Coverage for python / lsst / daf / butler / registry / interfaces / _bridge.py: 81%
72 statements
« prev ^ index » next coverage.py v7.13.2, created at 2026-02-03 01:20 -0800
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/>.
27from __future__ import annotations
29__all__ = ("DatasetIdRef", "DatastoreRegistryBridge", "DatastoreRegistryBridgeManager", "FakeDatasetRef")
31from abc import ABC, abstractmethod
32from collections.abc import Collection, Iterable
33from contextlib import AbstractContextManager
34from typing import TYPE_CHECKING, Any
36from lsst.utils.classes import immutable
38from ..._dataset_ref import DatasetId, DatasetRef
39from ._versioning import VersionedExtension, VersionTuple
41if TYPE_CHECKING:
42 from ..._dataset_ref import DatasetDatastoreRecords
43 from ..._dataset_type import DatasetType
44 from ...datastore import DatastoreTransaction
45 from ...datastore.stored_file_info import StoredDatastoreItemInfo
46 from ...dimensions import DimensionUniverse
47 from ._database import Database, StaticTablesContext
48 from ._datasets import DatasetRecordStorageManager
49 from ._opaque import OpaqueTableStorage, OpaqueTableStorageManager
52@immutable
53class FakeDatasetRef:
54 """A fake `DatasetRef` that can be used internally by butler where
55 only the dataset ID is available.
57 Should only be used when registry can not be used to create a full
58 `DatasetRef` from the ID. A particular use case is during dataset
59 deletion when solely the ID is available.
61 Parameters
62 ----------
63 id : `DatasetId`
64 The dataset ID.
65 """
67 __slots__ = ("id",)
69 def __init__(self, id: DatasetId):
70 self.id = id
72 def __str__(self) -> str:
73 return f"dataset_id={self.id}"
75 def __repr__(self) -> str:
76 return f"FakeDatasetRef({self.id})"
78 def __eq__(self, other: Any) -> bool:
79 try:
80 return self.id == other.id
81 except AttributeError:
82 return NotImplemented
84 def __hash__(self) -> int:
85 return hash(self.id)
87 id: DatasetId
88 """Unique identifier for this dataset.
89 """
91 @property
92 def datasetType(self) -> DatasetType:
93 raise AttributeError("A FakeDatasetRef can not be associated with a valid DatasetType")
95 @property
96 def _datastore_records(self) -> DatasetDatastoreRecords | None:
97 raise AttributeError("A FakeDatasetRef can not be associated with datastore records")
100DatasetIdRef = DatasetRef | FakeDatasetRef
101"""A type-annotation alias that matches both `DatasetRef` and `FakeDatasetRef`.
102"""
105class DatastoreRegistryBridge(ABC):
106 """An abstract base class that defines the interface that a `Datastore`
107 uses to communicate with a `Registry`.
109 Parameters
110 ----------
111 datastoreName : `str`
112 Name of the `Datastore` as it should appear in `Registry` tables
113 referencing it.
114 """
116 def __init__(self, datastoreName: str):
117 self.datastoreName = datastoreName
119 @abstractmethod
120 def insert(self, refs: Iterable[DatasetIdRef]) -> None:
121 """Record that a datastore holds the given datasets.
123 Parameters
124 ----------
125 refs : `~collections.abc.Iterable` of `DatasetIdRef`
126 References to the datasets.
127 """
128 raise NotImplementedError()
130 @abstractmethod
131 def ensure(self, refs: Iterable[DatasetIdRef]) -> None:
132 """Record that a datastore holds the given datasets, skipping if
133 the ref is already registered.
135 Parameters
136 ----------
137 refs : `~collections.abc.Iterable` of `DatasetIdRef`
138 References to the datasets.
139 """
140 raise NotImplementedError()
142 @abstractmethod
143 def forget(self, refs: Iterable[DatasetIdRef]) -> None:
144 """Remove dataset location information without any attempt to put it
145 in the trash while waiting for external deletes.
147 This should be used only to implement `Datastore.forget`, or in cases
148 where deleting the actual datastore artifacts cannot fail.
150 Parameters
151 ----------
152 refs : `~collections.abc.Iterable` of `DatasetIdRef`
153 References to the datasets.
154 """
155 raise NotImplementedError()
157 @abstractmethod
158 def moveToTrash(self, refs: Iterable[DatasetIdRef], transaction: DatastoreTransaction | None) -> None:
159 """Move dataset location information to trash.
161 Parameters
162 ----------
163 refs : `~collections.abc.Iterable` of `DatasetIdRef`
164 References to the datasets.
165 transaction : `DatastoreTransaction` or `None`
166 Transaction object. Can be `None` in some bridges or if no rollback
167 is required.
168 """
169 raise NotImplementedError()
171 @abstractmethod
172 def check(self, refs: Iterable[DatasetIdRef]) -> Iterable[DatasetIdRef]:
173 """Check which refs are listed for this datastore.
175 Parameters
176 ----------
177 refs : `~collections.abc.Iterable` of `DatasetIdRef`
178 References to the datasets.
180 Returns
181 -------
182 present : `~collections.abc.Iterable` [ `DatasetIdRef` ]
183 Datasets from ``refs`` that are recorded as being in this
184 datastore.
185 """
186 raise NotImplementedError()
188 @abstractmethod
189 def emptyTrash(
190 self,
191 records_table: OpaqueTableStorage | None = None,
192 record_class: type[StoredDatastoreItemInfo] | None = None,
193 record_column: str | None = None,
194 selected_ids: Collection[DatasetId] | None = None,
195 dry_run: bool = False,
196 ) -> AbstractContextManager[
197 tuple[Iterable[tuple[DatasetIdRef, StoredDatastoreItemInfo | None]], set[str] | None]
198 ]:
199 """Retrieve all the dataset ref IDs that are in the trash
200 associated for this datastore, and then remove them if the context
201 exists without an exception being raised.
203 Parameters
204 ----------
205 records_table : `OpaqueTableStorage`, optional
206 Table of records to query with the trash records.
207 record_class : `type` of `StoredDatastoreItemInfo`, optional
208 Class to use when reading records from ``records_table``.
209 record_column : `str`, optional
210 Name of the column in records_table that refers to the artifact.
211 selected_ids : `collections.abc.Collection` [ `DatasetId` ] \
212 or `None`, optional
213 If provided, collection of IDs that should be trashed. Only records
214 within this selection will be yielded and then removed. This
215 can be used to allow a subset of the trash table to be emptied.
216 If an empty set is given no artifacts will be trashed. If `None`
217 the full list from the trash table will be used.
218 dry_run : `bool`, optional
219 If `True`, the trash table will be queried and results reported
220 but no artifacts will be removed.
222 Yields
223 ------
224 matches : iterable of (`DatasetIdRef`, `StoredDatastoreItemInfo`)
225 The IDs of datasets that can be safely removed from this datastore
226 and the corresponding information from the records table.
227 Can be empty.
228 artifacts_to_keep : `set` of `str`, optional
229 Any external artifacts that are known to the table but which should
230 not be deleted. If `None`, the caller should check themselves.
232 Examples
233 --------
234 Typical usage by a Datastore is something like::
236 with self.bridge.emptyTrash() as trashed:
237 iter, to_keep = trashed
238 for ref, info in iter:
239 # Remove artifacts associated with id,
240 # raise an exception if something goes wrong.
242 Notes
243 -----
244 The object yielded by the context manager may be a single-pass
245 iterator. If multiple passes are required, it should be converted to
246 a `list` or other container.
248 Datastores should never raise (except perhaps in testing) when an
249 artifact cannot be removed only because it is already gone - this
250 condition is an unavoidable outcome of concurrent delete operations,
251 and must not be considered and error for those to be safe.
253 If a table record is provided the trashed records will be deleted
254 when the context manager completes.
255 """
256 raise NotImplementedError()
258 datastoreName: str
259 """The name of the `Datastore` as it should appear in `Registry` tables
260 (`str`).
261 """
264class DatastoreRegistryBridgeManager(VersionedExtension):
265 """An abstract base class that defines the interface between `Registry`
266 and `Datastore` when a new `Datastore` is constructed.
268 Parameters
269 ----------
270 opaque : `OpaqueTableStorageManager`
271 Manager object for opaque table storage in the `Registry`.
272 universe : `DimensionUniverse`
273 All dimensions know to the `Registry`.
274 registry_schema_version : `VersionTuple` or `None`, optional
275 Version of registry schema.
277 Notes
278 -----
279 Datastores are passed an instance of `DatastoreRegistryBridgeManager` at
280 construction, and should use it to obtain and keep any of the following:
282 - a `DatastoreRegistryBridge` instance to record in the `Registry` what is
283 present in the datastore (needed by all datastores that are not just
284 forwarders);
286 - one or more `OpaqueTableStorage` instance if they wish to store internal
287 records in the `Registry` database;
289 - the `DimensionUniverse`, if they need it to (e.g.) construct or validate
290 filename templates.
291 """
293 def __init__(
294 self,
295 *,
296 opaque: OpaqueTableStorageManager,
297 universe: DimensionUniverse,
298 registry_schema_version: VersionTuple | None = None,
299 ):
300 super().__init__(registry_schema_version=registry_schema_version)
301 self.opaque = opaque
302 self.universe = universe
304 @abstractmethod
305 def clone(
306 self,
307 *,
308 db: Database,
309 opaque: OpaqueTableStorageManager,
310 ) -> DatastoreRegistryBridgeManager:
311 """Make an independent copy of this manager instance bound to new
312 instances of `Database` and other managers.
314 Parameters
315 ----------
316 db : `Database`
317 New `Database` object to use when instantiating the manager.
318 opaque : `OpaqueTableStorageManager`
319 New `OpaqueTableStorageManager` object to use when instantiating
320 the manager.
322 Returns
323 -------
324 instance : `DatastoreRegistryBridgeManager`
325 New manager instance with the same configuration as this instance,
326 but bound to a new Database object.
327 """
328 raise NotImplementedError()
330 @classmethod
331 @abstractmethod
332 def initialize(
333 cls,
334 db: Database,
335 context: StaticTablesContext,
336 *,
337 opaque: OpaqueTableStorageManager,
338 datasets: type[DatasetRecordStorageManager],
339 universe: DimensionUniverse,
340 registry_schema_version: VersionTuple | None = None,
341 ) -> DatastoreRegistryBridgeManager:
342 """Construct an instance of the manager.
344 Parameters
345 ----------
346 db : `Database`
347 Interface to the underlying database engine and namespace.
348 context : `StaticTablesContext`
349 Context object obtained from `Database.declareStaticTables`; used
350 to declare any tables that should always be present in a layer
351 implemented with this manager.
352 opaque : `OpaqueTableStorageManager`
353 Registry manager object for opaque (to Registry) tables, provided
354 to allow Datastores to store their internal information inside the
355 Registry database.
356 datasets : subclass of `DatasetRecordStorageManager`
357 Concrete class that will be used to manage the core dataset tables
358 in this registry; should be used only to create foreign keys to
359 those tables.
360 universe : `DimensionUniverse`
361 All dimensions known to the registry.
362 registry_schema_version : `VersionTuple` or `None`
363 Schema version of this extension as defined in registry.
365 Returns
366 -------
367 manager : `DatastoreRegistryBridgeManager`
368 An instance of a concrete `DatastoreRegistryBridgeManager`
369 subclass.
370 """
371 raise NotImplementedError()
373 @abstractmethod
374 def refresh(self) -> None:
375 """Ensure all other operations on this manager are aware of any
376 collections that may have been registered by other clients since it
377 was initialized or last refreshed.
378 """
379 raise NotImplementedError()
381 @abstractmethod
382 def register(self, name: str, *, ephemeral: bool = True) -> DatastoreRegistryBridge:
383 """Register a new `Datastore` associated with this `Registry`.
385 This method should be called by all `Datastore` classes aside from
386 those that only forward storage to other datastores.
388 Parameters
389 ----------
390 name : `str`
391 Name of the datastore, as it should appear in `Registry` tables.
392 ephemeral : `bool`, optional
393 If `True` (`False` is default), return a bridge object that is
394 backed by storage that will not last past the end of the current
395 process. This should be used whenever the same is true of the
396 dataset's artifacts.
398 Returns
399 -------
400 bridge : `DatastoreRegistryBridge`
401 Object that provides the interface this `Datastore` should use to
402 communicate with the `Registry`.
403 """
404 raise NotImplementedError()
406 @abstractmethod
407 def findDatastores(self, ref: DatasetIdRef) -> Iterable[str]:
408 """Retrieve datastore locations for a given dataset.
410 Parameters
411 ----------
412 ref : `DatasetIdRef`
413 A reference to the dataset for which to retrieve storage
414 information.
416 Returns
417 -------
418 datastores : `~collections.abc.Iterable` [ `str` ]
419 All the matching datastores holding this dataset. Empty if the
420 dataset does not exist anywhere.
422 Raises
423 ------
424 AmbiguousDatasetError
425 Raised if ``ref.id`` is `None`.
426 """
427 raise NotImplementedError()
429 opaque: OpaqueTableStorageManager
430 """Registry manager object for opaque (to Registry) tables, provided
431 to allow Datastores to store their internal information inside the
432 Registry database.
433 """
435 universe: DimensionUniverse
436 """All dimensions known to the `Registry`.
437 """