Coverage for python/felis/datamodel.py: 53%
301 statements
« prev ^ index » next coverage.py v7.4.4, created at 2024-04-18 09:50 +0000
« prev ^ index » next coverage.py v7.4.4, created at 2024-04-18 09:50 +0000
1# This file is part of felis.
2#
3# Developed for the LSST Data Management System.
4# This product includes software developed by the LSST Project
5# (https://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 <https://www.gnu.org/licenses/>.
22from __future__ import annotations
24import logging
25import re
26from collections.abc import Mapping, Sequence
27from enum import StrEnum, auto
28from typing import Annotated, Any, Literal, TypeAlias
30from astropy import units as units # type: ignore
31from astropy.io.votable import ucd # type: ignore
32from pydantic import BaseModel, ConfigDict, Field, ValidationInfo, field_validator, model_validator
33from sqlalchemy import dialects
34from sqlalchemy import types as sqa_types
35from sqlalchemy.engine import create_mock_engine
36from sqlalchemy.engine.interfaces import Dialect
37from sqlalchemy.types import TypeEngine
39from .db.sqltypes import get_type_func
40from .types import FelisType
42logger = logging.getLogger(__name__)
44__all__ = (
45 "BaseObject",
46 "Column",
47 "CheckConstraint",
48 "Constraint",
49 "DescriptionStr",
50 "ForeignKeyConstraint",
51 "Index",
52 "Schema",
53 "SchemaVersion",
54 "Table",
55 "UniqueConstraint",
56)
58CONFIG = ConfigDict(
59 populate_by_name=True, # Populate attributes by name.
60 extra="forbid", # Do not allow extra fields.
61 str_strip_whitespace=True, # Strip whitespace from string fields.
62)
63"""Pydantic model configuration as described in:
64https://docs.pydantic.dev/2.0/api/config/#pydantic.config.ConfigDict
65"""
67DESCR_MIN_LENGTH = 3
68"""Minimum length for a description field."""
70DescriptionStr: TypeAlias = Annotated[str, Field(min_length=DESCR_MIN_LENGTH)]
71"""Define a type for a description string, which must be three or more
72characters long. Stripping of whitespace is done globally on all str fields."""
75class BaseObject(BaseModel):
76 """Base class for all Felis objects."""
78 model_config = CONFIG
79 """Pydantic model configuration."""
81 name: str
82 """The name of the database object.
84 All Felis database objects must have a name.
85 """
87 id: str = Field(alias="@id")
88 """The unique identifier of the database object.
90 All Felis database objects must have a unique identifier.
91 """
93 description: DescriptionStr | None = None
94 """A description of the database object."""
96 @model_validator(mode="after") # type: ignore[arg-type]
97 @classmethod
98 def check_description(cls, object: BaseObject, info: ValidationInfo) -> BaseObject:
99 """Check that the description is present if required."""
100 context = info.context
101 if not context or not context.get("require_description", False):
102 return object
103 if object.description is None or object.description == "":
104 raise ValueError("Description is required and must be non-empty")
105 if len(object.description) < DESCR_MIN_LENGTH:
106 raise ValueError(f"Description must be at least {DESCR_MIN_LENGTH} characters long")
107 return object
110class DataType(StrEnum):
111 """`Enum` representing the data types supported by Felis."""
113 boolean = auto()
114 byte = auto()
115 short = auto()
116 int = auto()
117 long = auto()
118 float = auto()
119 double = auto()
120 char = auto()
121 string = auto()
122 unicode = auto()
123 text = auto()
124 binary = auto()
125 timestamp = auto()
128_DIALECTS = {
129 "mysql": create_mock_engine("mysql://", executor=None).dialect,
130 "postgresql": create_mock_engine("postgresql://", executor=None).dialect,
131}
132"""Dictionary of dialect names to SQLAlchemy dialects."""
134_DIALECT_MODULES = {"mysql": getattr(dialects, "mysql"), "postgresql": getattr(dialects, "postgresql")}
135"""Dictionary of dialect names to SQLAlchemy dialect modules."""
137_DATATYPE_REGEXP = re.compile(r"(\w+)(\((.*)\))?")
138"""Regular expression to match data types in the form "type(length)"""
141def string_to_typeengine(
142 type_string: str, dialect: Dialect | None = None, length: int | None = None
143) -> TypeEngine:
144 match = _DATATYPE_REGEXP.search(type_string)
145 if not match:
146 raise ValueError(f"Invalid type string: {type_string}")
148 type_name, _, params = match.groups()
149 if dialect is None:
150 type_class = getattr(sqa_types, type_name.upper(), None)
151 else:
152 try:
153 dialect_module = _DIALECT_MODULES[dialect.name]
154 except KeyError:
155 raise ValueError(f"Unsupported dialect: {dialect}")
156 type_class = getattr(dialect_module, type_name.upper(), None)
158 if not type_class:
159 raise ValueError(f"Unsupported type: {type_class}")
161 if params:
162 params = [int(param) if param.isdigit() else param for param in params.split(",")]
163 type_obj = type_class(*params)
164 else:
165 type_obj = type_class()
167 if hasattr(type_obj, "length") and getattr(type_obj, "length") is None and length is not None:
168 type_obj.length = length
170 return type_obj
173class Column(BaseObject):
174 """A column in a table."""
176 datatype: DataType
177 """The datatype of the column."""
179 length: int | None = None
180 """The length of the column."""
182 nullable: bool | None = None
183 """Whether the column can be ``NULL``.
185 If `None`, this value was not set explicitly in the YAML data. In this
186 case, it will be set to `False` for columns with numeric types and `True`
187 otherwise.
188 """
190 value: Any = None
191 """The default value of the column."""
193 autoincrement: bool | None = None
194 """Whether the column is autoincremented."""
196 mysql_datatype: str | None = Field(None, alias="mysql:datatype")
197 """The MySQL datatype of the column."""
199 postgresql_datatype: str | None = Field(None, alias="postgresql:datatype")
200 """The PostgreSQL datatype of the column."""
202 ivoa_ucd: str | None = Field(None, alias="ivoa:ucd")
203 """The IVOA UCD of the column."""
205 fits_tunit: str | None = Field(None, alias="fits:tunit")
206 """The FITS TUNIT of the column."""
208 ivoa_unit: str | None = Field(None, alias="ivoa:unit")
209 """The IVOA unit of the column."""
211 tap_column_index: int | None = Field(None, alias="tap:column_index")
212 """The TAP_SCHEMA column index of the column."""
214 tap_principal: int | None = Field(0, alias="tap:principal", ge=0, le=1)
215 """Whether this is a TAP_SCHEMA principal column; can be either 0 or 1.
216 """
218 votable_arraysize: int | Literal["*"] | None = Field(None, alias="votable:arraysize")
219 """The VOTable arraysize of the column."""
221 tap_std: int | None = Field(0, alias="tap:std", ge=0, le=1)
222 """TAP_SCHEMA indication that this column is defined by an IVOA standard.
223 """
225 votable_utype: str | None = Field(None, alias="votable:utype")
226 """The VOTable utype (usage-specific or unique type) of the column."""
228 votable_xtype: str | None = Field(None, alias="votable:xtype")
229 """The VOTable xtype (extended type) of the column."""
231 @field_validator("ivoa_ucd")
232 @classmethod
233 def check_ivoa_ucd(cls, ivoa_ucd: str) -> str:
234 """Check that IVOA UCD values are valid."""
235 if ivoa_ucd is not None:
236 try:
237 ucd.parse_ucd(ivoa_ucd, check_controlled_vocabulary=True, has_colon=";" in ivoa_ucd)
238 except ValueError as e:
239 raise ValueError(f"Invalid IVOA UCD: {e}")
240 return ivoa_ucd
242 @model_validator(mode="before")
243 @classmethod
244 def check_units(cls, values: dict[str, Any]) -> dict[str, Any]:
245 """Check that units are valid."""
246 fits_unit = values.get("fits:tunit")
247 ivoa_unit = values.get("ivoa:unit")
249 if fits_unit and ivoa_unit:
250 raise ValueError("Column cannot have both FITS and IVOA units")
251 unit = fits_unit or ivoa_unit
253 if unit is not None:
254 try:
255 units.Unit(unit)
256 except ValueError as e:
257 raise ValueError(f"Invalid unit: {e}")
259 return values
261 @model_validator(mode="after") # type: ignore[arg-type]
262 @classmethod
263 def validate_datatypes(cls, col: Column, info: ValidationInfo) -> Column:
264 """Check for redundant datatypes on columns."""
265 context = info.context
266 if not context or not context.get("check_redundant_datatypes", False):
267 return col
268 if all(getattr(col, f"{dialect}:datatype", None) is not None for dialect in _DIALECTS.keys()):
269 return col
271 datatype = col.datatype
272 length: int | None = col.length or None
274 datatype_func = get_type_func(datatype)
275 felis_type = FelisType.felis_type(datatype)
276 if felis_type.is_sized:
277 if length is not None:
278 datatype_obj = datatype_func(length)
279 else:
280 raise ValueError(f"Length must be provided for sized type '{datatype}' in column '{col.id}'")
281 else:
282 datatype_obj = datatype_func()
284 for dialect_name, dialect in _DIALECTS.items():
285 db_annotation = f"{dialect_name}_datatype"
286 if datatype_string := col.model_dump().get(db_annotation):
287 db_datatype_obj = string_to_typeengine(datatype_string, dialect, length)
288 if datatype_obj.compile(dialect) == db_datatype_obj.compile(dialect):
289 raise ValueError(
290 "'{}: {}' is the same as 'datatype: {}' in column '{}'".format(
291 db_annotation, datatype_string, col.datatype, col.id
292 )
293 )
294 else:
295 logger.debug(
296 "Valid type override of 'datatype: {}' with '{}: {}' in column '{}'".format(
297 col.datatype, db_annotation, datatype_string, col.id
298 )
299 )
300 logger.debug(
301 "Compiled datatype '{}' with {} compiled override '{}'".format(
302 datatype_obj.compile(dialect), dialect_name, db_datatype_obj.compile(dialect)
303 )
304 )
306 return col
309class Constraint(BaseObject):
310 """A database table constraint."""
312 deferrable: bool = False
313 """If `True` then this constraint will be declared as deferrable."""
315 initially: str | None = None
316 """Value for ``INITIALLY`` clause, only used if ``deferrable`` is True."""
318 annotations: Mapping[str, Any] = Field(default_factory=dict)
319 """Additional annotations for this constraint."""
321 type: str | None = Field(None, alias="@type")
322 """The type of the constraint."""
325class CheckConstraint(Constraint):
326 """A check constraint on a table."""
328 expression: str
329 """The expression for the check constraint."""
332class UniqueConstraint(Constraint):
333 """A unique constraint on a table."""
335 columns: list[str]
336 """The columns in the unique constraint."""
339class Index(BaseObject):
340 """A database table index.
342 An index can be defined on either columns or expressions, but not both.
343 """
345 columns: list[str] | None = None
346 """The columns in the index."""
348 expressions: list[str] | None = None
349 """The expressions in the index."""
351 @model_validator(mode="before")
352 @classmethod
353 def check_columns_or_expressions(cls, values: dict[str, Any]) -> dict[str, Any]:
354 """Check that columns or expressions are specified, but not both."""
355 if "columns" in values and "expressions" in values:
356 raise ValueError("Defining columns and expressions is not valid")
357 elif "columns" not in values and "expressions" not in values:
358 raise ValueError("Must define columns or expressions")
359 return values
362class ForeignKeyConstraint(Constraint):
363 """A foreign key constraint on a table.
365 These will be reflected in the TAP_SCHEMA keys and key_columns data.
366 """
368 columns: list[str]
369 """The columns comprising the foreign key."""
371 referenced_columns: list[str] = Field(alias="referencedColumns")
372 """The columns referenced by the foreign key."""
375class Table(BaseObject):
376 """A database table."""
378 columns: Sequence[Column]
379 """The columns in the table."""
381 constraints: list[Constraint] = Field(default_factory=list)
382 """The constraints on the table."""
384 indexes: list[Index] = Field(default_factory=list)
385 """The indexes on the table."""
387 primary_key: str | list[str] | None = Field(None, alias="primaryKey")
388 """The primary key of the table."""
390 tap_table_index: int | None = Field(None, alias="tap:table_index")
391 """The IVOA TAP_SCHEMA table index of the table."""
393 mysql_engine: str | None = Field(None, alias="mysql:engine")
394 """The mysql engine to use for the table.
396 For now this is a freeform string but it could be constrained to a list of
397 known engines in the future.
398 """
400 mysql_charset: str | None = Field(None, alias="mysql:charset")
401 """The mysql charset to use for the table.
403 For now this is a freeform string but it could be constrained to a list of
404 known charsets in the future.
405 """
407 @model_validator(mode="before")
408 @classmethod
409 def create_constraints(cls, values: dict[str, Any]) -> dict[str, Any]:
410 """Create constraints from the ``constraints`` field."""
411 if "constraints" in values:
412 new_constraints: list[Constraint] = []
413 for item in values["constraints"]:
414 if item["@type"] == "ForeignKey":
415 new_constraints.append(ForeignKeyConstraint(**item))
416 elif item["@type"] == "Unique":
417 new_constraints.append(UniqueConstraint(**item))
418 elif item["@type"] == "Check":
419 new_constraints.append(CheckConstraint(**item))
420 else:
421 raise ValueError(f"Unknown constraint type: {item['@type']}")
422 values["constraints"] = new_constraints
423 return values
425 @field_validator("columns", mode="after")
426 @classmethod
427 def check_unique_column_names(cls, columns: list[Column]) -> list[Column]:
428 """Check that column names are unique."""
429 if len(columns) != len(set(column.name for column in columns)):
430 raise ValueError("Column names must be unique")
431 return columns
434class SchemaVersion(BaseModel):
435 """The version of the schema."""
437 current: str
438 """The current version of the schema."""
440 compatible: list[str] = Field(default_factory=list)
441 """The compatible versions of the schema."""
443 read_compatible: list[str] = Field(default_factory=list)
444 """The read compatible versions of the schema."""
447class SchemaIdVisitor:
448 """Visitor to build a Schema object's map of IDs to objects.
450 Duplicates are added to a set when they are encountered, which can be
451 accessed via the `duplicates` attribute. The presence of duplicates will
452 not throw an error. Only the first object with a given ID will be added to
453 the map, but this should not matter, since a ValidationError will be thrown
454 by the `model_validator` method if any duplicates are found in the schema.
456 This class is intended for internal use only.
457 """
459 def __init__(self) -> None:
460 """Create a new SchemaVisitor."""
461 self.schema: "Schema" | None = None
462 self.duplicates: set[str] = set()
464 def add(self, obj: BaseObject) -> None:
465 """Add an object to the ID map."""
466 if hasattr(obj, "id"):
467 obj_id = getattr(obj, "id")
468 if self.schema is not None:
469 if obj_id in self.schema.id_map:
470 self.duplicates.add(obj_id)
471 else:
472 self.schema.id_map[obj_id] = obj
474 def visit_schema(self, schema: "Schema") -> None:
475 """Visit the schema object that was added during initialization.
477 This will set an internal variable pointing to the schema object.
478 """
479 self.schema = schema
480 self.duplicates.clear()
481 self.add(self.schema)
482 for table in self.schema.tables:
483 self.visit_table(table)
485 def visit_table(self, table: Table) -> None:
486 """Visit a table object."""
487 self.add(table)
488 for column in table.columns:
489 self.visit_column(column)
490 for constraint in table.constraints:
491 self.visit_constraint(constraint)
493 def visit_column(self, column: Column) -> None:
494 """Visit a column object."""
495 self.add(column)
497 def visit_constraint(self, constraint: Constraint) -> None:
498 """Visit a constraint object."""
499 self.add(constraint)
502class Schema(BaseObject):
503 """The database schema containing the tables."""
505 version: SchemaVersion | str | None = None
506 """The version of the schema."""
508 tables: Sequence[Table]
509 """The tables in the schema."""
511 id_map: dict[str, Any] = Field(default_factory=dict, exclude=True)
512 """Map of IDs to objects."""
514 @field_validator("tables", mode="after")
515 @classmethod
516 def check_unique_table_names(cls, tables: list[Table]) -> list[Table]:
517 """Check that table names are unique."""
518 if len(tables) != len(set(table.name for table in tables)):
519 raise ValueError("Table names must be unique")
520 return tables
522 def _create_id_map(self: Schema) -> Schema:
523 """Create a map of IDs to objects.
525 This method should not be called by users. It is called automatically
526 by the ``model_post_init()`` method. If the ID map is already
527 populated, this method will return immediately.
528 """
529 if len(self.id_map):
530 logger.debug("Ignoring call to create_id_map() - ID map was already populated")
531 return self
532 visitor: SchemaIdVisitor = SchemaIdVisitor()
533 visitor.visit_schema(self)
534 logger.debug(f"Created schema ID map with {len(self.id_map.keys())} objects")
535 if len(visitor.duplicates):
536 raise ValueError(
537 "Duplicate IDs found in schema:\n " + "\n ".join(visitor.duplicates) + "\n"
538 )
539 return self
541 def model_post_init(self, ctx: Any) -> None:
542 """Post-initialization hook for the model."""
543 self._create_id_map()
545 def __getitem__(self, id: str) -> BaseObject:
546 """Get an object by its ID."""
547 if id not in self:
548 raise KeyError(f"Object with ID '{id}' not found in schema")
549 return self.id_map[id]
551 def __contains__(self, id: str) -> bool:
552 """Check if an object with the given ID is in the schema."""
553 return id in self.id_map