Coverage for python/lsst/pipe/base/graph/_implDetails.py : 47%

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
22from collections import defaultdict
24__all__ = ("_DatasetTracker", "DatasetTypeName")
26from dataclasses import dataclass, field
27import networkx as nx
28from typing import (DefaultDict, Generic, Optional, Set, TypeVar, Generator, Tuple, NewType,)
30from lsst.daf.butler import DatasetRef
32from .quantumNode import QuantumNode
33from ..pipeline import TaskDef
35# NewTypes
36DatasetTypeName = NewType("DatasetTypeName", str)
38# Generic type parameters
39_T = TypeVar("_T", DatasetTypeName, DatasetRef)
40_U = TypeVar("_U", TaskDef, QuantumNode)
43@dataclass
44class _DatasetTrackerElement(Generic[_U]):
45 inputs: Set[_U] = field(default_factory=set)
46 output: Optional[_U] = None
49class _DatasetTracker(Generic[_T, _U]):
50 def __init__(self):
51 self._container: DefaultDict[_T, _DatasetTrackerElement[_U]] = defaultdict(_DatasetTrackerElement)
53 def addInput(self, key: _T, value: _U):
54 self._container[key].inputs.add(value)
56 def addOutput(self, key: _T, value: _U):
57 element = self._container[key]
58 if element.output is not None:
59 raise ValueError(f"Only one output for key {key} is allowed, "
60 f"the current output is set to {element.output}")
61 element.output = value
63 def getInputs(self, key: _T) -> Set[_U]:
64 return self._container[key].inputs
66 def getOutput(self, key: _T) -> Optional[_U]:
67 return self._container[key].output
69 def getAll(self, key: _T) -> Set[_U]:
70 output = self._container[key].output
71 if output is not None:
72 return self._container[key].inputs.union((output,))
73 return set(self._container[key].inputs)
75 def makeNetworkXGraph(self) -> nx.DiGraph:
76 graph = nx.DiGraph()
77 graph.add_edges_from(self._datasetDictToEdgeIterator())
78 if None in graph.nodes():
79 graph.remove_node(None)
80 return graph
82 def _datasetDictToEdgeIterator(self) -> Generator[Tuple[Optional[_U], Optional[_U]], None, None]:
83 """Helper function designed to be used in conjunction with
84 `networkx.DiGraph.add_edges_from`. This takes a mapping of keys to
85 `_DatasetTrackers` and yields successive pairs of elements that are to
86 be considered connected by the graph.
87 """
88 for entry in self._container.values():
89 # If there is no inputs and only outputs (likely in test cases or
90 # building inits or something) use None as a Node, that will then
91 # be removed later
92 inputs = entry.inputs or (None,)
93 for inpt in inputs:
94 yield (entry.output, inpt)
96 def keys(self) -> Set[_T]:
97 return set(self._container.keys())