htsworkflow/submission/daf.py

   1 """Parse UCSC DAF File
   2 """
   3 import logging
   4 import re
   5 import string
   6 from StringIO import StringIO
   7 import types
   8
   9 from htsworkflow.util.rdfhelp import blankOrUri, toTypedNode
  10
  11 logger = logging.getLogger(__name__)
  12
  13 # STATES
  14 DAF_HEADER = 1
  15 DAF_VIEW = 2
  16
  17
  18 def parse(filename):
  19     stream = open(filename,'r')
  20     attributes =  parse_stream(stream)
  21     stream.close()
  22     return stream
  23
  24 def fromstring(daf_string):
  25     stream = StringIO(daf_string)
  26     return parse_stream(stream)
  27
  28 def parse_stream(stream):
  29     comment_re = re.compile("#.*$")
  30
  31     state = DAF_HEADER
  32     attributes = {'views': {}}
  33     view_name = None
  34     view_attributes = {}
  35     for line in stream:
  36         #remove comments
  37         line = comment_re.sub("", line)
  38         nstop = _extract_name_index(line)
  39         name = line[0:nstop]
  40         sstop = _consume_whitespace(line, start=nstop)
  41         vstop = _extract_value_index(line, start=sstop)
  42         value = line[sstop:vstop]
  43
  44         if value.lower() in ('yes',):
  45             value = True
  46         elif value.lower() in ('no',):
  47             value = False
  48
  49         if len(name) == 0:
  50             if view_name is not None:
  51                 attributes['views'][view_name] = view_attributes
  52                 view_name = None
  53                 view_attributes = {}
  54             state = DAF_HEADER
  55         elif state == DAF_HEADER and name == 'variables':
  56             attributes[name] = [ x.strip() for x in value.split(',')]
  57         elif state == DAF_HEADER and name == 'view':
  58             view_name = value
  59             view_attributes['view'] = value
  60             state = DAF_VIEW
  61         elif state == DAF_HEADER:
  62             attributes[name] = value
  63         elif state == DAF_VIEW:
  64             view_attributes[name] = value
  65
  66     # save last block
  67     if view_name is not None:
  68         attributes['views'][view_name] = view_attributes
  69
  70     return attributes
  71
  72 def _consume_whitespace(line, start=0):
  73     for i in xrange(start, len(line)):
  74         if line[i] not in string.whitespace:
  75             return i
  76
  77     return len(line)
  78
  79 def _extract_name_index(line, start=0):
  80     for i in xrange(start, len(line)):
  81         if line[i] in string.whitespace:
  82             return i
  83
  84     return len(line)
  85
  86 def _extract_value_index(line, start=0):
  87     shortline = line.rstrip()
  88     return len(shortline)
  89
  90 try:
  91     import RDF
  92     def convert_to_rdf_statements(attributes, source=None):
  93         ddfNS = RDF.NS("http://encodesubmit.ucsc.edu/pipeline/download_ddf#")
  94
  95         subject = blankOrUri(source)
  96
  97         statements = []
  98         for name in attributes:
  99             predicate = ddfNS[name]
 100             if name == 'views':
 101                 predicate = ddfNS['views']
 102                 for view_name in attributes.get('views', []):
 103                     view = attributes['views'][view_name]
 104                     viewNode = RDF.Node()
 105                     statements.append(RDF.Statement(subject, predicate, viewNode))
 106                     statements.extend(convert_to_rdf_statements(view, viewNode))
 107             elif name == 'variables':
 108                 predicate = ddfNS['variables']
 109                 for var in attributes.get('variables', []):
 110                     obj = toTypedNode(var)
 111                     statements.append(RDF.Statement(subject, predicate, obj))
 112             else:
 113                 value = attributes[name]
 114                 obj = toTypedNode(value)
 115                 statements.append(RDF.Statement(subject,predicate,obj))
 116
 117         return statements
 118
 119
 120     def add_to_model(model, attributes, source=None):
 121         for statement in convert_to_rdf_statements(attributes, source):
 122             model.add_statement(statement)
 123
 124 except ImportError, e:
 125     def convert_to_rdf_statements(attributes, source=None):
 126         raise NotImplementedError("librdf not installed")
 127     def add_to_model(model, attributes, source=None):
 128         raise NotImplementedError("librdf not installed")
 129