Coverage for python/lsst/daf/butler/dimensions/_records.py: 21%
167 statements
« prev ^ index » next coverage.py v7.4.1, created at 2024-02-13 10:57 +0000
« prev ^ index » next coverage.py v7.4.1, created at 2024-02-13 10:57 +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/>.
28from __future__ import annotations
30__all__ = ("DimensionRecord", "SerializedDimensionRecord")
32from collections.abc import Hashable
33from typing import TYPE_CHECKING, Any, ClassVar, Optional, Tuple
35import lsst.sphgeom
36from lsst.utils.classes import immutable
37from pydantic import BaseModel, Field, StrictBool, StrictFloat, StrictInt, StrictStr, create_model
39from .._timespan import Timespan
40from ..json import from_json_pydantic, to_json_pydantic
41from ..persistence_context import PersistenceContextVars
42from ._elements import Dimension, DimensionElement
44if TYPE_CHECKING: # Imports needed only for type annotations; may be circular.
45 from ..registry import Registry
46 from ._coordinate import DataCoordinate
47 from ._schema import DimensionElementFields
48 from ._universe import DimensionUniverse
51def _reconstructDimensionRecord(definition: DimensionElement, mapping: dict[str, Any]) -> DimensionRecord:
52 """Unpickle implementation for `DimensionRecord` subclasses.
54 For internal use by `DimensionRecord`.
55 """
56 return definition.RecordClass(**mapping)
59def _subclassDimensionRecord(definition: DimensionElement) -> type[DimensionRecord]:
60 """Create a dynamic subclass of `DimensionRecord` for the given element.
62 For internal use by `DimensionRecord`.
63 """
64 from ._schema import DimensionElementFields
66 fields = DimensionElementFields(definition)
67 slots = list(fields.standard.names)
68 if definition.spatial:
69 slots.append("region")
70 if definition.temporal:
71 slots.append("timespan")
72 d = {"definition": definition, "__slots__": tuple(slots), "fields": fields}
73 return type(definition.name + ".RecordClass", (DimensionRecord,), d)
76class SpecificSerializedDimensionRecord(BaseModel, extra="forbid"):
77 """Base model for a specific serialized record content."""
80_SIMPLE_RECORD_CLASS_CACHE: dict[
81 tuple[DimensionElement, DimensionUniverse], type[SpecificSerializedDimensionRecord]
82] = {}
85def _createSimpleRecordSubclass(definition: DimensionElement) -> type[SpecificSerializedDimensionRecord]:
86 from ._schema import DimensionElementFields
88 # Cache on the definition (which hashes as the name) and the
89 # associated universe.
90 cache_key = (definition, definition.universe)
91 if cache_key in _SIMPLE_RECORD_CLASS_CACHE:
92 return _SIMPLE_RECORD_CLASS_CACHE[cache_key]
94 fields = DimensionElementFields(definition)
95 members = {}
96 # Prefer strict typing for external data
97 type_map = {
98 str: StrictStr,
99 float: StrictFloat,
100 bool: StrictBool,
101 int: StrictInt,
102 }
104 for field in fields.standard:
105 field_type = field.getPythonType()
106 field_type = type_map.get(field_type, field_type)
107 if field.nullable:
108 field_type = Optional[field_type] # type: ignore
109 members[field.name] = (field_type, ...)
110 if definition.temporal:
111 members["timespan"] = (Optional[Tuple[int, int]], ...) # type: ignore
112 if definition.spatial:
113 members["region"] = (str, ...)
115 # mypy does not seem to like create_model
116 model = create_model(
117 f"SpecificSerializedDimensionRecord{definition.name.capitalize()}",
118 __base__=SpecificSerializedDimensionRecord,
119 **members, # type: ignore
120 )
122 _SIMPLE_RECORD_CLASS_CACHE[cache_key] = model
123 return model
126# While supporting pydantic v1 and v2 keep this outside the model.
127_serialized_dimension_record_schema_extra = {
128 "examples": [
129 {
130 "definition": "detector",
131 "record": {
132 "instrument": "HSC",
133 "id": 72,
134 "full_name": "0_01",
135 "name_in_raft": "01",
136 "raft": "0",
137 "purpose": "SCIENCE",
138 },
139 }
140 ]
141}
144class SerializedDimensionRecord(BaseModel):
145 """Simplified model for serializing a `DimensionRecord`."""
147 definition: str = Field(
148 ...,
149 title="Name of dimension associated with this record.",
150 examples=["exposure"],
151 )
153 # Use strict types to prevent casting
154 record: dict[str, None | StrictBool | StrictInt | StrictFloat | StrictStr | tuple[int, int]] = Field(
155 ...,
156 title="Dimension record keys and values.",
157 examples=[
158 {
159 "definition": "exposure",
160 "record": {
161 "instrument": "LATISS",
162 "exposure": 2021050300044,
163 "obs_id": "AT_O_20210503_00044",
164 },
165 }
166 ],
167 )
169 model_config = {
170 "json_schema_extra": _serialized_dimension_record_schema_extra, # type: ignore[typeddict-item]
171 }
173 @classmethod
174 def direct(
175 cls,
176 *,
177 definition: str,
178 record: dict[str, None | StrictFloat | StrictStr | StrictBool | StrictInt | tuple[int, int]],
179 ) -> SerializedDimensionRecord:
180 """Construct a `SerializedDimensionRecord` directly without validators.
182 Parameters
183 ----------
184 definition : `str`
185 The name of the record.
186 record : `dict`
187 A dictionary representation of the record content.
189 Returns
190 -------
191 rec : `SerializedDimensionRecord`
192 A model representing the dimension records.
194 Notes
195 -----
196 This differs from the pydantic "construct" method in that the arguments
197 are explicitly what the model requires, and it will recurse through
198 members, constructing them from their corresponding `direct` methods.
200 This method should only be called when the inputs are trusted.
201 """
202 # This method requires tuples as values of the mapping, but JSON
203 # readers will read things in as lists. Be kind and transparently
204 # transform to tuples
205 _recItems = {
206 k: v if type(v) != list else tuple(v) for k, v in record.items() # type: ignore # noqa: E721
207 }
209 # Type ignore because the ternary statement seems to confuse mypy
210 # based on conflicting inferred types of v.
211 key = (
212 definition,
213 frozenset(_recItems.items()),
214 )
215 cache = PersistenceContextVars.serializedDimensionRecordMapping.get()
216 if cache is not None and (result := cache.get(key)) is not None:
217 return result
219 node = cls.model_construct(definition=definition, record=_recItems) # type: ignore
221 if cache is not None:
222 cache[key] = node
223 return node
226@immutable
227class DimensionRecord:
228 """Base class for the Python representation of database records.
230 Parameters
231 ----------
232 **kwargs
233 Field values for this record. Unrecognized keys are ignored. If this
234 is the record for a `Dimension`, its primary key value may be provided
235 with the actual name of the field (e.g. "id" or "name"), the name of
236 the `Dimension`, or both. If this record class has a "timespan"
237 attribute, "datetime_begin" and "datetime_end" keyword arguments may
238 be provided instead of a single "timespan" keyword argument (but are
239 ignored if a "timespan" argument is provided).
241 Notes
242 -----
243 `DimensionRecord` subclasses are created dynamically for each
244 `DimensionElement` in a `DimensionUniverse`, and are accessible via the
245 `DimensionElement.RecordClass` attribute. The `DimensionRecord` base class
246 itself is pure abstract, but does not use the `abc` module to indicate this
247 because it does not have overridable methods.
249 Record classes have attributes that correspond exactly to the
250 `~DimensionElementFields.standard` fields in the related database table,
251 plus "region" and "timespan" attributes for spatial and/or temporal
252 elements (respectively).
254 Instances are usually obtained from a `Registry`, but can be constructed
255 directly from Python as well.
257 `DimensionRecord` instances are immutable.
258 """
260 # Derived classes are required to define __slots__ as well, and it's those
261 # derived-class slots that other methods on the base class expect to see
262 # when they access self.__slots__.
263 __slots__ = ("dataId",)
265 _serializedType = SerializedDimensionRecord
267 def __init__(self, **kwargs: Any):
268 # Accept either the dimension name or the actual name of its primary
269 # key field; ensure both are present in the dict for convenience below.
270 if isinstance(self.definition, Dimension):
271 v = kwargs.get(self.definition.primaryKey.name)
272 if v is None:
273 v = kwargs.get(self.definition.name)
274 if v is None:
275 raise ValueError(
276 f"No value provided for {self.definition.name}.{self.definition.primaryKey.name}."
277 )
278 kwargs[self.definition.primaryKey.name] = v
279 else:
280 v2 = kwargs.setdefault(self.definition.name, v)
281 if v != v2:
282 raise ValueError(
283 "Multiple inconsistent values for "
284 f"{self.definition.name}.{self.definition.primaryKey.name}: {v!r} != {v2!r}."
285 )
287 from ._coordinate import DataCoordinate
289 object.__setattr__(
290 self,
291 "dataId",
292 DataCoordinate.from_required_values(
293 self.definition.minimal_group,
294 tuple(kwargs[dimension] for dimension in self.definition.required.names),
295 ),
296 )
297 # Don't need the primary key value aliased to the dimension name
298 # anymore.
299 kwargs.pop(self.definition.name, None)
301 for name in self.__slots__:
302 # Note that we remove from kwargs as we go, to make sure there's
303 # nothing left at the end.
304 object.__setattr__(self, name, kwargs.pop(name, None))
305 # Support 'datetime_begin' and 'datetime_end' instead of 'timespan' for
306 # backwards compatibility, but if one is provided both must be.
307 if self.definition.temporal is not None and self.timespan is None and "datetime_begin" in kwargs:
308 object.__setattr__(
309 self,
310 "timespan",
311 Timespan(
312 kwargs.pop("datetime_begin"),
313 kwargs.pop("datetime_end"),
314 ),
315 )
317 if kwargs:
318 raise TypeError(f"Invalid fields for {self.definition} dimension record: {set(kwargs.keys())}.")
320 def __eq__(self, other: Any) -> bool:
321 if type(other) != type(self):
322 return False
323 return self.dataId == other.dataId
325 def __hash__(self) -> int:
326 return hash(self.dataId.required_values)
328 def __str__(self) -> str:
329 lines = [f"{self.definition.name}:"]
330 lines.extend(f" {name}: {getattr(self, name)!r}" for name in self.__slots__)
331 return "\n".join(lines)
333 def __repr__(self) -> str:
334 return "{}.RecordClass({})".format(
335 self.definition.name, ", ".join(f"{name}={getattr(self, name)!r}" for name in self.__slots__)
336 )
338 def __reduce__(self) -> tuple:
339 mapping = {name: getattr(self, name) for name in self.__slots__}
340 return (_reconstructDimensionRecord, (self.definition, mapping))
342 def _repr_html_(self) -> str:
343 """Override the default representation in IPython/Jupyter notebooks.
345 This gives a more readable output that understands embedded newlines.
346 """
347 return f"<pre>{self}<pre>"
349 def to_simple(self, minimal: bool = False) -> SerializedDimensionRecord:
350 """Convert this class to a simple python type.
352 This makes it suitable for serialization.
354 Parameters
355 ----------
356 minimal : `bool`, optional
357 Use minimal serialization. Has no effect on for this class.
359 Returns
360 -------
361 names : `list`
362 The names of the dimensions.
363 """
364 # The DataId is sufficient if you are willing to do a deferred
365 # query. This may not be overly useful since to reconstruct
366 # a collection of records will require repeated registry queries.
367 # For now do not implement minimal form.
368 key = (id(self.definition), self.dataId)
369 cache = PersistenceContextVars.serializedDimensionRecordMapping.get()
370 if cache is not None and (result := cache.get(key)) is not None:
371 return result
373 mapping = {name: getattr(self, name) for name in self.__slots__}
374 # If the item in mapping supports simplification update it
375 for k, v in mapping.items():
376 try:
377 mapping[k] = v.to_simple(minimal=minimal)
378 except AttributeError:
379 if isinstance(v, lsst.sphgeom.Region):
380 # YAML serialization specifies the class when it
381 # doesn't have to. This is partly for explicitness
382 # and also history. Here use a different approach.
383 # This code needs to be migrated to sphgeom
384 mapping[k] = v.encode().hex()
385 if isinstance(v, bytes):
386 # We actually can't handle serializing out to bytes for
387 # hash objects, encode it here to a hex string
388 mapping[k] = v.hex()
389 definition = self.definition.to_simple(minimal=minimal)
390 dimRec = SerializedDimensionRecord(definition=definition, record=mapping)
391 if cache is not None:
392 cache[key] = dimRec
393 return dimRec
395 @classmethod
396 def from_simple(
397 cls,
398 simple: SerializedDimensionRecord,
399 universe: DimensionUniverse | None = None,
400 registry: Registry | None = None,
401 cacheKey: Hashable | None = None,
402 ) -> DimensionRecord:
403 """Construct a new object from the simplified form.
405 This is generally data returned from the `to_simple`
406 method.
408 Parameters
409 ----------
410 simple : `SerializedDimensionRecord`
411 Value return from `to_simple`.
412 universe : `DimensionUniverse`
413 The special graph of all known dimensions of which this graph will
414 be a subset. Can be `None` if `Registry` is provided.
415 registry : `lsst.daf.butler.Registry`, optional
416 Registry from which a universe can be extracted. Can be `None`
417 if universe is provided explicitly.
418 cacheKey : `Hashable` or `None`
419 If this is not None, it will be used as a key for any cached
420 reconstruction instead of calculating a value from the serialized
421 format.
423 Returns
424 -------
425 record : `DimensionRecord`
426 Newly-constructed object.
427 """
428 if universe is None and registry is None:
429 raise ValueError("One of universe or registry is required to convert names to a DimensionGraph")
430 if universe is None and registry is not None:
431 universe = registry.dimensions
432 if universe is None:
433 # this is for mypy
434 raise ValueError("Unable to determine a usable universe")
435 # Type ignore because the ternary statement seems to confuse mypy
436 # based on conflicting inferred types of v.
437 key = cacheKey or (
438 simple.definition,
439 frozenset(simple.record.items()), # type: ignore
440 )
441 cache = PersistenceContextVars.dimensionRecords.get()
442 if cache is not None and (result := cache.get(key)) is not None:
443 return result
445 definition = DimensionElement.from_simple(simple.definition, universe=universe)
447 # Create a specialist subclass model with type validation.
448 # This allows us to do simple checks of external data (possibly
449 # sent as JSON) since for now _reconstructDimensionRecord does not
450 # do any validation.
451 record_model_cls = _createSimpleRecordSubclass(definition)
452 record_model = record_model_cls(**simple.record)
454 # Timespan and region have to be converted to native form
455 # for now assume that those keys are special
456 rec = record_model.model_dump()
458 if (ts := "timespan") in rec:
459 rec[ts] = Timespan.from_simple(rec[ts], universe=universe, registry=registry)
460 if (reg := "region") in rec:
461 encoded = bytes.fromhex(rec[reg])
462 rec[reg] = lsst.sphgeom.Region.decode(encoded)
463 if (hsh := "hash") in rec:
464 rec[hsh] = bytes.fromhex(rec[hsh].decode())
466 dimRec = _reconstructDimensionRecord(definition, rec)
467 if cache is not None:
468 cache[key] = dimRec
469 return dimRec
471 to_json = to_json_pydantic
472 from_json: ClassVar = classmethod(from_json_pydantic)
474 def toDict(self, splitTimespan: bool = False) -> dict[str, Any]:
475 """Return a vanilla `dict` representation of this record.
477 Parameters
478 ----------
479 splitTimespan : `bool`, optional
480 If `True` (`False` is default) transform any "timespan" key value
481 from a `Timespan` instance into a pair of regular
482 ("datetime_begin", "datetime_end") fields.
483 """
484 results = {name: getattr(self, name) for name in self.__slots__}
485 if splitTimespan:
486 timespan = results.pop("timespan", None)
487 if timespan is not None:
488 results["datetime_begin"] = timespan.begin
489 results["datetime_end"] = timespan.end
490 return results
492 # DimensionRecord subclasses are dynamically created, so static type
493 # checkers can't know about them or their attributes. To avoid having to
494 # put "type: ignore", everywhere, add a dummy __getattr__ that tells type
495 # checkers not to worry about missing attributes.
496 def __getattr__(self, name: str) -> Any:
497 raise AttributeError(name)
499 # Class attributes below are shadowed by instance attributes, and are
500 # present just to hold the docstrings for those instance attributes.
502 dataId: DataCoordinate
503 """A dict-like identifier for this record's primary keys
504 (`DataCoordinate`).
505 """
507 definition: ClassVar[DimensionElement]
508 """The `DimensionElement` whose records this class represents
509 (`DimensionElement`).
510 """
512 fields: ClassVar[DimensionElementFields]
513 """A categorized view of the fields in this class
514 (`DimensionElementFields`).
515 """