SPARQL-PLY

A SPARQL 1.1 query parser implemented with the PLY library.

Data Structure

This parser defines several data structures, referred to as query components, to represent the structure of a SPARQL query. Here are some figures to illustrate them. See the source code in sparql_ply/components.py for more details.

Query

GraphPattern

NodeTerm

PropertyPath

CollectionPath & BlankNodePath & TriplesPath

Expression

Usage

Parse SPARQL Query

Parse a SPARQL query string into a query component.

>>> from sparql_ply import parse_sparql
>>> query = parse_sparql('SELECT * WHERE { ?s ?p ?o }')

Collect Query Component

Collect components of specific types in the query.

For example, the following code collects all the variables in the query.

>>> from sparql_ply import parse_sparql, NodeTerm
>>> from sparql_ply.util import collect_component
>>>
>>> query = parse_sparql('SELECT * WHERE { ?s ?p ?o }')
>>> collect_component(query, NodeTerm.VAR)
[?s, ?p, ?o]

Expand Syntax Form

Expand syntax forms in a SPARQL query according to SPARQL 1.1 document.

The following syntax forms can be expanded:

1. Keyword a
2. Prefixed Names and Relative IRIs
3. RDF Collections
4. Blank Nodes
5. Predicate-Object Lists and Object Lists

>>> from sparql_ply.util import expand_syntax_form
>>> prefix_part = '\nPREFIX : <http://example.org/>\n'
>>> select_part = 'SELECT * WHERE {\n:Person ?p [a ?q]\n}'
>>> sparql = prefix_part + select_part
>>> print(expand_syntax_form(sparql))

PREFIX : <http://example.org/>
SELECT * WHERE {
<http://example.org/Person> ?p _:b0.
_:b0 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> ?q
}

Query Component Serialization

Convert a query component to a dictionary for JSON serialization, and vice versa for deserialization.

>>> from sparql_ply import parse_sparql
>>> from sparql_ply.util import serialize, deserialize
>>>
>>> q1 = parse_sparql('SELECT * WHERE { BIND (42 AS ?s) }')
>>> data = serialize(q1)
>>> q2 = deserialize(data)
>>> q1.to_str() == q2.to_str()
True
>>> import json
>>> print(json.dumps(data, indent=2))
{
  "type": "Query.SELECT",        
  "target": {
    "type": "NodeTerm.SPECIAL",  
    "value": "*"
  },
  "pattern": {
    "type": "GraphPattern.GROUP",
    "children": [
      {
        "type": "GraphPattern.BIND",
        "children": [
          {
            "type": "NodeTerm.INTEGER",
            "value": "42"
          },
          {
            "type": "NodeTerm.VAR",
            "value": "?s"
          }
        ]
      }
    ]
  }
}

Note that the following fields are omitted in the serialization if they are None.

• All fields in Query
• NodeTerm.language and NodeTerm.datatype

The following boolean fields are omitted in the serialization if they are False.

• GraphPattern.is_silent
• Expression.is_distinct

Get Variable Scopes

Get the scopes of variables in a SPARQL query.

>>> from sparql_ply import parse_sparql
>>> from sparql_ply.util import get_variable_scopes
>>> subselect = '{ SELECT ?b { ?b :p1 ?a. ?a :p2 :q. } }'
>>> sparql = 'SELECT ?a ?b { ?a :p :q. ' + subselect + ' }'
>>> for i in range(0, len(sparql), 25):
...     print(f'[{i:0{len(str(len(sparql)))}d}]', sparql[i:i + 25])
...
[00] SELECT ?a ?b { ?a :p :q. 
[25] { SELECT ?b { ?b :p1 ?a. 
[50] ?a :p2 :q. } } }
>>> for scope in get_variable_scopes(parse_sparql(sparql)):
...     print(sparql[scope[0][0]:scope[0][1]], scope)
...
?a [[46, 48], [50, 52]]
?a [[7, 9], [15, 17]]
?b [[10, 12], [34, 36], [39, 41]]