'c(?P<cycle>[\d]+)_l(?P<lane>[\d]+)(_r(?P<read>[\d]))?\.fastq')
class FastqName(collections.Mapping):
+ """Utility class to convert to the standardized submission fastq name.
+ """
def __init__(self, is_paired=None, **kwargs):
+ """Create a fastq name handler.
+
+ Takes filename or common attributes like flowcell, lib_id, lane, read, cycle
+ """
self._attributes = ('flowcell', 'lib_id', 'lane', 'read', 'cycle')
self._is_paired = is_paired
value = kwargs[k]
self[k] = value
-
def _init_by_filename(self, filename):
match = FASTQ_RE.match(filename)
if match is None:
toTypedNode, \
fromTypedNode
from htsworkflow.util.hashfile import make_md5sum
-
+from htsworkflow.submission.fastqname import FastqName
from htsworkflow.submission.daf import \
MetadataLookupException, \
get_submission_uri
an_analysis))
# add file specific information
- fileNode = self.link_file_to_classes(pathname,
- an_analysis,
- an_analysis_uri,
- analysis_dir)
+ fileNode = self.make_file_node(pathname, an_analysis)
self.add_md5s(filename, fileNode, analysis_dir)
+ self.add_fastq_metadata(filename, fileNode)
self.model.add_statement(
RDF.Statement(fileNode,
rdfNS['type'],
file_type))
LOGGER.debug("Done.")
- def link_file_to_classes(self, pathname, submissionNode, submission_uri, analysis_dir):
+ def make_file_node(self, pathname, submissionNode):
+ """Create file node and attach it to its submission.
+ """
# add file specific information
path, filename = os.path.split(pathname)
fileNode = RDF.Node(RDF.Uri('file://'+ os.path.abspath(pathname)))
self.model.add_statement(
RDF.Statement(fileNode, dafTermOntology['md5sum'], md5))
+ def add_fastq_metadata(self, filename, fileNode):
+ # How should I detect if this is actually a fastq file?
+ try:
+ fqname = FastqName(filename=filename)
+ except ValueError:
+ # currently its just ignore it if the fastq name parser fails
+ return
+
+ terms = [('flowcell', libraryOntology['flowcell_id']),
+ ('lib_id', libraryOntology['library_id']),
+ ('lane', libraryOntology['lane_number']),
+ ('read', libraryOntology['read']),
+ ('cycle', libraryOntology['read_length'])]
+ for file_term, model_term in terms:
+ value = fqname.get(file_term)
+ if value is not None:
+ s = RDF.Statement(fileNode, model_term, toTypedNode(value))
+ self.model.append(s)
+
def _add_library_details_to_model(self, libNode):
# attributes that can have multiple values
set_attributes = set((libraryOntology['has_lane'],
-{% for name, value in series %}
-{{name}}={{value}}{% endfor %}{% for row in samples %}
-^SAMPLE={{row.name}}
+{% for name, value in series %}{{name}}={{value}}
+{% endfor %}{% for row in samples %}!Series_sample_id={{row.name}}
+{% endfor %}{% for row in samples %}^SAMPLE={{row.name}}
!Sample_type=SRA
!Sample_title={{row.name}}
!Sample_series_id={{ series_id }}
!Sample_raw_file_checksum_{{forloop.counter}}={{raw.md5sum}}
!sample_raw_file_single_or_paired-end_{{forloop.counter}}={{raw.flowcell_type}}{% endfor %}{% for sup in row.supplimental %}
!Sample_supplementary_file_{{forloop.counter}}={{sup.filename}}
-!Sample_supplementary_file_checksum_{{forloop.counter}}={{sup.md5sum}}
-{% endfor %}{% endfor %}
+!Sample_supplementary_file_checksum_{{forloop.counter}}={{sup.md5sum}}{% endfor %}
+{% endfor %}
\ No newline at end of file
return node
-def toTypedNode(value):
+def toTypedNode(value, language="en"):
"""Convert a python variable to a RDF Node with its closest xsd type
"""
if type(value) == types.BooleanType:
if value_type is not None:
node = RDF.Node(literal=value, datatype=value_type)
else:
- node = RDF.Node(literal=unicode(value).encode('utf-8'))
+ node = RDF.Node(literal=unicode(value).encode('utf-8'), language=language)
return node