def to_graph(subject_specific_dir, project_id, output_dir, new_id=None):
# location of FreeSurfer $SUBJECTS_DIR
basedir = os.path.abspath(subject_specific_dir)
subject_id = basedir.rstrip(os.path.sep).split(os.path.sep)[-1]
graph = prov.ProvBundle()
graph.add_namespace(foaf)
graph.add_namespace(dcterms)
graph.add_namespace(fs)
graph.add_namespace(nidm)
graph.add_namespace(niiri)
graph.add_namespace(obo)
graph.add_namespace(nif)
graph.add_namespace(crypto)
graph = encode_fs_directory(graph, basedir, project_id, subject_id)
provn = graph.get_provn()
old_id = subject_id
if new_id:
provn = provn.replace(subject_id, new_id)
subject_id = new_id
filename = os.path.join(output_dir,
'%s_%s.provn' % (subject_id, project_id))
with open(filename, 'wt') as fp:
fp.writelines(provn)
filename_ttl = os.path.join(output_dir,
'%s_%s.ttl' % (subject_id, project_id))
graph.rdf().serialize(filename_ttl, format='turtle')
if new_id:
map_graph = rdflib.Graph()
map_graph.namespace_manager.bind('fs', fs.get_uri())
map_graph.namespace_manager.bind('nidm', nidm.get_uri())
map_graph.add((fs[new_id].rdf_representation(),
nidm['sameSubjectAs'].rdf_representation(),
fs[old_id].rdf_representation()))
if os.path.exists('mapper.ttl'):
map_graph.parse('mapper.ttl', format='turtle')
map_graph.serialize('mapper.ttl', format='turtle')
return graph, old_id
def cff2provn(filename):
"""Parse cml xml file and return a prov bundle object"""
#filename = "/Users/fariba/Desktop/UCI/freesurfer/scripts/meta-MC-SCA-023_tp1.cml"
tree = xml.dom.minidom.parse(filename)
collections = tree.documentElement
g = prov.ProvBundle()
g.add_namespace(xsd)
g.add_namespace(dcterms)
g.add_namespace(cml)
url_entity = g.entity(cml[get_id()])
url_entity.add_extra_attributes({
prov.PROV['type']:
nidm['nidm:ConnectomeFileFormat'],
prov.PROV['location']:
prov.Literal(filename, prov.XSD['String'])
})
cml_collection = g.collection(cml[get_id()])
cml_collection.add_extra_attributes({
prov.PROV['type']: cml['connectome'],
prov.PROV['label']: filename
})
g.wasDerivedFrom(cml_collection, url_entity)
# get species, subject_name, and subject_timepoint
species = tree.getElementsByTagName('cml:species')[0].toxml()
species = species.replace('<cml:species>',
'').replace('</cml:species>', '')
tp = ''
sub = ''
tags = collections.getElementsByTagName("cml:tag")
for t in tags:
if t.attributes['key'].value == 'subject_name':
sub = t.toxml()
if t.attributes['key'].value == 'subject_timepoint':
tp = t.toxml()
sub = sub.replace('<cml:tag key="subject_name">',
'').replace('</cml:tag>', '')
tp = tp.replace('<cml:tag key="subject_timepoint">',
'').replace('</cml:tag>', '')
#print species + " " + sub + " " + tp
cml_meta = g.entity(cml[get_id()])
cml_meta.add_extra_attributes({
prov.PROV['type']: cml['connectome-meta'],
cml['species']: species,
cml['timepoint']: tp,
cml['subject_name']: sub
})
g.hadMember(cml_collection, cml_meta)
volumes = collections.getElementsByTagName("cml:connectome-volume")
c = 0
for v in volumes:
c = c + 1
#print v.getAttribute("src") + " " + v.getAttribute("dtype") + " " + v.getAttribute("name") + " " + v.getAttribute("fileformat")
#print v.attributes['fileformat'].value
dtype = v.getAttribute('dtype')
src = v.getAttribute('src')
name = v.getAttribute('name')
fileformat = v.getAttribute('fileformat')
cml_volume = g.entity(cml[get_id()])
cml_volume.add_extra_attributes({
prov.PROV['type']:
cml['connectome-volume'],
cml['dtype']:
dtype,
cml['src']:
src,
cml['name']:
name,
cml['fileformat']:
fileformat
})
g.hadMember(cml_collection, cml_volume)
tracks = collections.getElementsByTagName("cml:connectome-track")
c = 0
for t in tracks:
c = c + 1
#print t.getAttribute("src") + " " + t.getAttribute("dtype") + " " + t.getAttribute("name") + " " + t.getAttribute("fileformat")
dtype = t.getAttribute('dtype')
src = t.getAttribute('src')
name = t.getAttribute('name')
fileformat = t.getAttribute('fileformat')
cml_track = g.entity(cml[get_id()])
cml_track.add_extra_attributes({
prov.PROV['type']:
cml['connectome-track'],
cml['dtype']:
dtype,
cml['src']:
src,
cml['name']:
name,
cml['fileformat']:
fileformat
})
g.hadMember(cml_collection, cml_track)
networks = collections.getElementsByTagName("cml:connectome-network")
c = 0
for n in networks:
c = c + 1
#print n.getAttribute("src") + " " + n.getAttribute("dtype") + " " + n.getAttribute("name") + " " + n.getAttribute("fileformat")
dtype = n.getAttribute('dtype')
src = n.getAttribute('src')
name = n.getAttribute('name')
fileformat = n.getAttribute('fileformat')
cml_network = g.entity(cml[get_id()])
cml_network.add_extra_attributes({
prov.PROV['type']:
cml['connectome-network'],
cml['dtype']:
dtype,
cml['src']:
src,
cml['name']:
name,
cml['fileformat']:
fileformat
})
g.hadMember(cml_collection, cml_network)
surfaces = collections.getElementsByTagName("cml:connectome-surface")
c = 0
for s in surfaces:
c = c + 1
#print s.getAttribute("src") + " " + s.getAttribute("dtype") + " " + s.getAttribute("name") + " " + s.getAttribute("fileformat")
dtype = s.getAttribute('dtype')
src = s.getAttribute('src')
name = s.getAttribute('name')
fileformat = s.getAttribute('fileformat')
cml_surface = g.entity(cml[get_id()])
cml_surface.add_extra_attributes({
prov.PROV['type']:
cml['connectome-surface'],
cml['dtype']:
dtype,
cml['src']:
src,
cml['name']:
name,
cml['fileformat']:
fileformat
})
g.hadMember(cml_collection, cml_surface)
data = collections.getElementsByTagName("cml:connectome-data")
c = 0
for d in data:
c = c + 1
#print d.getAttribute("src") + " " + d.getAttribute("dtype") + " " + d.getAttribute("name") + " " + d.getAttribute("fileformat")
dtype = d.getAttribute('dtype')
src = d.getAttribute('src')
name = d.getAttribute('name')
cml_data = g.entity(cml[get_id()])
cml_data.add_extra_attributes({
prov.PROV['type']: cml['connectome-data'],
cml['dtype']: dtype,
cml['src']: src,
cml['name']: name,
cml['fileformat']: fileformat
})
g.hadMember(cml_collection, cml_data)
return g
def __init__(self):
self.g = pm.ProvBundle(identifier=get_id())
self.g.add_namespace(foaf)
self.g.add_namespace(dcterms)
self.g.add_namespace(nipype_ns)
self.g.add_namespace(niiri)
r = requests.get('https://raw.githubusercontent.com/joejimbo/HCLSDatasetDescriptions/master/Overview.html')
soup = BeautifulSoup(r.text)
table = soup.find('table')
codes = table.findChildren('code')
code_block = [codes[pos:pos + 2] for pos in xrange(0, len(codes), 2)]
ns_map = {code[0].text[:-1]: code[1].text for code in code_block}
# Access namespace objects as attributes
class AttrDict(dict):
def __init__(self, *args, **kwargs):
super(AttrDict, self).__init__(*args, **kwargs)
self.__dict__ = self
# create a prov bundle to store the graph
bundle = prov.ProvBundle()
# add namespaces to the bundle
for k, v in ns_map.iteritems():
ns = prov.Namespace(k, v)
bundle.add_namespace(ns)
ns = AttrDict(bundle._namespaces)
# Core Metadata
# property: [summary, version, distribution]
core_metadata = {ns.rdf['type']: ["MUST", "MUST", "MUST"],
ns.dct['title']: ["MUST", "MUST", "MUST"],
ns.dct['alternative']: ["MAY", "MAY", "MAY"],
ns.dct['description']: ["MUST", "MUST", "MUST"],
ns.dct['created']: ["NEVER", "SHOULD", "SHOULD"],