Coverage for python/lsst/daf/butler/registry/dimensions/query.py: 81%

66 statements  

« prev     ^ index     » next       coverage.py v6.4.4, created at 2022-09-22 02:04 -0700

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 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 

22 

23__all__ = ["QueryDimensionRecordStorage"] 

24 

25from typing import Any, Iterable, Mapping, Optional 

26 

27import sqlalchemy 

28 

29from ...core import ( 

30 DatabaseDimension, 

31 DatabaseDimensionElement, 

32 DataCoordinateIterable, 

33 DimensionElement, 

34 DimensionRecord, 

35 GovernorDimension, 

36 NamedKeyDict, 

37 NamedKeyMapping, 

38 TimespanDatabaseRepresentation, 

39) 

40from ..interfaces import ( 

41 Database, 

42 DatabaseDimensionRecordStorage, 

43 GovernorDimensionRecordStorage, 

44 StaticTablesContext, 

45) 

46from ..queries import QueryBuilder 

47 

48 

49class QueryDimensionRecordStorage(DatabaseDimensionRecordStorage): 

50 """A read-only record storage implementation backed by SELECT query. 

51 

52 At present, the only query this class supports is a SELECT DISTINCT over 

53 the table for some other dimension that has this dimension as an implied 

54 dependency. For example, we can use this class to provide access to the 

55 set of ``band`` names referenced by any ``physical_filter``. 

56 

57 Parameters 

58 ---------- 

59 db : `Database` 

60 Interface to the database engine and namespace that will hold these 

61 dimension records. 

62 element : `DatabaseDimensionElement` 

63 The element whose records this storage will manage. 

64 """ 

65 

66 def __init__(self, db: Database, element: DatabaseDimensionElement, viewOf: str): 

67 assert isinstance( 

68 element, DatabaseDimension 

69 ), "An element cannot be a dependency unless it is a dimension." 

70 self._db = db 

71 self._element = element 

72 self._target = element.universe[viewOf] 

73 self._targetSpec = self._target.RecordClass.fields.makeTableSpec( 

74 TimespanReprClass=self._db.getTimespanRepresentation() 

75 ) 

76 self._viewOf = viewOf 

77 self._query = None # Constructed on first use. 

78 if element not in self._target.graph.dimensions: 78 ↛ 79line 78 didn't jump to line 79, because the condition on line 78 was never true

79 raise NotImplementedError("Query-backed dimension must be a dependency of its target.") 

80 if element.metadata: 80 ↛ 81line 80 didn't jump to line 81, because the condition on line 80 was never true

81 raise NotImplementedError("Cannot use query to back dimension with metadata.") 

82 if element.implied: 82 ↛ 83line 82 didn't jump to line 83, because the condition on line 82 was never true

83 raise NotImplementedError("Cannot use query to back dimension with implied dependencies.") 

84 if element.alternateKeys: 84 ↛ 85line 84 didn't jump to line 85, because the condition on line 84 was never true

85 raise NotImplementedError("Cannot use query to back dimension with alternate unique keys.") 

86 if element.spatial is not None: 86 ↛ 87line 86 didn't jump to line 87, because the condition on line 86 was never true

87 raise NotImplementedError("Cannot use query to back spatial dimension.") 

88 if element.temporal is not None: 88 ↛ 89line 88 didn't jump to line 89, because the condition on line 88 was never true

89 raise NotImplementedError("Cannot use query to back temporal dimension.") 

90 

91 @classmethod 

92 def initialize( 

93 cls, 

94 db: Database, 

95 element: DatabaseDimensionElement, 

96 *, 

97 context: Optional[StaticTablesContext] = None, 

98 config: Mapping[str, Any], 

99 governors: NamedKeyMapping[GovernorDimension, GovernorDimensionRecordStorage], 

100 ) -> DatabaseDimensionRecordStorage: 

101 # Docstring inherited from DatabaseDimensionRecordStorage. 

102 viewOf = config["view_of"] 

103 return cls(db, element, viewOf) 

104 

105 @property 

106 def element(self) -> DatabaseDimension: 

107 # Docstring inherited from DimensionRecordStorage.element. 

108 return self._element 

109 

110 def clearCaches(self) -> None: 

111 # Docstring inherited from DimensionRecordStorage.clearCaches. 

112 pass 

113 

114 def _ensureQuery(self) -> None: 

115 if self._query is None: 

116 targetTable = self._db.getExistingTable(self._target.name, self._targetSpec) 

117 assert targetTable is not None 

118 columns = [] 

119 # The only columns for this dimension are ones for its required 

120 # dependencies and its own primary key (guaranteed by the checks in 

121 # the ctor). 

122 for dimension in self.element.required: 

123 if dimension == self.element: 123 ↛ 126line 123 didn't jump to line 126, because the condition on line 123 was never false

124 columns.append(targetTable.columns[dimension.name].label(dimension.primaryKey.name)) 

125 else: 

126 columns.append(targetTable.columns[dimension.name].label(dimension.name)) 

127 # This query doesn't do a SELECT DISTINCT, because that's confusing 

128 # and potentially wasteful if we apply a restrictive WHERE clause, 

129 # as SelectableDimensionRecordStorage.fetch will do. 

130 # Instead, we add DISTINCT in join() only. 

131 self._query = ( 

132 sqlalchemy.sql.select(*columns).distinct().select_from(targetTable).alias(self.element.name) 

133 ) 

134 

135 def join( 

136 self, 

137 builder: QueryBuilder, 

138 *, 

139 regions: Optional[NamedKeyDict[DimensionElement, sqlalchemy.sql.ColumnElement]] = None, 

140 timespans: Optional[NamedKeyDict[DimensionElement, TimespanDatabaseRepresentation]] = None, 

141 ) -> None: 

142 # Docstring inherited from DimensionRecordStorage. 

143 assert regions is None, "Should be guaranteed by constructor checks." 

144 assert timespans is None, "Should be guaranteed by constructor checks." 

145 if self._target in builder.summary.mustHaveKeysJoined: 145 ↛ 150line 145 didn't jump to line 150, because the condition on line 145 was never true

146 # Do nothing; the target dimension is already being included, so 

147 # joining against a subquery referencing it would just produce a 

148 # more complicated query that's guaranteed to return the same 

149 # results. 

150 return 

151 self._ensureQuery() 

152 joinOn = builder.startJoin( 

153 self._query, self.element.required, self.element.RecordClass.fields.required.names 

154 ) 

155 builder.finishJoin(self._query, joinOn) 

156 return self._query 

157 

158 def insert(self, *records: DimensionRecord, replace: bool = False, skip_existing: bool = False) -> None: 

159 # Docstring inherited from DimensionRecordStorage.insert. 

160 raise TypeError( 

161 f"Cannot insert {self.element.name} records, define as part of {self._viewOf} instead." 

162 ) 

163 

164 def sync(self, record: DimensionRecord, update: bool = False) -> bool: 

165 # Docstring inherited from DimensionRecordStorage.sync. 

166 raise TypeError(f"Cannot sync {self.element.name} records, define as part of {self._viewOf} instead.") 

167 

168 def fetch(self, dataIds: DataCoordinateIterable) -> Iterable[DimensionRecord]: 

169 # Docstring inherited from DimensionRecordStorage.fetch. 

170 RecordClass = self.element.RecordClass 

171 for dataId in dataIds: 

172 # Given the restrictions imposed at construction, we know there's 

173 # nothing to actually fetch: everything we need is in the data ID. 

174 yield RecordClass(**dataId.byName()) 

175 

176 def digestTables(self) -> Iterable[sqlalchemy.schema.Table]: 

177 # Docstring inherited from DimensionRecordStorage.digestTables. 

178 return []