CollectionType, Registry
61from ..registry.interfaces import (
62 ChainedCollectionRecord,
63 CollectionRecord,
64 RunRecord,
65 VersionTuple,
66)
67from ..registry.versions import IncompatibleVersionError
68from ._interfaces import RepoExportBackend, RepoImportBackend
71EXPORT_FORMAT_VERSION = VersionTuple(1, 0, 0)
72"""Export format version.
74Files with a different major version or a newer minor version cannot be read by
75this version of the code.
76"""
79class YamlRepoExportBackend(RepoExportBackend):
80 """A repository export implementation that saves to a YAML file.
82 Parameters
83 ----------
84 stream
85 A writeable file-like object.
86 """
88 def __init__(self, stream: IO):
89 self.stream = stream
90 self.data: List[Dict[str, Any]] = []
92 def saveDimensionData(self, element: DimensionElement, *data: DimensionRecord) -> None:
93 # Docstring inherited from RepoExportBackend.saveDimensionData.
94 data_dicts = [record.toDict(splitTimespan=True) for record in data]
95 self.data.append({
96 "type": "dimension",
97 "element": element.name,
98 "records": data_dicts,
99 })
101 def saveCollection(self, record: CollectionRecord) -> None:
102 # Docstring inherited from RepoExportBackend.saveCollections.
103 data: Dict[str, Any] = {
104 "type": "collection",
105 "collection_type": record.type.name,
106 "name": record.name,
107 }
108 if isinstance(record, RunRecord):
109 data["host"] = record.host
110 data["timespan_begin"] = record.timespan.begin
111 data["timespan_end"] = record.timespan.end
112 elif isinstance(record, ChainedCollectionRecord):
113 data["children"] = list(record.children)
114 self.data.append(data)
116 def saveDatasets(self, datasetType: DatasetType, run: str, *datasets: FileDataset) -> None:
117 # Docstring inherited from RepoExportBackend.saveDatasets.
118 self.data.append({
119 "type": "dataset_type",
120 "name": datasetType.name,
121 "dimensions": [d.name for d in datasetType.dimensions],
122 "storage_class": datasetType.storageClass.name,
123 "is_calibration": datasetType.isCalibration(),
124 })
125 self.data.append({
126 "type": "dataset",
127 "dataset_type": datasetType.name,
128 "run": run,
129 "records": [
130 {
131 "dataset_id": [ref.id for ref in sorted(dataset.refs)],
132 "data_id": [ref.dataId.byName() for ref in sorted(dataset.refs)],
133 "path": dataset.path,
134 "formatter": dataset.formatter,
135 # TODO: look up and save other collections
136 }
137 for dataset in datasets
138 ]
139 })
141 def saveDatasetAssociations(self, collection: str, collectionType: CollectionType,
142 associations: Iterable[DatasetAssociation]) -> None:
143 # Docstring inherited from RepoExportBackend.saveDatasetAssociations.
144 if collectionType is CollectionType.TAGGED:
145 self.data.append({
146 "type": "associations",
147 "collection": collection,
148 "collection_type": collectionType.name,
149 "dataset_ids": [assoc.ref.id for assoc in associations],
150 })
151 elif collectionType is CollectionType.CALIBRATION:
152 idsByTimespan: Dict[Timespan, List[int]] = defaultdict(list)
153 for association in associations:
154 assert association.timespan is not None
155 assert association.ref.id is not None
156 idsByTimespan[association.timespan].append(association.ref.id)
157 self.data.append({
158 "type": "associations",
159 "collection": collection,
160 "collection_type": collectionType.name,
161 "validity_ranges": [
162 {
163 "begin": timespan.begin,
164 "end": timespan.end,
165 "dataset_ids": dataset_ids,
166 }
167 for timespan, dataset_ids in idsByTimespan.items()
168 ]
169 })
171 def finish(self) -> None:
172 # Docstring inherited from RepoExportBackend.
173 yaml.dump(
174 {
175 "description": "Butler Data Repository Export",
176 "version": str(EXPORT_FORMAT_VERSION),
177 "data": self.data,
178 },
179 stream=self.stream,
180 sort_keys=False,
181 )
184class YamlRepoImportBackend(RepoImportBackend):
185 """A repository import implementation that reads from a YAML file.
187 Parameters
188 ----------
189 stream
190 A readable file-like object.
191 registry : `Registry`
192 The registry datasets will be imported into. Only used to retreive
193 dataset types during construction; all write happen in `register`
194 and `load`.
195 """
197 def __init__(self, stream: IO, registry: Registry):
198 # We read the file fully and convert its contents to Python objects
199 # instead of loading incrementally so we can spot some problems early;
200 # because `register` can't be put inside a transaction, we'd rather not
201 # run that at all if there's going to be problem later in `load`.
202 wrapper = yaml.safe_load(stream)
203 if wrapper["version"] == 0:
204 # Grandfather-in 'version: 0' -> 1.0.0, which is what we wrote
205 # before we really tried to do versioning here.
206 fileVersion = VersionTuple(1, 0, 0)
207 else:
208 fileVersion = VersionTuple.fromString(wrapper["version"])
209 if fileVersion.major != EXPORT_FORMAT_VERSION.major:
210 raise IncompatibleVersionError(
211 f"Cannot read repository export file with version={fileVersion} "
212 f"({EXPORT_FORMAT_VERSION.major}.x.x required)."
213 )
214 if fileVersion.minor > EXPORT_FORMAT_VERSION.minor:
215 raise IncompatibleVersionError(
216 f"Cannot read repository export file with version={fileVersion} "
217 f"< {EXPORT_FORMAT_VERSION.major}.{EXPORT_FORMAT_VERSION.minor}.x required."
218 )
219 self.runs: Dict[str, Tuple[Optional[str], Timespan]] = {}
220 self.chains: Dict[str, List[str]] = {}
221 self.collections: Dict[str, CollectionType] = {}
222 self.datasetTypes: NamedValueSet[DatasetType] = NamedValueSet()
223 self.dimensions: Mapping[DimensionElement, List[DimensionRecord]] = defaultdict(list)
224 self.tagAssociations: Dict[str, List[int]] = defaultdict(list)
225 self.calibAssociations: Dict[str, Dict[Timespan, List[int]]] = defaultdict(dict)
226 self.refsByFileId: Dict[int, DatasetRef] = {}
227 self.registry: Registry = registry
228 datasetData = []
229 for data in wrapper["data"]:
230 if data["type"] == "dimension":
231 # convert all datetime values to astropy
232 for record in data["records"]:
233 for key in record:
234 # Some older YAML files were produced with native
235 # YAML support for datetime, we support reading that
236 # data back. Newer conversion uses _AstropyTimeToYAML
237 # class with special YAML tag.
238 if isinstance(record[key], datetime):
239 record[key] = astropy.time.Time(record[key], scale="utc")
240 element = self.registry.dimensions[data["element"]]
241 RecordClass: Type[DimensionRecord] = element.RecordClass
242 self.dimensions[element].extend(
243 RecordClass(**r) for r in data["records"]
244 )
245 elif data["type"] == "collection":
246 collectionType = CollectionType.__members__[data["collection_type"].upper()]
247 if collectionType is CollectionType.RUN:
248 self.runs[data["name"]] = (
249 data["host"],
250 Timespan(begin=data["timespan_begin"], end=data["timespan_end"])
251 )
252 elif collectionType is CollectionType.CHAINED:
253 children = []
254 for child in data["children"]:
255 if not isinstance(child, str):
256 warnings.warn(
257 f"CHAINED collection {data['name']} includes restrictions on child "
258 "collection searches, which are no longer suppored and will be ignored."
259 )
260 # Old form with dataset type restrictions only,
261 # supported for backwards compatibility.
262 child, _ = child
263 children.append(child)
264 self.chains[data["name"]] = children
265 else:
266 self.collections[data["name"]] = collectionType
267 elif data["type"] == "run":
268 # Also support old form of saving a run with no extra info.
269 self.runs[data["name"]] = (None, Timespan(None, None))
270 elif data["type"] == "dataset_type":
271 self.datasetTypes.add(
272 DatasetType(data["name"], dimensions=data["dimensions"],
273 storageClass=data["storage_class"], universe=self.registry.dimensions,
274 isCalibration=data.get("is_calibration", False))
275 )
276 elif data["type"] == "dataset":
277 # Save raw dataset data for a second loop, so we can ensure we
278 # know about all dataset types first.
279 datasetData.append(data)
280 elif data["type"] == "associations":
281 collectionType = CollectionType.__members__[data["collection_type"].upper()]
282 if collectionType is CollectionType.TAGGED:
283 self.tagAssociations[data["collection"]].extend(data["dataset_ids"])
284 elif collectionType is CollectionType.CALIBRATION:
285 assocsByTimespan = self.calibAssociations[data["collection"]]
286 for d in data["validity_ranges"]:
287 assocsByTimespan[Timespan(begin=d["begin"], end=d["end"])] = d["dataset_ids"]
288 else:
289 raise ValueError(f"Unexpected calibration type for association: {collectionType.name}.")
290 else:
291 raise ValueError(f"Unexpected dictionary type: {data['type']}.")
292 # key is (dataset type name, run)
293 self.datasets: Mapping[Tuple[str, str], List[FileDataset]] = defaultdict(list)
294 for data in datasetData:
295 datasetType = self.datasetTypes.get(data["dataset_type"])
296 if datasetType is None:
297 datasetType = self.registry.getDatasetType(data["dataset_type"])
298 self.datasets[data["dataset_type"], data["run"]].extend(
299 FileDataset(
300 d.get("path"),
301 [DatasetRef(datasetType, dataId, run=data["run"], id=refid)
302 for dataId, refid in zip(iterable(d["data_id"]), iterable(d["dataset_id"]))],
303 formatter=doImport(d.get("formatter")) if "formatter" in d else None
304 )
305 for d in data["records"]
306 )
308 def register(self) -> None:
309 # Docstring inherited from RepoImportBackend.register.
310 for datasetType in self.datasetTypes:
311 self.registry.registerDatasetType(datasetType)
312 for run in self.runs:
313 self.registry.registerRun(run)
314 # No way to add extra run info to registry yet.
315 for collection, collection_type in self.collections.items():
316 self.registry.registerCollection(collection, collection_type)
317 for chain, children in self.chains.items():
318 self.registry.registerCollection(chain, CollectionType.CHAINED)
319 self.registry.setCollectionChain(chain, children)
321 def load(self, datastore: Optional[Datastore], *,
322 directory: Optional[str] = None, transfer: Optional[str] = None,
323 skip_dimensions: Optional[Set] = None) -> None:
324 # Docstring inherited from RepoImportBackend.load.
325 for element, dimensionRecords in self.dimensions.items():
326 if skip_dimensions and element in skip_dimensions:
327 continue
328 self.registry.insertDimensionData(element, *dimensionRecords)
329 # FileDatasets to ingest into the datastore (in bulk):
330 fileDatasets = []
331 for (datasetTypeName, run), records in self.datasets.items():
332 datasetType = self.registry.getDatasetType(datasetTypeName)
333 # Make a big flattened list of all data IDs and dataset_ids, while
334 # remembering slices that associate them with the FileDataset
335 # instances they came from.
336 dataIds: List[DataCoordinate] = []
337 dataset_ids: List[int] = []
338 slices = []
339 for fileDataset in records:
340 start = len(dataIds)
341 dataIds.extend(ref.dataId for ref in fileDataset.refs)
342 dataset_ids.extend(ref.id for ref in fileDataset.refs) # type: ignore
343 stop = len(dataIds)
344 slices.append(slice(start, stop))
345 # Insert all of those DatasetRefs at once.
346 # For now, we ignore the dataset_id we pulled from the file
347 # and just insert without one to get a new autoincrement value.
348 # Eventually (once we have origin in IDs) we'll preserve them.
349 resolvedRefs = self.registry.insertDatasets(
350 datasetType,
351 dataIds=dataIds,
352 run=run,
353 )
354 # Populate our dictionary that maps int dataset_id values from the
355 # export file to the new DatasetRefs
356 for fileId, ref in zip(dataset_ids, resolvedRefs):
357 self.refsByFileId[fileId] = ref
358 # Now iterate over the original records, and install the new
359 # resolved DatasetRefs to replace the unresolved ones as we
360 # reorganize the collection information.
361 for sliceForFileDataset, fileDataset in zip(slices, records):
362 fileDataset.refs = resolvedRefs[sliceForFileDataset]
363 if directory is not None:
364 fileDataset.path = os.path.join(directory, fileDataset.path)
365 fileDatasets.append(fileDataset)
366 # Ingest everything into the datastore at once.
367 if datastore is not None and fileDatasets:
368 datastore.ingest(*fileDatasets, transfer=transfer)
369 # Associate datasets with tagged collections.
370 for collection, dataset_ids in self.tagAssociations.items():
371 self.registry.associate(collection, [self.refsByFileId[i] for i in dataset_ids])
372 # Associate datasets with calibration collections.
373 for collection, idsByTimespan in self.calibAssociations.items():
374 for timespan, dataset_ids in idsByTimespan.items():
375 self.registry.certify(collection, [self.refsByFileId[i] for i in dataset_ids], timespan)