SPARQL Extensions

Owl Datatypes

Native support for the owl:real and owl:rational datatypes. This includes the support for comparison and arithmetic operations that involve these datatypes.

Trigonometric and exponential functions

We implement all Trigonometric and exponential functions from XPath and XQuery Functions and Operators 3.1. These functions accept input that are convertible to xsd:double. Following the XPath 3.1 specs, the output of these functions is always xsd:double.

The table below provides an overview of the trigonometric and exponential functions.

Function	Meaning
`PI()`	Returns an approximation to the mathematical constant π.
`EXP(?x)`	Returns the value of e^x.
`EXP10(?x)`	Returns the value of 10^x.
`LOG(?arg)`	Returns the natural logarithm of the argument.
`LOG10(?arg)`	Returns the base-ten logarithm of the argument.
`POW(?x, ?y)`	Returns the value of x^y.
`SQRT(?arg)`	Returns the non-negative square root of the argument.
`SIN(?θ)`	Returns the sine of the argument. The argument is an angle in radians.
`COS(?θ)`	Returns the cosine of the argument. The argument is an angle in radians.
`TAN(?θ)`	Returns the tangent of the argument. The argument is an angle in radians.
`ASIN(?arg)`	Returns the arc sine of the argument.
`ACOS(?arg)`	Returns the arc cosine of the argument.
`ATAN(?arg)`	Returns the arc tangent of the argument.
`ATAN2(?y, ?x)`	Returns the angle in radians subtended at the origin by the point on a plane with coordinates (x, y) and the positive x-axis.
`TO_UNIX_TIME(?dt)`	Returns the unix time representation of the given `xsd:dateTimeStamp` (in whole seconds since `1970-01-01T00:00:00Z` as `xsd:integer`).
`FROM_UNIX_TIME(?unix)`	Returns the `xsd:dateTimeStamp` corresponding to the unix time representation (in whole seconds since `1970-01-01T00:00:00Z`).

The SPARQL standard allows implementations to extend the capabilities of operations (e.g., <, + and =). As described here, these extensions lead to FILTER expressions potentially producing additional results.

Examples of operations that involve extended operations are provided below

Extended comparison operations

xsd:date > xsd:date => xsd:boolean
xsd:duration != xsd:duration => xsd:boolean
owl:real < owl:real => xsd:boolean

Extended arithmetic operations on date and time types

xsd:date - xsd:date => xsd:dayTimeDuration
xsd:dayTimeDuration / xsd:int => xsd:dayTimeDuration

The extended capabilities of the operators are detailed in the chapter Extensions of Operators.

Note: These extensions do not discard any of the expected results.

`DESCRIBE` Queries

Tentris implements two algorithms for the evaluation DESCRIBE queries.

DESCRIBE / DESCRIBE DIRECT

DESCRIBE <http://www.example.org/entity>

DESCRIBE DIRECT <http://www.example.org/entity>

Both queries provided above return the triples of the underlying knowledge graph that have <http://www.example.org/entity> as their subject.

DESCRIBE CBD

DESCRIBE CBD <http://www.example.org/entity>

The query provided above returns the Consise Bounded Description of <http://www.example.org/entity>.

Online Loading of Datasets

Tentris enables the online loading of datasets in two ways. First, Tentris extends the LOAD operation of the SPARQL Update language as shown below.

LOAD <http://www.example.org/datasen.nq> INTO DATASET

Second, Tentris allows the loading of datasets via the Graph Store Protocol via /graph-store?dataset. It supports all operations that are also available for graphs, usage is analogous to what is specified in the standard.

For example, to download the whole dataset, use the following:

curl -H "Accept: application/n-quads" "http://localhost:9080/graph-store?dataset"

For more details on usage see Chapter 6 (/graph-store).

Explicit Content-Type specification in LOAD

When a LOAD query is issued, Tentris uses the Content-Type header from the server's response to determine the file format of the given RDF file.

If the server does not answer with a Content-Type at all or with text/plain, Tentris looks at the URL and attempts to detect the Content-Type based on the file extension (if any exists). For example in LOAD <https://example.com/data.nt> the format is going to be detected as N-Triples. This, however, only works when the URL contains such information.

As a last resort, Tentris allows the user to explicitly specify the format.

Examples

LOAD <https://example.com/graph> AS NTRIPLES

LOAD <https://example.com/dataset> INTO DATASET AS NQUADS

The allowed formats after AS are NTRIPLES, NQUADS, TURTLE, TRIG and RDFXML.

`DROP DATASET` and `CLEAR DATASET`

For consistency reasons, Tentris extends the Update language with the operations DROP DATASET and CLEAR DATASET, which are aliases for DROP ALL and CLEAR ALL, respectively.

Tentris Beta Documentation