Hide keyboard shortcuts

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/>. 

21 

22"""Module defining a butler like object specialized to a specific quantum. 

23""" 

24 

25__all__ = ("ButlerQuantumContext",) 

26 

27import types 

28import typing 

29 

30from .connections import InputQuantizedConnection, OutputQuantizedConnection, DeferredDatasetRef 

31from .struct import Struct 

32from lsst.daf.butler import DatasetRef, Butler, Quantum 

33 

34 

35class ButlerQuantumContext: 

36 """A Butler-like class specialized for a single quantum 

37 

38 A ButlerQuantumContext class wraps a standard butler interface and 

39 specializes it to the context of a given quantum. What this means 

40 in practice is that the only gets and puts that this class allows 

41 are DatasetRefs that are contained in the quantum. 

42 

43 In the future this class will also be used to record provenance on 

44 what was actually get and put. This is in contrast to what the 

45 preflight expects to be get and put by looking at the graph before 

46 execution. 

47 

48 Parameters 

49 ---------- 

50 butler : `lsst.daf.butler.Butler` 

51 Butler object from/to which datasets will be get/put 

52 quantum : `lsst.daf.butler.core.Quantum` 

53 Quantum object that describes the datasets which will be get/put by a 

54 single execution of this node in the pipeline graph. All input 

55 dataset references must be resolved (i.e. satisfy 

56 ``DatasetRef.id is not None``) prior to constructing the 

57 `ButlerQuantumContext`. 

58 

59 Notes 

60 ----- 

61 Most quanta in any non-trivial graph will not start with resolved dataset 

62 references, because they represent processing steps that can only run 

63 after some other quanta have produced their inputs. At present, it is the 

64 responsibility of ``lsst.ctrl.mpexec.SingleQuantumExecutor`` to resolve all 

65 datasets prior to constructing `ButlerQuantumContext` and calling 

66 `runQuantum`, and the fact that this precondition is satisfied by code in 

67 a downstream package is considered a problem with the 

68 ``pipe_base/ctrl_mpexec`` separation of concerns that will be addressed in 

69 the future. 

70 """ 

71 def __init__(self, butler: Butler, quantum: Quantum): 

72 self.quantum = quantum 

73 self.registry = butler.registry 

74 self.allInputs = set() 

75 self.allOutputs = set() 

76 for refs in quantum.inputs.values(): 

77 for ref in refs: 

78 self.allInputs.add((ref.datasetType, ref.dataId)) 

79 for refs in quantum.outputs.values(): 

80 for ref in refs: 

81 self.allOutputs.add((ref.datasetType, ref.dataId)) 

82 

83 # Create closures over butler to discourage anyone from directly 

84 # using a butler reference 

85 def _get(self, ref): 

86 # Butler methods below will check for unresolved DatasetRefs and 

87 # raise appropriately, so no need for us to do that here. 

88 if isinstance(ref, DeferredDatasetRef): 

89 self._checkMembership(ref.datasetRef, self.allInputs) 

90 return butler.getDirectDeferred(ref.datasetRef) 

91 

92 else: 

93 self._checkMembership(ref, self.allInputs) 

94 return butler.getDirect(ref) 

95 

96 def _put(self, value, ref): 

97 self._checkMembership(ref, self.allOutputs) 

98 butler.put(value, ref) 

99 

100 self._get = types.MethodType(_get, self) 

101 self._put = types.MethodType(_put, self) 

102 

103 def get(self, dataset: typing.Union[InputQuantizedConnection, 

104 typing.List[DatasetRef], 

105 DatasetRef]) -> object: 

106 """Fetches data from the butler 

107 

108 Parameters 

109 ---------- 

110 dataset 

111 This argument may either be an `InputQuantizedConnection` which describes 

112 all the inputs of a quantum, a list of `~lsst.daf.butler.DatasetRef`, or 

113 a single `~lsst.daf.butler.DatasetRef`. The function will get and return 

114 the corresponding datasets from the butler. 

115 

116 Returns 

117 ------- 

118 return : `object` 

119 This function returns arbitrary objects fetched from the bulter. The 

120 structure these objects are returned in depends on the type of the input 

121 argument. If the input dataset argument is a InputQuantizedConnection, then 

122 the return type will be a dictionary with keys corresponding to the attributes 

123 of the `InputQuantizedConnection` (which in turn are the attribute identifiers 

124 of the connections). If the input argument is of type `list` of 

125 `~lsst.daf.butler.DatasetRef` then the return type will be a list of objects. 

126 If the input argument is a single `~lsst.daf.butler.DatasetRef` then a single 

127 object will be returned. 

128 

129 Raises 

130 ------ 

131 ValueError 

132 If a `DatasetRef` is passed to get that is not defined in the quantum object 

133 """ 

134 if isinstance(dataset, InputQuantizedConnection): 

135 retVal = {} 

136 for name, ref in dataset: 

137 if isinstance(ref, list): 

138 val = [self._get(r) for r in ref] 

139 else: 

140 val = self._get(ref) 

141 retVal[name] = val 

142 return retVal 

143 elif isinstance(dataset, list): 

144 return [self._get(x) for x in dataset] 

145 elif isinstance(dataset, DatasetRef) or isinstance(dataset, DeferredDatasetRef): 

146 return self._get(dataset) 

147 else: 

148 raise TypeError("Dataset argument is not a type that can be used to get") 

149 

150 def put(self, values: typing.Union[Struct, typing.List[typing.Any], object], 

151 dataset: typing.Union[OutputQuantizedConnection, typing.List[DatasetRef], DatasetRef]): 

152 """Puts data into the butler 

153 

154 Parameters 

155 ---------- 

156 values : `Struct` or `list` of `object` or `object` 

157 The data that should be put with the butler. If the type of the dataset 

158 is `OutputQuantizedConnection` then this argument should be a `Struct` 

159 with corresponding attribute names. Each attribute should then correspond 

160 to either a list of object or a single object depending of the type of the 

161 corresponding attribute on dataset. I.e. if dataset.calexp is [datasetRef1, 

162 datasetRef2] then values.calexp should be [calexp1, calexp2]. Like wise 

163 if there is a single ref, then only a single object need be passed. The same 

164 restriction applies if dataset is directly a `list` of `DatasetRef` or a 

165 single `DatasetRef`. 

166 dataset 

167 This argument may either be an `InputQuantizedConnection` which describes 

168 all the inputs of a quantum, a list of `lsst.daf.butler.DatasetRef`, or 

169 a single `lsst.daf.butler.DatasetRef`. The function will get and return 

170 the corresponding datasets from the butler. 

171 

172 Raises 

173 ------ 

174 ValueError 

175 If a `DatasetRef` is passed to put that is not defined in the quantum object, or 

176 the type of values does not match what is expected from the type of dataset. 

177 """ 

178 if isinstance(dataset, OutputQuantizedConnection): 

179 if not isinstance(values, Struct): 

180 raise ValueError("dataset is a OutputQuantizedConnection, a Struct with corresponding" 

181 " attributes must be passed as the values to put") 

182 for name, refs in dataset: 

183 valuesAttribute = getattr(values, name) 

184 if isinstance(refs, list): 

185 if len(refs) != len(valuesAttribute): 

186 raise ValueError(f"There must be a object to put for every Dataset ref in {name}") 

187 for i, ref in enumerate(refs): 

188 self._put(valuesAttribute[i], ref) 

189 else: 

190 self._put(valuesAttribute, refs) 

191 elif isinstance(dataset, list): 

192 if len(dataset) != len(values): 

193 raise ValueError("There must be a common number of references and values to put") 

194 for i, ref in enumerate(dataset): 

195 self._put(values[i], ref) 

196 elif isinstance(dataset, DatasetRef): 

197 self._put(values, dataset) 

198 else: 

199 raise TypeError("Dataset argument is not a type that can be used to put") 

200 

201 def _checkMembership(self, ref: typing.Union[typing.List[DatasetRef], DatasetRef], inout: set): 

202 """Internal function used to check if a DatasetRef is part of the input quantum 

203 

204 This function will raise an exception if the ButlerQuantumContext is used to 

205 get/put a DatasetRef which is not defined in the quantum. 

206 

207 Parameters 

208 ---------- 

209 ref : `list` of `DatasetRef` or `DatasetRef` 

210 Either a list or a single `DatasetRef` to check 

211 inout : `set` 

212 The connection type to check, e.g. either an input or an output. This prevents 

213 both types needing to be checked for every operation, which may be important 

214 for Quanta with lots of `DatasetRef`s. 

215 """ 

216 if not isinstance(ref, list): 

217 ref = [ref] 

218 for r in ref: 

219 if (r.datasetType, r.dataId) not in inout: 

220 raise ValueError("DatasetRef is not part of the Quantum being processed")