Coverage for python/lsst/pipe/base/graphBuilder.py : 22%

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 pipe_base.
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/>.
21from __future__ import annotations
23"""Module defining GraphBuilder class and related methods.
24"""
26__all__ = ['GraphBuilder']
28# -------------------------------
29# Imports of standard modules --
30# -------------------------------
31import itertools
32from collections import ChainMap
33from contextlib import contextmanager
34from dataclasses import dataclass
35from typing import Dict, Iterable, Iterator, List, Set
36import logging
39# -----------------------------
40# Imports for other modules --
41# -----------------------------
42from .connections import iterConnections
43from .pipeline import PipelineDatasetTypes, TaskDatasetTypes, TaskDef, Pipeline
44from .graph import QuantumGraph
45from lsst.daf.butler import (
46 DataCoordinate,
47 DatasetRef,
48 DatasetType,
49 DimensionGraph,
50 DimensionUniverse,
51 NamedKeyDict,
52 Quantum,
53)
54from lsst.daf.butler.registry.queries.exprParser import ParseError, ParserYacc, TreeVisitor
55from lsst.utils import doImport
57# ----------------------------------
58# Local non-exported definitions --
59# ----------------------------------
61_LOG = logging.getLogger(__name__.partition(".")[2])
64class _DatasetDict(NamedKeyDict[DatasetType, Dict[DataCoordinate, DatasetRef]]):
65 """A custom dictionary that maps `DatasetType` to a nested dictionary of
66 the known `DatasetRef` instances of that type.
68 Parameters
69 ----------
70 args
71 Positional arguments are forwarded to the `dict` constructor.
72 universe : `DimensionUniverse`
73 Universe of all possible dimensions.
74 """
75 def __init__(self, *args, universe: DimensionGraph):
76 super().__init__(*args)
77 self.universe = universe
79 @classmethod
80 def fromDatasetTypes(cls, datasetTypes: Iterable[DatasetType], *,
81 universe: DimensionUniverse) -> _DatasetDict:
82 """Construct a dictionary from a flat iterable of `DatasetType` keys.
84 Parameters
85 ----------
86 datasetTypes : `iterable` of `DatasetType`
87 DatasetTypes to use as keys for the dict. Values will be empty
88 dictionaries.
89 universe : `DimensionUniverse`
90 Universe of all possible dimensions.
92 Returns
93 -------
94 dictionary : `_DatasetDict`
95 A new `_DatasetDict` instance.
96 """
97 return cls({datasetType: {} for datasetType in datasetTypes}, universe=universe)
99 @classmethod
100 def fromSubset(cls, datasetTypes: Iterable[DatasetType], first: _DatasetDict, *rest: _DatasetDict
101 ) -> _DatasetDict:
102 """Return a new dictionary by extracting items corresponding to the
103 given keys from one or more existing dictionaries.
105 Parameters
106 ----------
107 datasetTypes : `iterable` of `DatasetType`
108 DatasetTypes to use as keys for the dict. Values will be obtained
109 by lookups against ``first`` and ``rest``.
110 first : `_DatasetDict`
111 Another dictionary from which to extract values.
112 rest
113 Additional dictionaries from which to extract values.
115 Returns
116 -------
117 dictionary : `_DatasetDict`
118 A new dictionary instance.
119 """
120 combined = ChainMap(first, *rest)
121 return cls({datasetType: combined[datasetType] for datasetType in datasetTypes},
122 universe=first.universe)
124 @property
125 def dimensions(self) -> DimensionGraph:
126 """The union of all dimensions used by all dataset types in this
127 dictionary, including implied dependencies (`DimensionGraph`).
128 """
129 base = self.universe.empty
130 if len(self) == 0:
131 return base
132 return base.union(*[datasetType.dimensions for datasetType in self.keys()])
134 def unpackSingleRefs(self) -> NamedKeyDict[DatasetType, DatasetRef]:
135 """Unpack nested single-element `DatasetRef` dicts into a new
136 mapping with `DatasetType` keys and `DatasetRef` values.
138 This method assumes that each nest contains exactly one item, as is the
139 case for all "init" datasets.
141 Returns
142 -------
143 dictionary : `NamedKeyDict`
144 Dictionary mapping `DatasetType` to `DatasetRef`, with both
145 `DatasetType` instances and string names usable as keys.
146 """
147 def getOne(refs: Dict[DataCoordinate, DatasetRef]) -> DatasetRef:
148 ref, = refs.values()
149 return ref
150 return NamedKeyDict({datasetType: getOne(refs) for datasetType, refs in self.items()})
152 def unpackMultiRefs(self) -> NamedKeyDict[DatasetType, DatasetRef]:
153 """Unpack nested multi-element `DatasetRef` dicts into a new
154 mapping with `DatasetType` keys and `set` of `DatasetRef` values.
156 Returns
157 -------
158 dictionary : `NamedKeyDict`
159 Dictionary mapping `DatasetType` to `DatasetRef`, with both
160 `DatasetType` instances and string names usable as keys.
161 """
162 return NamedKeyDict({datasetType: list(refs.values()) for datasetType, refs in self.items()})
164 def extract(self, datasetType: DatasetType, dataIds: Iterable[DataCoordinate]
165 ) -> Iterator[DatasetRef]:
166 """Iterate over the contained `DatasetRef` instances that match the
167 given `DatasetType` and data IDs.
169 Parameters
170 ----------
171 datasetType : `DatasetType`
172 Dataset type to match.
173 dataIds : `Iterable` [ `DataCoordinate` ]
174 Data IDs to match.
176 Returns
177 -------
178 refs : `Iterator` [ `DatasetRef` ]
179 DatasetRef instances for which ``ref.datasetType == datasetType``
180 and ``ref.dataId`` is in ``dataIds``.
181 """
182 refs = self[datasetType]
183 return (refs[dataId] for dataId in dataIds)
186class _QuantumScaffolding:
187 """Helper class aggregating information about a `Quantum`, used when
188 constructing a `QuantumGraph`.
190 See `_PipelineScaffolding` for a top-down description of the full
191 scaffolding data structure.
193 Parameters
194 ----------
195 task : _TaskScaffolding
196 Back-reference to the helper object for the `PipelineTask` this quantum
197 represents an execution of.
198 dataId : `DataCoordinate`
199 Data ID for this quantum.
200 """
201 def __init__(self, task: _TaskScaffolding, dataId: DataCoordinate):
202 self.task = task
203 self.dataId = dataId
204 self.inputs = _DatasetDict.fromDatasetTypes(task.inputs.keys(), universe=dataId.universe)
205 self.outputs = _DatasetDict.fromDatasetTypes(task.outputs.keys(), universe=dataId.universe)
206 self.prerequisites = _DatasetDict.fromDatasetTypes(task.prerequisites.keys(),
207 universe=dataId.universe)
209 __slots__ = ("task", "dataId", "inputs", "outputs", "prerequisites")
211 def __repr__(self):
212 return f"_QuantumScaffolding(taskDef={self.task.taskDef}, dataId={self.dataId}, ...)"
214 task: _TaskScaffolding
215 """Back-reference to the helper object for the `PipelineTask` this quantum
216 represents an execution of.
217 """
219 dataId: DataCoordinate
220 """Data ID for this quantum.
221 """
223 inputs: _DatasetDict
224 """Nested dictionary containing `DatasetRef` inputs to this quantum.
226 This is initialized to map each `DatasetType` to an empty dictionary at
227 construction. Those nested dictionaries are populated (with data IDs as
228 keys) with unresolved `DatasetRef` instances in
229 `_PipelineScaffolding.connectDataIds`.
230 """
232 outputs: _DatasetDict
233 """Nested dictionary containing `DatasetRef` outputs this quantum.
234 """
236 prerequisites: _DatasetDict
237 """Nested dictionary containing `DatasetRef` prerequisite inputs to this
238 quantum.
239 """
241 def makeQuantum(self) -> Quantum:
242 """Transform the scaffolding object into a true `Quantum` instance.
244 Returns
245 -------
246 quantum : `Quantum`
247 An actual `Quantum` instance.
248 """
249 allInputs = self.inputs.unpackMultiRefs()
250 allInputs.update(self.prerequisites.unpackMultiRefs())
251 # Give the task's Connections class an opportunity to remove some
252 # inputs, or complain if they are unacceptable.
253 # This will raise if one of the check conditions is not met, which is the intended
254 # behavior
255 allInputs = self.task.taskDef.connections.adjustQuantum(allInputs)
256 return Quantum(
257 taskName=self.task.taskDef.taskName,
258 taskClass=self.task.taskDef.taskClass,
259 dataId=self.dataId,
260 initInputs=self.task.initInputs.unpackSingleRefs(),
261 inputs=allInputs,
262 outputs=self.outputs.unpackMultiRefs(),
263 )
266@dataclass
267class _TaskScaffolding:
268 """Helper class aggregating information about a `PipelineTask`, used when
269 constructing a `QuantumGraph`.
271 See `_PipelineScaffolding` for a top-down description of the full
272 scaffolding data structure.
274 Parameters
275 ----------
276 taskDef : `TaskDef`
277 Data structure that identifies the task class and its config.
278 parent : `_PipelineScaffolding`
279 The parent data structure that will hold the instance being
280 constructed.
281 datasetTypes : `TaskDatasetTypes`
282 Data structure that categorizes the dataset types used by this task.
283 """
284 def __init__(self, taskDef: TaskDef, parent: _PipelineScaffolding, datasetTypes: TaskDatasetTypes):
285 universe = parent.dimensions.universe
286 self.taskDef = taskDef
287 self.dimensions = DimensionGraph(universe, names=taskDef.connections.dimensions)
288 assert self.dimensions.issubset(parent.dimensions)
289 # Initialize _DatasetDicts as subsets of the one or two
290 # corresponding dicts in the parent _PipelineScaffolding.
291 self.initInputs = _DatasetDict.fromSubset(datasetTypes.initInputs, parent.initInputs,
292 parent.initIntermediates)
293 self.initOutputs = _DatasetDict.fromSubset(datasetTypes.initOutputs, parent.initIntermediates,
294 parent.initOutputs)
295 self.inputs = _DatasetDict.fromSubset(datasetTypes.inputs, parent.inputs, parent.intermediates)
296 self.outputs = _DatasetDict.fromSubset(datasetTypes.outputs, parent.intermediates, parent.outputs)
297 self.prerequisites = _DatasetDict.fromSubset(datasetTypes.prerequisites, parent.prerequisites)
298 self.dataIds = set()
299 self.quanta = {}
301 def __repr__(self):
302 # Default dataclass-injected __repr__ gets caught in an infinite loop
303 # because of back-references.
304 return f"_TaskScaffolding(taskDef={self.taskDef}, ...)"
306 taskDef: TaskDef
307 """Data structure that identifies the task class and its config
308 (`TaskDef`).
309 """
311 dimensions: DimensionGraph
312 """The dimensions of a single `Quantum` of this task (`DimensionGraph`).
313 """
315 initInputs: _DatasetDict
316 """Dictionary containing information about datasets used to construct this
317 task (`_DatasetDict`).
318 """
320 initOutputs: _DatasetDict
321 """Dictionary containing information about datasets produced as a
322 side-effect of constructing this task (`_DatasetDict`).
323 """
325 inputs: _DatasetDict
326 """Dictionary containing information about datasets used as regular,
327 graph-constraining inputs to this task (`_DatasetDict`).
328 """
330 outputs: _DatasetDict
331 """Dictionary containing information about datasets produced by this task
332 (`_DatasetDict`).
333 """
335 prerequisites: _DatasetDict
336 """Dictionary containing information about input datasets that must be
337 present in the repository before any Pipeline containing this task is run
338 (`_DatasetDict`).
339 """
341 quanta: Dict[DataCoordinate, _QuantumScaffolding]
342 """Dictionary mapping data ID to a scaffolding object for the Quantum of
343 this task with that data ID.
344 """
346 def makeQuantumSet(self) -> Set[Quantum]:
347 """Create a `set` of `Quantum` from the information in ``self``.
349 Returns
350 -------
351 nodes : `set` of `Quantum
352 The `Quantum` elements corresponding to this task.
353 """
354 return set(q.makeQuantum() for q in self.quanta.values())
357@dataclass
358class _PipelineScaffolding:
359 """A helper data structure that organizes the information involved in
360 constructing a `QuantumGraph` for a `Pipeline`.
362 Parameters
363 ----------
364 pipeline : `Pipeline`
365 Sequence of tasks from which a graph is to be constructed. Must
366 have nested task classes already imported.
367 universe : `DimensionUniverse`
368 Universe of all possible dimensions.
370 Notes
371 -----
372 The scaffolding data structure contains nested data structures for both
373 tasks (`_TaskScaffolding`) and datasets (`_DatasetDict`). The dataset
374 data structures are shared between the pipeline-level structure (which
375 aggregates all datasets and categorizes them from the perspective of the
376 complete pipeline) and the individual tasks that use them as inputs and
377 outputs.
379 `QuantumGraph` construction proceeds in four steps, with each corresponding
380 to a different `_PipelineScaffolding` method:
382 1. When `_PipelineScaffolding` is constructed, we extract and categorize
383 the DatasetTypes used by the pipeline (delegating to
384 `PipelineDatasetTypes.fromPipeline`), then use these to construct the
385 nested `_TaskScaffolding` and `_DatasetDict` objects.
387 2. In `connectDataIds`, we construct and run the "Big Join Query", which
388 returns related tuples of all dimensions used to identify any regular
389 input, output, and intermediate datasets (not prerequisites). We then
390 iterate over these tuples of related dimensions, identifying the subsets
391 that correspond to distinct data IDs for each task and dataset type,
392 and then create `_QuantumScaffolding` objects.
394 3. In `resolveDatasetRefs`, we run follow-up queries against all of the
395 dataset data IDs previously identified, transforming unresolved
396 DatasetRefs into resolved DatasetRefs where appropriate. We then look
397 up prerequisite datasets for all quanta.
399 4. In `makeQuantumGraph`, we construct a `QuantumGraph` from the lists of
400 per-task `_QuantumScaffolding` objects.
401 """
402 def __init__(self, pipeline, *, registry):
403 _LOG.debug("Initializing data structures for QuantumGraph generation.")
404 self.tasks = []
405 # Aggregate and categorize the DatasetTypes in the Pipeline.
406 datasetTypes = PipelineDatasetTypes.fromPipeline(pipeline, registry=registry)
407 # Construct dictionaries that map those DatasetTypes to structures
408 # that will (later) hold addiitonal information about them.
409 for attr in ("initInputs", "initIntermediates", "initOutputs",
410 "inputs", "intermediates", "outputs", "prerequisites"):
411 setattr(self, attr, _DatasetDict.fromDatasetTypes(getattr(datasetTypes, attr),
412 universe=registry.dimensions))
413 # Aggregate all dimensions for all non-init, non-prerequisite
414 # DatasetTypes. These are the ones we'll include in the big join query.
415 self.dimensions = self.inputs.dimensions.union(self.intermediates.dimensions,
416 self.outputs.dimensions)
417 # Construct scaffolding nodes for each Task, and add backreferences
418 # to the Task from each DatasetScaffolding node.
419 # Note that there's only one scaffolding node for each DatasetType, shared by
420 # _PipelineScaffolding and all _TaskScaffoldings that reference it.
421 if isinstance(pipeline, Pipeline):
422 pipeline = pipeline.toExpandedPipeline()
423 self.tasks = [_TaskScaffolding(taskDef=taskDef, parent=self, datasetTypes=taskDatasetTypes)
424 for taskDef, taskDatasetTypes in zip(pipeline,
425 datasetTypes.byTask.values())]
427 def __repr__(self):
428 # Default dataclass-injected __repr__ gets caught in an infinite loop
429 # because of back-references.
430 return f"_PipelineScaffolding(tasks={self.tasks}, ...)"
432 tasks: List[_TaskScaffolding]
433 """Scaffolding data structures for each task in the pipeline
434 (`list` of `_TaskScaffolding`).
435 """
437 initInputs: _DatasetDict
438 """Datasets consumed but not produced when constructing the tasks in this
439 pipeline (`_DatasetDict`).
440 """
442 initIntermediates: _DatasetDict
443 """Datasets that are both consumed and produced when constructing the tasks
444 in this pipeline (`_DatasetDict`).
445 """
447 initOutputs: _DatasetDict
448 """Datasets produced but not consumed when constructing the tasks in this
449 pipeline (`_DatasetDict`).
450 """
452 inputs: _DatasetDict
453 """Datasets that are consumed but not produced when running this pipeline
454 (`_DatasetDict`).
455 """
457 intermediates: _DatasetDict
458 """Datasets that are both produced and consumed when running this pipeline
459 (`_DatasetDict`).
460 """
462 outputs: _DatasetDict
463 """Datasets produced but not consumed when when running this pipeline
464 (`_DatasetDict`).
465 """
467 prerequisites: _DatasetDict
468 """Datasets that are consumed when running this pipeline and looked up
469 per-Quantum when generating the graph (`_DatasetDict`).
470 """
472 dimensions: DimensionGraph
473 """All dimensions used by any regular input, intermediate, or output
474 (not prerequisite) dataset; the set of dimension used in the "Big Join
475 Query" (`DimensionGraph`).
477 This is required to be a superset of all task quantum dimensions.
478 """
480 @contextmanager
481 def connectDataIds(self, registry, collections, userQuery):
482 """Query for the data IDs that connect nodes in the `QuantumGraph`.
484 This method populates `_TaskScaffolding.dataIds` and
485 `_DatasetScaffolding.dataIds` (except for those in `prerequisites`).
487 Parameters
488 ----------
489 registry : `lsst.daf.butler.Registry`
490 Registry for the data repository; used for all data ID queries.
491 collections : `lsst.daf.butler.CollectionSearch`
492 Object representing the collections to search for input datasets.
493 userQuery : `str`, optional
494 User-provided expression to limit the data IDs processed.
496 Returns
497 -------
498 commonDataIds : \
499 `lsst.daf.butler.registry.queries.DataCoordinateQueryResults`
500 An interface to a database temporary table containing all data IDs
501 that will appear in this `QuantumGraph`. Returned inside a
502 context manager, which will drop the temporary table at the end of
503 the `with` block in which this method is called.
504 """
505 _LOG.debug("Building query for data IDs.")
506 # Initialization datasets always have empty data IDs.
507 emptyDataId = DataCoordinate.makeEmpty(registry.dimensions)
508 for datasetType, refs in itertools.chain(self.initInputs.items(),
509 self.initIntermediates.items(),
510 self.initOutputs.items()):
511 refs[emptyDataId] = DatasetRef(datasetType, emptyDataId)
512 # Run one big query for the data IDs for task dimensions and regular
513 # inputs and outputs. We limit the query to only dimensions that are
514 # associated with the input dataset types, but don't (yet) try to
515 # obtain the dataset_ids for those inputs.
516 _LOG.debug("Submitting data ID query and materializing results.")
517 with registry.queryDataIds(self.dimensions,
518 datasets=list(self.inputs),
519 collections=collections,
520 where=userQuery,
521 ).materialize() as commonDataIds:
522 _LOG.debug("Expanding data IDs.")
523 commonDataIds = commonDataIds.expanded()
524 _LOG.debug("Iterating over query results to associate quanta with datasets.")
525 # Iterate over query results, populating data IDs for datasets and
526 # quanta and then connecting them to each other.
527 n = 0
528 for n, commonDataId in enumerate(commonDataIds):
529 # Create DatasetRefs for all DatasetTypes from this result row,
530 # noting that we might have created some already.
531 # We remember both those that already existed and those that we
532 # create now.
533 refsForRow = {}
534 for datasetType, refs in itertools.chain(self.inputs.items(), self.intermediates.items(),
535 self.outputs.items()):
536 datasetDataId = commonDataId.subset(datasetType.dimensions)
537 ref = refs.get(datasetDataId)
538 if ref is None:
539 ref = DatasetRef(datasetType, datasetDataId)
540 refs[datasetDataId] = ref
541 refsForRow[datasetType.name] = ref
542 # Create _QuantumScaffolding objects for all tasks from this result
543 # row, noting that we might have created some already.
544 for task in self.tasks:
545 quantumDataId = commonDataId.subset(task.dimensions)
546 quantum = task.quanta.get(quantumDataId)
547 if quantum is None:
548 quantum = _QuantumScaffolding(task=task, dataId=quantumDataId)
549 task.quanta[quantumDataId] = quantum
550 # Whether this is a new quantum or an existing one, we can now
551 # associate the DatasetRefs for this row with it. The fact
552 # the fact that a Quantum data ID and a dataset data ID both
553 # came from the same result row is what tells us they should
554 # be associated.
555 # Many of these associates will be duplicates (because another
556 # query row that differed from this one only in irrelevant
557 # dimensions already added them), and we use sets to skip.
558 for datasetType in task.inputs:
559 ref = refsForRow[datasetType.name]
560 quantum.inputs[datasetType.name][ref.dataId] = ref
561 for datasetType in task.outputs:
562 ref = refsForRow[datasetType.name]
563 quantum.outputs[datasetType.name][ref.dataId] = ref
564 _LOG.debug("Finished processing %d rows from data ID query.", n)
565 yield commonDataIds
567 def resolveDatasetRefs(self, registry, collections, run, commonDataIds, *, skipExisting=True):
568 """Perform follow up queries for each dataset data ID produced in
569 `fillDataIds`.
571 This method populates `_DatasetScaffolding.refs` (except for those in
572 `prerequisites`).
574 Parameters
575 ----------
576 registry : `lsst.daf.butler.Registry`
577 Registry for the data repository; used for all data ID queries.
578 collections : `lsst.daf.butler.CollectionSearch`
579 Object representing the collections to search for input datasets.
580 run : `str`, optional
581 Name of the `~lsst.daf.butler.CollectionType.RUN` collection for
582 output datasets, if it already exists.
583 commonDataIds : \
584 `lsst.daf.butler.registry.queries.DataCoordinateQueryResults`
585 Result of a previous call to `connectDataIds`.
586 skipExisting : `bool`, optional
587 If `True` (default), a Quantum is not created if all its outputs
588 already exist in ``run``. Ignored if ``run`` is `None`.
590 Raises
591 ------
592 OutputExistsError
593 Raised if an output dataset already exists in the output run
594 and ``skipExisting`` is `False`. The case where some but not all
595 of a quantum's outputs are present and ``skipExisting`` is `True`
596 cannot be identified at this stage, and is handled by `fillQuanta`
597 instead.
598 """
599 # Look up [init] intermediate and output datasets in the output
600 # collection, if there is an output collection.
601 if run is not None:
602 for datasetType, refs in itertools.chain(self.initIntermediates.items(),
603 self.initOutputs.items(),
604 self.intermediates.items(),
605 self.outputs.items()):
606 _LOG.debug("Resolving %d datasets for intermediate and/or output dataset %s.",
607 len(refs), datasetType.name)
608 isInit = datasetType in self.initIntermediates or datasetType in self.initOutputs
609 resolvedRefQueryResults = commonDataIds.subset(
610 datasetType.dimensions,
611 unique=True
612 ).findDatasets(
613 datasetType,
614 collections=run,
615 deduplicate=True
616 )
617 for resolvedRef in resolvedRefQueryResults:
618 # TODO: we could easily support per-DatasetType
619 # skipExisting and I could imagine that being useful - it's
620 # probably required in order to support writing initOutputs
621 # before QuantumGraph generation.
622 assert resolvedRef.dataId in refs
623 if skipExisting or isInit:
624 refs[resolvedRef.dataId] = resolvedRef
625 else:
626 raise OutputExistsError(f"Output dataset {datasetType.name} already exists in "
627 f"output RUN collection '{run}' with data ID"
628 f" {resolvedRef.dataId}.")
629 # Look up input and initInput datasets in the input collection(s).
630 for datasetType, refs in itertools.chain(self.initInputs.items(), self.inputs.items()):
631 _LOG.debug("Resolving %d datasets for input dataset %s.", len(refs), datasetType.name)
632 resolvedRefQueryResults = commonDataIds.subset(
633 datasetType.dimensions,
634 unique=True
635 ).findDatasets(
636 datasetType,
637 collections=collections,
638 deduplicate=True
639 )
640 dataIdsNotFoundYet = set(refs.keys())
641 for resolvedRef in resolvedRefQueryResults:
642 dataIdsNotFoundYet.discard(resolvedRef.dataId)
643 refs[resolvedRef.dataId] = resolvedRef
644 if dataIdsNotFoundYet:
645 raise RuntimeError(
646 f"{len(dataIdsNotFoundYet)} dataset(s) of type "
647 f"'{datasetType.name}' was/were present in a previous "
648 f"query, but could not be found now."
649 f"This is either a logic bug in QuantumGraph generation "
650 f"or the input collections have been modified since "
651 f"QuantumGraph generation began."
652 )
653 # Copy the resolved DatasetRefs to the _QuantumScaffolding objects,
654 # replacing the unresolved refs there, and then look up prerequisites.
655 for task in self.tasks:
656 _LOG.debug(
657 "Applying resolutions and finding prerequisites for %d quanta of task with label '%s'.",
658 len(task.quanta),
659 task.taskDef.label
660 )
661 lookupFunctions = {
662 c.name: c.lookupFunction
663 for c in iterConnections(task.taskDef.connections, "prerequisiteInputs")
664 if c.lookupFunction is not None
665 }
666 dataIdsToSkip = []
667 for quantum in task.quanta.values():
668 # Process outputs datasets only if there is a run to look for
669 # outputs in and skipExisting is True. Note that if
670 # skipExisting is False, any output datasets that already exist
671 # would have already caused an exception to be raised.
672 # We never update the DatasetRefs in the quantum because those
673 # should never be resolved.
674 if run is not None and skipExisting:
675 resolvedRefs = []
676 unresolvedRefs = []
677 for datasetType, originalRefs in quantum.outputs.items():
678 for ref in task.outputs.extract(datasetType, originalRefs.keys()):
679 if ref.id is not None:
680 resolvedRefs.append(ref)
681 else:
682 unresolvedRefs.append(ref)
683 if resolvedRefs:
684 if unresolvedRefs:
685 raise OutputExistsError(
686 f"Quantum {quantum.dataId} of task with label "
687 f"'{quantum.task.taskDef.label}' has some outputs that exist "
688 f"({resolvedRefs}) "
689 f"and others that don't ({unresolvedRefs})."
690 )
691 else:
692 # All outputs are already present; skip this
693 # quantum and continue to the next.
694 dataIdsToSkip.append(quantum.dataId)
695 continue
696 # Update the input DatasetRefs to the resolved ones we already
697 # searched for.
698 for datasetType, refs in quantum.inputs.items():
699 for ref in task.inputs.extract(datasetType, refs.keys()):
700 refs[ref.dataId] = ref
701 # Look up prerequisite datasets in the input collection(s).
702 # These may have dimensions that extend beyond those we queried
703 # for originally, because we want to permit those data ID
704 # values to differ across quanta and dataset types.
705 for datasetType in task.prerequisites:
706 lookupFunction = lookupFunctions.get(datasetType.name)
707 if lookupFunction is not None:
708 # PipelineTask has provided its own function to do the
709 # lookup. This always takes precedence.
710 refs = list(
711 lookupFunction(datasetType, registry, quantum.dataId, collections)
712 )
713 elif (datasetType.isCalibration()
714 and datasetType.dimensions <= quantum.dataId.graph
715 and quantum.dataId.graph.temporal):
716 # This is a master calibration lookup, which we have to
717 # handle specially because the query system can't do a
718 # temporal join on a non-dimension-based timespan yet.
719 timespan = quantum.dataId.timespan
720 try:
721 refs = [registry.findDataset(datasetType, quantum.dataId,
722 collections=collections,
723 timespan=timespan)]
724 except KeyError:
725 # This dataset type is not present in the registry,
726 # which just means there are no datasets here.
727 refs = []
728 else:
729 # Most general case.
730 refs = list(registry.queryDatasets(datasetType,
731 collections=collections,
732 dataId=quantum.dataId,
733 deduplicate=True).expanded())
734 quantum.prerequisites[datasetType].update({ref.dataId: ref for ref in refs
735 if ref is not None})
736 # Actually remove any quanta that we decided to skip above.
737 if dataIdsToSkip:
738 _LOG.debug("Pruning %d quanta for task with label '%s' because all of their outputs exist.",
739 len(dataIdsToSkip), task.taskDef.label)
740 for dataId in dataIdsToSkip:
741 del task.quanta[dataId]
743 def makeQuantumGraph(self):
744 """Create a `QuantumGraph` from the quanta already present in
745 the scaffolding data structure.
747 Returns
748 -------
749 graph : `QuantumGraph`
750 The full `QuantumGraph`.
751 """
752 graph = QuantumGraph({task.taskDef: task.makeQuantumSet() for task in self.tasks})
753 return graph
756class _InstrumentFinder(TreeVisitor):
757 """Implementation of TreeVisitor which looks for instrument name
759 Instrument should be specified as a boolean expression
761 instrument = 'string'
762 'string' = instrument
764 so we only need to find a binary operator where operator is "=",
765 one side is a string literal and other side is an identifier.
766 All visit methods return tuple of (type, value), non-useful nodes
767 return None for both type and value.
768 """
769 def __init__(self):
770 self.instruments = []
772 def visitNumericLiteral(self, value, node):
773 # do not care about numbers
774 return (None, None)
776 def visitStringLiteral(self, value, node):
777 # return type and value
778 return ("str", value)
780 def visitTimeLiteral(self, value, node):
781 # do not care about these
782 return (None, None)
784 def visitRangeLiteral(self, start, stop, stride, node):
785 # do not care about these
786 return (None, None)
788 def visitIdentifier(self, name, node):
789 if name.lower() == "instrument":
790 return ("id", "instrument")
791 return (None, None)
793 def visitUnaryOp(self, operator, operand, node):
794 # do not care about these
795 return (None, None)
797 def visitBinaryOp(self, operator, lhs, rhs, node):
798 if operator == "=":
799 if lhs == ("id", "instrument") and rhs[0] == "str":
800 self.instruments.append(rhs[1])
801 elif rhs == ("id", "instrument") and lhs[0] == "str":
802 self.instruments.append(lhs[1])
803 return (None, None)
805 def visitIsIn(self, lhs, values, not_in, node):
806 # do not care about these
807 return (None, None)
809 def visitParens(self, expression, node):
810 # do not care about these
811 return (None, None)
814def _findInstruments(queryStr):
815 """Get the names of any instrument named in the query string by searching
816 for "instrument = <value>" and similar patterns.
818 Parameters
819 ----------
820 queryStr : `str` or None
821 The query string to search, or None if there is no query.
823 Returns
824 -------
825 instruments : `list` [`str`]
826 The list of instrument names found in the query.
828 Raises
829 ------
830 ValueError
831 If the query expression can not be parsed.
832 """
833 if not queryStr:
834 return []
835 parser = ParserYacc()
836 finder = _InstrumentFinder()
837 try:
838 tree = parser.parse(queryStr)
839 except ParseError as exc:
840 raise ValueError(f"failed to parse query expression: {queryStr}") from exc
841 tree.visit(finder)
842 return finder.instruments
845# ------------------------
846# Exported definitions --
847# ------------------------
850class GraphBuilderError(Exception):
851 """Base class for exceptions generated by graph builder.
852 """
853 pass
856class OutputExistsError(GraphBuilderError):
857 """Exception generated when output datasets already exist.
858 """
859 pass
862class PrerequisiteMissingError(GraphBuilderError):
863 """Exception generated when a prerequisite dataset does not exist.
864 """
865 pass
868class GraphBuilder(object):
869 """GraphBuilder class is responsible for building task execution graph from
870 a Pipeline.
872 Parameters
873 ----------
874 registry : `~lsst.daf.butler.Registry`
875 Data butler instance.
876 skipExisting : `bool`, optional
877 If `True` (default), a Quantum is not created if all its outputs
878 already exist.
879 """
881 def __init__(self, registry, skipExisting=True):
882 self.registry = registry
883 self.dimensions = registry.dimensions
884 self.skipExisting = skipExisting
886 def makeGraph(self, pipeline, collections, run, userQuery):
887 """Create execution graph for a pipeline.
889 Parameters
890 ----------
891 pipeline : `Pipeline`
892 Pipeline definition, task names/classes and their configs.
893 collections : `lsst.daf.butler.CollectionSearch`
894 Object representing the collections to search for input datasets.
895 run : `str`, optional
896 Name of the `~lsst.daf.butler.CollectionType.RUN` collection for
897 output datasets, if it already exists.
898 userQuery : `str`
899 String which defines user-defined selection for registry, should be
900 empty or `None` if there is no restrictions on data selection.
902 Returns
903 -------
904 graph : `QuantumGraph`
906 Raises
907 ------
908 UserExpressionError
909 Raised when user expression cannot be parsed.
910 OutputExistsError
911 Raised when output datasets already exist.
912 Exception
913 Other exceptions types may be raised by underlying registry
914 classes.
915 """
916 scaffolding = _PipelineScaffolding(pipeline, registry=self.registry)
918 instrument = pipeline.getInstrument()
919 if isinstance(instrument, str):
920 instrument = doImport(instrument)
921 instrumentName = instrument.getName() if instrument else None
922 userQuery = self._verifyInstrumentRestriction(instrumentName, userQuery)
924 with scaffolding.connectDataIds(self.registry, collections, userQuery) as commonDataIds:
925 scaffolding.resolveDatasetRefs(self.registry, collections, run, commonDataIds,
926 skipExisting=self.skipExisting)
927 return scaffolding.makeQuantumGraph()
929 @staticmethod
930 def _verifyInstrumentRestriction(instrumentName, query):
931 """Add an instrument restriction to the query if it does not have one,
932 and verify that if given an instrument name that there are no other
933 instrument restrictions in the query.
935 Parameters
936 ----------
937 instrumentName : `str`
938 The name of the instrument that should appear in the query.
939 query : `str`
940 The query string.
942 Returns
943 -------
944 query : `str`
945 The query string with the instrument added to it if needed.
947 Raises
948 ------
949 RuntimeError
950 If the pipeline names an instrument and the query contains more
951 than one instrument or the name of the instrument in the query does
952 not match the instrument named by the pipeline.
953 """
954 if not instrumentName:
955 return query
956 queryInstruments = _findInstruments(query)
957 if len(queryInstruments) > 1:
958 raise RuntimeError(f"When the pipeline has an instrument (\"{instrumentName}\") the query must "
959 "have zero instruments or one instrument that matches the pipeline. "
960 f"Found these instruments in the query: {queryInstruments}.")
961 if not queryInstruments:
962 # There is not an instrument in the query, add it:
963 restriction = f"instrument = '{instrumentName}'"
964 _LOG.debug(f"Adding restriction \"{restriction}\" to query.")
965 query = f"{restriction} AND ({query})" if query else restriction # (there may not be a query)
966 elif queryInstruments[0] != instrumentName:
967 # Since there is an instrument in the query, it should match
968 # the instrument in the pipeline.
969 raise RuntimeError(f"The instrument named in the query (\"{queryInstruments[0]}\") does not "
970 f"match the instrument named by the pipeline (\"{instrumentName}\")")
971 return query