concepts.dsl.dsl_types#

Data structures for custom types in a DSL.

All types extend the base class TypeBase.

TypeBase has four derived types:

- :class:`ObjectType`: corresponds to an object in the physical world.
- :class:`ValueType`: corresponds to an encoding of a feature of the object.
- :class:`UnionType`: corresponds to a union of types.
- :class:`SequenceType`: corresponds to a tuple or list type.
- :class:`~concepts.dsl.dsl_functions.FunctionType`: corresponds to a function over the encodings.
- :class:`~concepts.dsl.dsl_functions.OverloadedFunctionType`: corresponds to a collection of functions that share the same name.

Furthermore, ValueType has four derived types:

- :class:`ConstantType`: corresponds to a constant.
- :class:`TensorValueTypeBase`: corresponds to a tensor-based encoding.
- :class:`PyObjValueType`: corresponds to an arbitrary Python object.

Module attributes

`AnyType`	AutoType means the type will be automatically inferred later.
`QINDEX`	The "quantified index" object, used to indicate a quantified dimension in a tensor value type.

Functions

`is_any_type`(t)	Check if the type is AnyType.
`is_auto_type`(t)	Check if the type is AutoType.

Classes

`ConstantType`	The ConstantType corresponds to a constant value.
`FormatContext`	FormatContext is a context manager that controls the format of types and objects.
`ListType`
`NamedTensorValueType`
`ObjectConstant`	The class representing a constant object in a DSL.
`ObjectType`	The ObjectType corresponds to the type of "real-world" objects.
`PyObjValueType`
`ScalarValueType`
`SequenceType`	The basic sequence type.
`TensorValueTypeBase`	A value type refers to a type of some features associated with physical objects.
`TupleType`
`TypeBase`	Base class for all types.
`UnionType`	The UnionType is a type that is the union of multiple types.
`UnnamedPlaceholder`
`ValueType`	The ValueType corresponds to a value of a certain type.
`Variable`	The class representing a variable in a function.
`VectorValueType`