FieldSpec#

class lsst.daf.butler.ddl.FieldSpec(name: str, dtype: type, length: int | None = None, nbytes: int | None = None, primaryKey: bool = False, autoincrement: bool = False, nullable: bool = True, default: Any = None, doc: str | None = None)#

Bases: object

A data class for defining a column in a logical Registry table.

Attributes Summary

`autoincrement`	Whether the database should insert automatically incremented values when no value is provided in an INSERT.
`default`	A server-side default value for this field.
`doc`	Documentation for this field.
`dtype`	Type of the column; usually a `type` subclass provided by SQLAlchemy that defines both a Python type and a corresponding precise SQL type.
`length`	Length of the type in the database, for variable-length types.
`name`	Name of the column.
`nbytes`	Natural length used for hash and encoded-region columns, to be converted into the post-encoding length.
`nullable`	Whether this field is allowed to be NULL.
`primaryKey`	Whether this field is (part of) its table's primary key.

Methods Summary

`for_region`([name, nullable, nbytes])	Create a `FieldSpec` for a spatial region column.
`fromConfig`(config, args, *kwargs)
`getPythonType`()	Return the Python type associated with this field's (SQL) dtype.
`getSizedColumnType`()	Return a sized version of the column type.
`isStringType`()	Indicate that this is a sqlalchemy.String field spec.

Attributes Documentation

autoincrement: bool = False#: Whether the database should insert automatically incremented values when no value is provided in an INSERT.

default: Any = None#

A server-side default value for this field.

This is passed directly as the server_default argument to sqlalchemy.schema.Column. It does _not_ go through SQLAlchemy’s usual type conversion or quoting for Python literals, and should hence be used with care. See the SQLAlchemy documentation for more information.

doc: str | None = None#: Documentation for this field.

dtype: type = <dataclasses._MISSING_TYPE object>#: Type of the column; usually a type subclass provided by SQLAlchemy that defines both a Python type and a corresponding precise SQL type.

length: int | None = None#: Length of the type in the database, for variable-length types.

name: str = <dataclasses._MISSING_TYPE object>#: Name of the column.

nbytes: int | None = None#: Natural length used for hash and encoded-region columns, to be converted into the post-encoding length.

nullable: bool = True#: Whether this field is allowed to be NULL. If primaryKey is True, during construction this value will be forced to False.

primaryKey: bool = False#: Whether this field is (part of) its table’s primary key.

Methods Documentation

classmethod for_region(name: str = 'region', nullable: bool = True, nbytes: int = 2048) → FieldSpec#

Create a FieldSpec for a spatial region column.

Parameters#

namestr, optional: Name for the field.
nullablebool, optional: Whether NULL values are permitted.
nbytesint, optional: Maximum number of bytes for serialized regions. The actual column size will be larger to allow for base-64 encoding.

Returns#

specFieldSpec: Specification structure for a region column.

classmethod fromConfig(config: Config, *args: Any, **kwargs: Any) → Any#

getPythonType() → type#

Return the Python type associated with this field’s (SQL) dtype.

Returns#

typetype: Python type associated with this field’s (SQL) dtype.

getSizedColumnType() → TypeEngine | type#

Return a sized version of the column type.

Utilizes either (or neither) of self.length and self.nbytes.

Returns#

dtypesqlalchemy.types.TypeEngine: A SQLAlchemy column type object.

isStringType() → bool#

Indicate that this is a sqlalchemy.String field spec.

Returns#

isStringbool: The field refers to a sqlalchemy.String and not any other type. This can return False even if the object was created with a string type if it has been decided that it should be implemented as a sqlalchemy.Text type.