Coverage for python/lsst/dax/apdb/apdbSqlSchema.py: 18%

211 statements  

« prev     ^ index     » next       coverage.py v7.2.5, created at 2023-05-23 08:50 +0000

1# This file is part of dax_apdb. 

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 responsible for APDB schema operations. 

23""" 

24 

25from __future__ import annotations 

26 

27__all__ = ["ApdbSqlSchema", "ExtraTables"] 

28 

29import enum 

30import logging 

31import uuid 

32from typing import Any, Dict, List, Mapping, Optional, Type 

33 

34import felis.types 

35import sqlalchemy 

36from felis import simple 

37from sqlalchemy import ( 

38 DDL, 

39 Column, 

40 ForeignKeyConstraint, 

41 Index, 

42 MetaData, 

43 PrimaryKeyConstraint, 

44 Table, 

45 UniqueConstraint, 

46 event, 

47 inspect, 

48) 

49from sqlalchemy.dialects.postgresql import UUID 

50 

51from .apdbSchema import ApdbSchema, ApdbTables 

52 

53_LOG = logging.getLogger(__name__) 

54 

55 

56# 

57# Copied from daf_butler. 

58# 

59class GUID(sqlalchemy.TypeDecorator): 

60 """Platform-independent GUID type. 

61 

62 Uses PostgreSQL's UUID type, otherwise uses CHAR(32), storing as 

63 stringified hex values. 

64 """ 

65 

66 impl = sqlalchemy.CHAR 

67 

68 cache_ok = True 

69 

70 def load_dialect_impl(self, dialect: sqlalchemy.engine.Dialect) -> sqlalchemy.types.TypeEngine: 

71 if dialect.name == "postgresql": 

72 return dialect.type_descriptor(UUID()) 

73 else: 

74 return dialect.type_descriptor(sqlalchemy.CHAR(32)) 

75 

76 def process_bind_param(self, value: Any, dialect: sqlalchemy.engine.Dialect) -> Optional[str]: 

77 if value is None: 

78 return value 

79 

80 # Coerce input to UUID type, in general having UUID on input is the 

81 # only thing that we want but there is code right now that uses ints. 

82 if isinstance(value, int): 

83 value = uuid.UUID(int=value) 

84 elif isinstance(value, bytes): 

85 value = uuid.UUID(bytes=value) 

86 elif isinstance(value, str): 

87 # hexstring 

88 value = uuid.UUID(hex=value) 

89 elif not isinstance(value, uuid.UUID): 

90 raise TypeError(f"Unexpected type of a bind value: {type(value)}") 

91 

92 if dialect.name == "postgresql": 

93 return str(value) 

94 else: 

95 return "%.32x" % value.int 

96 

97 def process_result_value( 

98 self, value: str | uuid.UUID | None, dialect: sqlalchemy.engine.Dialect 

99 ) -> Optional[uuid.UUID]: 

100 if value is None: 

101 return value 

102 elif isinstance(value, uuid.UUID): 

103 # sqlalchemy 2 converts to UUID internally 

104 return value 

105 else: 

106 return uuid.UUID(hex=value) 

107 

108 

109@enum.unique 

110class ExtraTables(enum.Enum): 

111 """Names of the tables used for tracking insert IDs.""" 

112 

113 DiaInsertId = "DiaInsertId" 

114 """Name of the table for insert ID records.""" 

115 

116 DiaObjectInsertId = "DiaObjectInsertId" 

117 """Name of the table for DIAObject insert ID records.""" 

118 

119 DiaSourceInsertId = "DiaSourceInsertId" 

120 """Name of the table for DIASource insert ID records.""" 

121 

122 DiaForcedSourceInsertId = "DiaFSourceInsertId" 

123 """Name of the table for DIAForcedSource insert ID records.""" 

124 

125 def table_name(self, prefix: str = "") -> str: 

126 """Return full table name.""" 

127 return prefix + self.value 

128 

129 @classmethod 

130 def insert_id_tables(cls) -> Mapping[ExtraTables, ApdbTables]: 

131 """Return mapping of tables used for insert ID tracking to their 

132 corresponding regular tables.""" 

133 return { 

134 cls.DiaObjectInsertId: ApdbTables.DiaObject, 

135 cls.DiaSourceInsertId: ApdbTables.DiaSource, 

136 cls.DiaForcedSourceInsertId: ApdbTables.DiaForcedSource, 

137 } 

138 

139 

140class ApdbSqlSchema(ApdbSchema): 

141 """Class for management of APDB schema. 

142 

143 Attributes 

144 ---------- 

145 objects : `sqlalchemy.Table` 

146 DiaObject table instance 

147 objects_last : `sqlalchemy.Table` 

148 DiaObjectLast table instance, may be None 

149 sources : `sqlalchemy.Table` 

150 DiaSource table instance 

151 forcedSources : `sqlalchemy.Table` 

152 DiaForcedSource table instance 

153 has_insert_id : `bool` 

154 If true then schema has tables for insert ID tracking. 

155 

156 Parameters 

157 ---------- 

158 engine : `sqlalchemy.engine.Engine` 

159 SQLAlchemy engine instance 

160 dia_object_index : `str` 

161 Indexing mode for DiaObject table, see `ApdbSqlConfig.dia_object_index` 

162 for details. 

163 htm_index_column : `str` 

164 Name of a HTM index column for DiaObject and DiaSource tables. 

165 schema_file : `str` 

166 Name of the YAML schema file. 

167 schema_name : `str`, optional 

168 Name of the schema in YAML files. 

169 prefix : `str`, optional 

170 Prefix to add to all schema elements. 

171 namespace : `str`, optional 

172 Namespace (or schema name) to use for all APDB tables. 

173 use_insert_id : `bool`, optional 

174 

175 """ 

176 

177 pixel_id_tables = (ApdbTables.DiaObject, ApdbTables.DiaObjectLast, ApdbTables.DiaSource) 

178 """Tables that need pixelId column for spatial indexing.""" 

179 

180 def __init__( 

181 self, 

182 engine: sqlalchemy.engine.Engine, 

183 dia_object_index: str, 

184 htm_index_column: str, 

185 schema_file: str, 

186 schema_name: str = "ApdbSchema", 

187 prefix: str = "", 

188 namespace: Optional[str] = None, 

189 use_insert_id: bool = False, 

190 ): 

191 super().__init__(schema_file, schema_name) 

192 

193 self._engine = engine 

194 self._dia_object_index = dia_object_index 

195 self._htm_index_column = htm_index_column 

196 self._prefix = prefix 

197 self._use_insert_id = use_insert_id 

198 

199 self._metadata = MetaData(schema=namespace) 

200 

201 # map YAML column types to SQLAlchemy 

202 self._type_map = { 

203 felis.types.Double: self._getDoubleType(engine), 

204 felis.types.Float: sqlalchemy.types.Float, 

205 felis.types.Timestamp: sqlalchemy.types.TIMESTAMP, 

206 felis.types.Long: sqlalchemy.types.BigInteger, 

207 felis.types.Int: sqlalchemy.types.Integer, 

208 felis.types.Short: sqlalchemy.types.Integer, 

209 felis.types.Byte: sqlalchemy.types.Integer, 

210 felis.types.Binary: sqlalchemy.types.LargeBinary, 

211 felis.types.Text: sqlalchemy.types.CHAR, 

212 felis.types.String: sqlalchemy.types.CHAR, 

213 felis.types.Char: sqlalchemy.types.CHAR, 

214 felis.types.Unicode: sqlalchemy.types.CHAR, 

215 felis.types.Boolean: sqlalchemy.types.Boolean, 

216 } 

217 

218 # Add pixelId column and index to tables that need it 

219 for table in self.pixel_id_tables: 

220 tableDef = self.tableSchemas.get(table) 

221 if not tableDef: 

222 continue 

223 column = simple.Column( 

224 id=f"#{htm_index_column}", 

225 name=htm_index_column, 

226 datatype=felis.types.Long, 

227 nullable=False, 

228 value=None, 

229 description="Pixelization index column.", 

230 table=tableDef, 

231 ) 

232 tableDef.columns.append(column) 

233 

234 # Adjust index if needed 

235 if table == ApdbTables.DiaObject and self._dia_object_index == "pix_id_iov": 

236 tableDef.primary_key.insert(0, column) 

237 

238 if table is ApdbTables.DiaObjectLast: 

239 # use it as a leading PK column 

240 tableDef.primary_key.insert(0, column) 

241 else: 

242 # make a regular index 

243 name = f"IDX_{tableDef.name}_{htm_index_column}" 

244 index = simple.Index(id=f"#{name}", name=name, columns=[column]) 

245 tableDef.indexes.append(index) 

246 

247 # generate schema for all tables, must be called last 

248 self._apdb_tables = self._make_apdb_tables() 

249 self._extra_tables = self._make_extra_tables(self._apdb_tables) 

250 

251 self._has_insert_id: bool | None = None 

252 

253 def makeSchema(self, drop: bool = False) -> None: 

254 """Create or re-create all tables. 

255 

256 Parameters 

257 ---------- 

258 drop : `bool`, optional 

259 If True then drop tables before creating new ones. 

260 """ 

261 # Create namespace if it does not exist yet, for now this only makes 

262 # sense for postgres. 

263 if self._metadata.schema: 

264 dialect = self._engine.dialect 

265 quoted_schema = dialect.preparer(dialect).quote_schema(self._metadata.schema) 

266 create_schema = DDL( 

267 "CREATE SCHEMA IF NOT EXISTS %(schema)s", context={"schema": quoted_schema} 

268 ).execute_if(dialect="postgresql") 

269 event.listen(self._metadata, "before_create", create_schema) 

270 

271 # create all tables (optionally drop first) 

272 if drop: 

273 _LOG.info("dropping all tables") 

274 self._metadata.drop_all(self._engine) 

275 _LOG.info("creating all tables") 

276 self._metadata.create_all(self._engine) 

277 

278 # Reset possibly cached value. 

279 self._has_insert_id = None 

280 

281 def get_table(self, table_enum: ApdbTables | ExtraTables) -> Table: 

282 """Return SQLAlchemy table instance for a specified table type/enum. 

283 

284 Parameters 

285 ---------- 

286 table_enum : `ApdbTables` or `ExtraTables` 

287 Type of table to return. 

288 

289 Returns 

290 ------- 

291 table : `sqlalchemy.schema.Table` 

292 Table instance. 

293 

294 Raises 

295 ------ 

296 ValueError 

297 Raised if ``table_enum`` is not valid for this database. 

298 """ 

299 try: 

300 if isinstance(table_enum, ApdbTables): 

301 return self._apdb_tables[table_enum] 

302 else: 

303 return self._extra_tables[table_enum] 

304 except LookupError: 

305 raise ValueError(f"Table type {table_enum} does not exist in the schema") from None 

306 

307 def get_apdb_columns(self, table_enum: ApdbTables | ExtraTables) -> list[Column]: 

308 """Return list of columns defined for a table in APDB schema. 

309 

310 Returned list excludes columns that are implementation-specific, e.g. 

311 ``pixelId`` column is not include in the returned list. 

312 

313 Parameters 

314 ---------- 

315 table_enum : `ApdbTables` or `ExtraTables` 

316 Type of table. 

317 

318 Returns 

319 ------- 

320 table : `list` [`sqlalchemy.schema.Column`] 

321 Table instance. 

322 

323 Raises 

324 ------ 

325 ValueError 

326 Raised if ``table_enum`` is not valid for this database. 

327 """ 

328 table = self.get_table(table_enum) 

329 exclude_columns = set() 

330 if table_enum in self.pixel_id_tables: 

331 exclude_columns.add(self._htm_index_column) 

332 return [column for column in table.columns if column.name not in exclude_columns] 

333 

334 @property 

335 def has_insert_id(self) -> bool: 

336 """Whether insert ID tables are to be used (`bool`).""" 

337 if self._has_insert_id is None: 

338 self._has_insert_id = self._use_insert_id and self._check_insert_id() 

339 return self._has_insert_id 

340 

341 def _check_insert_id(self) -> bool: 

342 """Check whether database has tables for tracking insert IDs.""" 

343 inspector = inspect(self._engine) 

344 db_tables = set(inspector.get_table_names(schema=self._metadata.schema)) 

345 return ExtraTables.DiaInsertId.table_name(self._prefix) in db_tables 

346 

347 def _make_apdb_tables(self, mysql_engine: str = "InnoDB") -> Mapping[ApdbTables, Table]: 

348 """Generate schema for regular tables. 

349 

350 Parameters 

351 ---------- 

352 mysql_engine : `str`, optional 

353 MySQL engine type to use for new tables. 

354 """ 

355 tables = {} 

356 for table_enum in ApdbTables: 

357 if table_enum is ApdbTables.DiaObjectLast and self._dia_object_index != "last_object_table": 

358 continue 

359 

360 columns = self._tableColumns(table_enum) 

361 constraints = self._tableIndices(table_enum) 

362 table = Table( 

363 table_enum.table_name(self._prefix), 

364 self._metadata, 

365 *columns, 

366 *constraints, 

367 mysql_engine=mysql_engine, 

368 ) 

369 tables[table_enum] = table 

370 

371 return tables 

372 

373 def _make_extra_tables( 

374 self, apdb_tables: Mapping[ApdbTables, Table], mysql_engine: str = "InnoDB" 

375 ) -> Mapping[ExtraTables, Table]: 

376 """Generate schema for insert ID tables.""" 

377 tables: dict[ExtraTables, Table] = {} 

378 if not self._use_insert_id: 

379 return tables 

380 

381 # Parent table needs to be defined first 

382 column_defs = [ 

383 Column("insert_id", GUID, primary_key=True), 

384 Column("insert_time", sqlalchemy.types.TIMESTAMP, nullable=False), 

385 ] 

386 parent_table = Table( 

387 ExtraTables.DiaInsertId.table_name(self._prefix), 

388 self._metadata, 

389 *column_defs, 

390 mysql_engine=mysql_engine, 

391 ) 

392 tables[ExtraTables.DiaInsertId] = parent_table 

393 

394 for table_enum, apdb_enum in ExtraTables.insert_id_tables().items(): 

395 apdb_table = apdb_tables[apdb_enum] 

396 columns = self._insertIdColumns(table_enum) 

397 constraints = self._insertIdIndices(table_enum, apdb_table, parent_table) 

398 table = Table( 

399 table_enum.table_name(self._prefix), 

400 self._metadata, 

401 *columns, 

402 *constraints, 

403 mysql_engine=mysql_engine, 

404 ) 

405 tables[table_enum] = table 

406 

407 return tables 

408 

409 def _tableColumns(self, table_name: ApdbTables) -> List[Column]: 

410 """Return set of columns in a table 

411 

412 Parameters 

413 ---------- 

414 table_name : `ApdbTables` 

415 Name of the table. 

416 

417 Returns 

418 ------- 

419 column_defs : `list` 

420 List of `Column` objects. 

421 """ 

422 # get the list of columns in primary key, they are treated somewhat 

423 # specially below 

424 table_schema = self.tableSchemas[table_name] 

425 

426 # convert all column dicts into alchemy Columns 

427 column_defs: list[Column] = [] 

428 for column in table_schema.columns: 

429 kwargs: Dict[str, Any] = dict(nullable=column.nullable) 

430 if column.value is not None: 

431 kwargs.update(server_default=str(column.value)) 

432 if column in table_schema.primary_key: 

433 kwargs.update(autoincrement=False) 

434 ctype = self._type_map[column.datatype] 

435 column_defs.append(Column(column.name, ctype, **kwargs)) 

436 

437 return column_defs 

438 

439 def _tableIndices(self, table_name: ApdbTables) -> List[sqlalchemy.schema.SchemaItem]: 

440 """Return set of constraints/indices in a table 

441 

442 Parameters 

443 ---------- 

444 table_name : `ApdbTables` 

445 Name of the table. 

446 info : `dict` 

447 Additional options passed to SQLAlchemy index constructor. 

448 

449 Returns 

450 ------- 

451 index_defs : `list` 

452 List of SQLAlchemy index/constraint objects. 

453 """ 

454 

455 table_schema = self.tableSchemas[table_name] 

456 

457 # convert all index dicts into alchemy Columns 

458 index_defs: List[sqlalchemy.schema.SchemaItem] = [] 

459 if table_schema.primary_key: 

460 index_defs.append(PrimaryKeyConstraint(*[column.name for column in table_schema.primary_key])) 

461 for index in table_schema.indexes: 

462 name = self._prefix + index.name if index.name else "" 

463 index_defs.append(Index(name, *[column.name for column in index.columns])) 

464 for constraint in table_schema.constraints: 

465 constr_name: str | None = None 

466 if constraint.name: 

467 constr_name = self._prefix + constraint.name 

468 if isinstance(constraint, simple.UniqueConstraint): 

469 index_defs.append( 

470 UniqueConstraint(*[column.name for column in constraint.columns], name=constr_name) 

471 ) 

472 

473 return index_defs 

474 

475 def _insertIdColumns(self, table_enum: ExtraTables) -> List[Column]: 

476 """Return list of columns for insert ID tables.""" 

477 column_defs: list[Column] = [Column("insert_id", GUID, nullable=False)] 

478 insert_id_tables = ExtraTables.insert_id_tables() 

479 if table_enum in insert_id_tables: 

480 column_defs += self._tablePkColumns(insert_id_tables[table_enum]) 

481 else: 

482 assert False, "Above branches have to cover all enum values" 

483 return column_defs 

484 

485 def _tablePkColumns(self, table_enum: ApdbTables) -> list[Column]: 

486 """Return a list of columns for table PK.""" 

487 table_schema = self.tableSchemas[table_enum] 

488 column_defs: list[Column] = [] 

489 for column in table_schema.primary_key: 

490 ctype = self._type_map[column.datatype] 

491 column_defs.append(Column(column.name, ctype, nullable=False, autoincrement=False)) 

492 return column_defs 

493 

494 def _insertIdIndices( 

495 self, 

496 table_enum: ExtraTables, 

497 apdb_table: sqlalchemy.schema.Table, 

498 parent_table: sqlalchemy.schema.Table, 

499 ) -> List[sqlalchemy.schema.SchemaItem]: 

500 """Return set of constraints/indices for insert ID tables.""" 

501 index_defs: List[sqlalchemy.schema.SchemaItem] = [] 

502 

503 # Special case for insert ID tables that are not in felis schema. 

504 insert_id_tables = ExtraTables.insert_id_tables() 

505 if table_enum in insert_id_tables: 

506 # PK is the same as for original table 

507 pk_names = [column.name for column in self._tablePkColumns(insert_id_tables[table_enum])] 

508 index_defs.append(PrimaryKeyConstraint(*pk_names)) 

509 # Non-unique index on insert_id column. 

510 name = self._prefix + table_enum.name + "_idx" 

511 index_defs.append(Index(name, "insert_id")) 

512 # Foreign key to original table 

513 pk_columns = [apdb_table.columns[column] for column in pk_names] 

514 index_defs.append( 

515 ForeignKeyConstraint(pk_names, pk_columns, onupdate="CASCADE", ondelete="CASCADE") 

516 ) 

517 # Foreign key to parent table 

518 index_defs.append( 

519 ForeignKeyConstraint( 

520 ["insert_id"], [parent_table.columns["insert_id"]], onupdate="CASCADE", ondelete="CASCADE" 

521 ) 

522 ) 

523 else: 

524 assert False, "Above branches have to cover all enum values" 

525 return index_defs 

526 

527 @classmethod 

528 def _getDoubleType(cls, engine: sqlalchemy.engine.Engine) -> Type | sqlalchemy.types.TypeEngine: 

529 """DOUBLE type is database-specific, select one based on dialect. 

530 

531 Parameters 

532 ---------- 

533 engine : `sqlalchemy.engine.Engine` 

534 Database engine. 

535 

536 Returns 

537 ------- 

538 type_object : `object` 

539 Database-specific type definition. 

540 """ 

541 if engine.name == "mysql": 

542 from sqlalchemy.dialects.mysql import DOUBLE 

543 

544 return DOUBLE(asdecimal=False) 

545 elif engine.name == "postgresql": 

546 from sqlalchemy.dialects.postgresql import DOUBLE_PRECISION 

547 

548 return DOUBLE_PRECISION 

549 elif engine.name == "oracle": 

550 from sqlalchemy.dialects.oracle import DOUBLE_PRECISION 

551 

552 return DOUBLE_PRECISION 

553 elif engine.name == "sqlite": 

554 # all floats in sqlite are 8-byte 

555 from sqlalchemy.dialects.sqlite import REAL 

556 

557 return REAL 

558 else: 

559 raise TypeError("cannot determine DOUBLE type, unexpected dialect: " + engine.name)