htsworkflow/util/rdfhelp.py

   1 """Helper features for working with librdf
   2 """
   3 import logging
   4 import os
   5 import types
   6
   7 import RDF
   8
   9 logger = logging.getLogger(__name__)
  10
  11 # standard ontology namespaces
  12 owlNS = RDF.NS('http://www.w3.org/2002/07/owl#')
  13 dublinCoreNS = RDF.NS("http://purl.org/dc/elements/1.1/")
  14 rdfNS = RDF.NS("http://www.w3.org/1999/02/22-rdf-syntax-ns#")
  15 rdfsNS= RDF.NS("http://www.w3.org/2000/01/rdf-schema#")
  16 xsdNS = RDF.NS("http://www.w3.org/2001/XMLSchema#")
  17
  18 # internal ontologies
  19 submissionOntology = RDF.NS("http://jumpgate.caltech.edu/wiki/UcscSubmissionOntology#")
  20 dafTermOntology = RDF.NS("http://jumpgate.caltech.edu/wiki/UcscDaf#")
  21 libraryOntology = RDF.NS("http://jumpgate.caltech.edu/wiki/LibraryOntology#")
  22 submissionLog = RDF.NS("http://jumpgate.caltech.edu/wiki/SubmissionsLog/")
  23
  24 def sparql_query(model, query_filename):
  25     """Execute sparql query from file
  26     """
  27     logger.info("Opening: %s" % (query_filename,))
  28     query_body = open(query_filename,'r').read()
  29     query = RDF.SPARQLQuery(query_body)
  30     results = query.execute(model)
  31     for row in results:
  32         output = []
  33         for k,v in row.items()[::-1]:
  34             print "{0}: {1}".format(k,v)
  35         print
  36
  37
  38 def blankOrUri(value=None):
  39     node = None
  40     if value is None:
  41         node = RDF.Node()
  42     elif type(value) in types.StringTypes:
  43         node = RDF.Node(uri_string=value)
  44     elif isinstance(value, RDF.Node):
  45         node = value
  46
  47     return node
  48
  49
  50 def toTypedNode(value):
  51     if type(value) == types.BooleanType:
  52         value_type = xsdNS['boolean'].uri
  53         if value:
  54             value = u'1'
  55         else:
  56             value = u'0'
  57     elif type(value) in (types.IntType, types.LongType):
  58         value_type = xsdNS['decimal'].uri
  59         value = unicode(value)
  60     elif type(value) == types.FloatType:
  61         value_type = xsdNS['float'].uri
  62         value = unicode(value)
  63     elif type(value) in types.StringTypes:
  64         value_type = xsdNS['string'].uri
  65     else:
  66         value_type = None
  67         value = unicode(value)
  68
  69     return RDF.Node(literal=value, datatype=value_type)
  70
  71 def fromTypedNode(node):
  72     if node is None:
  73         return None
  74
  75     value_type = str(node.literal_value['datatype'])
  76     # chop off xml schema declaration
  77     value_type = value_type.replace(str(xsdNS[''].uri),'')
  78     literal = node.literal_value['string']
  79     literal_lower = literal.lower()
  80
  81     if value_type == 'boolean':
  82         if literal_lower in ('1', 'yes', 'true'):
  83             return True
  84         elif literal_lower in ('0', 'no', 'false'):
  85             return False
  86         else:
  87             raise ValueError("Unrecognized boolean %s" % (literal,))
  88     elif value_type == 'decimal' and literal.find('.') == -1:
  89         return int(literal)
  90     elif value_type in ('decimal', 'float', 'double'):
  91         return float(literal)
  92     elif value_type in ('string'):
  93         return literal
  94     elif value_type in ('dateTime'):
  95         raise NotImplemented('need to parse isoformat date-time')
  96
  97     return literal
  98
  99
 100 def get_model(model_name=None, directory=None):
 101     if directory is None:
 102         directory = os.getcwd()
 103
 104     if model_name is None:
 105         storage = RDF.MemoryStorage()
 106         logger.info("Using RDF Memory model")
 107     else:
 108         options = "hash-type='bdb',dir='{0}'".format(directory)
 109         storage = RDF.HashStorage(model_name,
 110                       options=options)
 111         logger.info("Using {0} with options {1}".format(model_name, options))
 112     model = RDF.Model(storage)
 113     return model
 114
 115
 116 def load_into_model(model, parser_name, filename, ns=None):
 117     if not os.path.exists(filename):
 118         raise IOError("Can't find {0}".format(filename))
 119
 120     data = open(filename, 'r').read()
 121     rdf_parser = RDF.Parser(name=parser_name)
 122     rdf_parser.parse_string_into_model(model, data, ns)
 123
 124
 125 def get_serializer(name='turtle'):
 126     """Return a serializer with our standard prefixes loaded
 127     """
 128     writer = RDF.Serializer(name=name)
 129     # really standard stuff
 130     writer.set_namespace('owl', owlNS._prefix)
 131     writer.set_namespace('rdf', rdfNS._prefix)
 132     writer.set_namespace('rdfs', rdfsNS._prefix)
 133     writer.set_namespace('xsd', xsdNS._prefix)
 134
 135     # should these be here, kind of specific to an application
 136     writer.set_namespace('libraryOntology', libraryOntology._prefix)
 137     writer.set_namespace('ucscSubmission', submissionOntology._prefix)
 138     writer.set_namespace('ucscDaf', dafTermOntology._prefix)
 139     return writer
 140