View Source RDF.XSD.Datatype behaviour (RDF.ex v2.0.0)

A behaviour for XSD datatypes.

A XSD datatype has three properties:

  • A value space, which is a set of values.
  • A lexical space, which is a set of literals used to denote the values.
  • A collection of functions associated with the datatype.

Builtin XSD datatypes

RDF.ex comes with the following builtin implementations of XSD datatypes:

xsd:booleanRDF.XSD.Boolean
xsd:floatRDF.XSD.Float
xsd:doubleRDF.XSD.Double
xsd:decimalRDF.XSD.Decimal
xsd:integerRDF.XSD.Integer
xsd:longRDF.XSD.Long
xsd:intRDF.XSD.Int
xsd:shortRDF.XSD.Short
xsd:byteRDF.XSD.Byte
xsd:nonPositiveIntegerRDF.XSD.NonPositiveInteger
xsd:negativeIntegerRDF.XSD.NegativeInteger
xsd:nonNegativeIntegerRDF.XSD.NonNegativeInteger
xsd:positiveIntegerRDF.XSD.PositiveInteger
xsd:unsignedLongRDF.XSD.UnsignedLong
xsd:unsignedIntRDF.XSD.UnsignedInt
xsd:unsignedShortRDF.XSD.UnsignedShort
xsd:unsignedByteRDF.XSD.UnsignedByte
xsd:stringRDF.XSD.String
xsd:normalizedString
xsd:token
xsd:language
xsd:Name
xsd:NCName
xsd:ID
xsd:IDREF
xsd:ENTITY
xsd:NMTOKEN
xsd:dateTimeRDF.XSD.DateTime
xsd:dateTimeStamp
xsd:dateRDF.XSD.Date
xsd:timeRDF.XSD.Time
xsd:duration
xsd:dayTimeDuration
xsd:yearMonthDuration
xsd:gYearMonth
xsd:gYear
xsd:gMonthDay
xsd:gDay
xsd:gMonth
xsd:base64BinaryRDF.XSD.Base64Binary
xsd:hexBinary
xsd:anyURIRDF.XSD.AnyURI
xsd:QName
xsd:NOTATION

There are some notable difference in the implementations of some datatypes compared to the original spec:

see https://www.w3.org/TR/xmlschema11-2/#built-in-datatypes

Summary

Callbacks

The set of applicable facets of a RDF.XSD.Datatype.

The base datatype from which a RDF.XSD.Datatype is derived.

The primitive RDF.XSD.Datatype from which a RDF.XSD.Datatype is derived.

Returns the standard lexical representation for a value of the value space of a RDF.XSD.Datatype.

Checks if the RDF.XSD.Datatype is directly or indirectly derived from the given RDF.XSD.Datatype.

A mapping from Elixir values into the value space of a RDF.XSD.Datatype.

Produces the lexical representation of an invalid value.

Produces the lexical representation to be used for a RDF.XSD.Datatype literal.

A mapping from the lexical space of a RDF.XSD.Datatype into its value space.

Returns if the RDF.XSD.Datatype is a primitive datatype.

Functions

Returns the RDF.XSD.Datatype for a datatype IRI.

Types

@type literal() :: %{
  __struct__: t(),
  value: any(),
  uncanonical_lexical: uncanonical_lexical()
}
@type t() :: module()
@type uncanonical_lexical() :: String.t() | nil

Callbacks

@callback applicable_facets() :: [RDF.XSD.Facet.t()]

The set of applicable facets of a RDF.XSD.Datatype.

@callback base() :: t() | nil

The base datatype from which a RDF.XSD.Datatype is derived.

Note: Since this library focuses on atomic types and the special xsd:anyAtomicType specified as the base type of all primitive types in the W3C spec wouldn't serve any purpose here, all primitive datatypes just return nil instead.

@callback base_primitive() :: t()

The primitive RDF.XSD.Datatype from which a RDF.XSD.Datatype is derived.

In case of a primitive RDF.XSD.Datatype this function returns this RDF.XSD.Datatype itself.

@callback canonical_mapping(any()) :: String.t()

Returns the standard lexical representation for a value of the value space of a RDF.XSD.Datatype.

@callback derived_from?(t()) :: boolean()

Checks if the RDF.XSD.Datatype is directly or indirectly derived from the given RDF.XSD.Datatype.

Note that this is just a basic datatype reflection function on the module level and does not work with RDF.Literals. See RDF.Literal.Datatype.datatype?/1 instead.

@callback elixir_mapping(any(), Keyword.t()) :: any() | {any(), uncanonical_lexical()}

A mapping from Elixir values into the value space of a RDF.XSD.Datatype.

If the Elixir mapping for the given value can not be mapped into value space of the XSD datatype an implementation should return @invalid_value (which is just nil at the moment, so nil is never a valid value of a value space).

Otherwise, a tuple {value, lexical} with value being the internal representation of the mapped value from the value space and lexical being the lexical representation to be used for the Elixir value or nil if init_valid_lexical/3 should be used to determine the lexical form in general (i.e. also when initialized with a string via the lexical_mapping/2). Since the later case is most often what you want, you can also return value directly, as long as it is not a two element tuple.

Link to this callback

init_invalid_lexical(any, t)

View Source
@callback init_invalid_lexical(any(), Keyword.t()) :: String.t()

Produces the lexical representation of an invalid value.

The default implementation of the _using__ macro just returns the to_string/1 representation of the value.

Link to this callback

init_valid_lexical(any, uncanonical_lexical, t)

View Source
@callback init_valid_lexical(any(), uncanonical_lexical(), Keyword.t()) ::
  uncanonical_lexical()

Produces the lexical representation to be used for a RDF.XSD.Datatype literal.

By default, the lexical representation of a RDF.XSD.Datatype is either the canonical form in case it is created from a non-string Elixir value or, if it is created from a string, just with that string as the lexical form.

But there can be various reasons for why this should be different for certain datatypes. For example, for RDF.XSD.Doubles given as Elixir floats, we want the default lexical representation to be the decimal and not the canonical exponential form. Another reason might be that additional options are given which should be taken into account in the lexical form.

If the lexical representation for a given value and lexical should be the canonical one, an implementation should return nil.

@callback lexical_mapping(String.t(), Keyword.t()) :: any()

A mapping from the lexical space of a RDF.XSD.Datatype into its value space.

@callback primitive?() :: boolean()

Returns if the RDF.XSD.Datatype is a primitive datatype.

Functions

Returns the RDF.XSD.Datatype for a datatype IRI.