def download_articles(self):
"""
Function that downloads articles' text.
:return: Results list.
"""
logger.info('')
m = multiprocessing.Manager()
results = m.list()
utils.async_getter(
function_pointer=self.download_article,
input_list=self.parent.articles,
pass_data={"results": results},
)
return results
def get_rss_feeds(self, num_feeds_max):
"""
Function that downloads RSS feeds (links are stored in RSS_FEEDS),
and returns articles in the required format in a list.
:param num_feeds_max: Maximum number of feeds.
:return: Articles in a list.
"""
logger.info('num_feeds_max:%s' % num_feeds_max)
m = multiprocessing.Manager()
articles = m.list()
rss_feeds = RSS_FEEDS[:num_feeds_max]
utils.async_getter(function_pointer=self.get_rss_feed,
input_list=rss_feeds,
pass_data={"results": articles})
return articles
def main():
files = glob(os.path.expanduser('~/git/methodsOntology-upstream/to_be_integrated_in_NIF/*'))
rows = []
got_header = False
for file in files:
with open(file, 'rt') as f:
r = [r for r in csv.reader(f, delimiter='|')]
if got_header:
r = r[1:]
else:
got_header = True
rows.extend(r)
def async_func(row):
resps = sgv.findByTerm(row[2])
if resps:
n = resps[0]
c, l = n['curie'], n['labels'][0]
else:
c, l = None, None
r = row + [c, l]
return r
matched = [rows[0] + ['e_curie', 'e_label']] + async_getter(async_func, [(r,) for r in rows[1:]])
embed()
def main():
files = glob(os.path.expanduser('~/git/methodsOntology-upstream/to_be_integrated_in_NIF/*'))
rows = []
got_header = False
for file in files:
with open(file, 'rt') as f:
r = [r for r in csv.reader(f, delimiter='|')]
if got_header:
r = r[1:]
else:
got_header = True
rows.extend(r)
def async_func(row):
resps = sgv.findByTerm(row[2])
if resps:
n = resps[0]
c, l = n['curie'], n['labels'][0]
else:
c, l = None, None
r = row + [c, l]
return r
matched = [rows[0] + ['e_curie', 'e_label']] + async_getter(async_func, [(r,) for r in rows[1:]])
embed()
def swanson():
""" not really a parcellation scheme """
source = 'resources/swanson_aligned.txt'
ONT_PATH = GENERATED
filename = 'swanson_hierarchies'
ontid = ONT_PATH + filename + '.ttl'
PREFIXES = makePrefixes('', 'ilx', 'owl', 'skos', 'NIFRID', 'ILXREPLACE')
PREFIXES.update({
#'':ontid + '/', # looking for better options
'SWAN': interlex_namespace('swanson/nt/term'),
'SWAA': interlex_namespace('swanson/nt/appendix'),
})
new_graph = makeGraph(filename, PREFIXES, writeloc=WRITELOC)
new_graph.add_ont(
ontid, 'Swanson brain partomies', 'Swanson 2014 Partonomies',
'This file is automatically generated from ' + source + '.' + NOTICE,
TODAY)
# FIXME citations should really go on the ... anatomy? scheme artifact
definingCitation = 'Swanson, Larry W. Neuroanatomical Terminology: a lexicon of classical origins and historical foundations. Oxford University Press, USA, 2014.'
definingCitationID = 'ISBN:9780195340624'
new_graph.add_trip(ontid, 'NIFRID:definingCitation', definingCitation)
new_graph.add_trip(ontid, 'NIFRID:definingCitationID', definingCitationID)
with open(source, 'rt') as f:
lines = [l.strip() for l in f.readlines()]
# join header on page 794
lines[635] += ' ' + lines.pop(636)
#fix for capitalization since this header is reused
fixed = ' or '.join([
' ('.join([n.capitalize() for n in _.split(' (')])
for _ in lines[635].lower().split(' or ')
]).replace('human', 'HUMAN')
lines[635] = fixed
data = []
for l in lines:
if not l.startswith('#'):
level = l.count('.' * 5)
l = l.strip('.')
if ' (' in l:
if ') or' in l:
n1, l = l.split(') or')
area_name, citationP = n1.strip().split(' (')
citation = citationP.rstrip(')')
d = (level, area_name, citation, 'NEXT SYN')
data.append(d)
#print(tc.red(tc.bold(repr(d))))
area_name, citationP = l.strip().split(' (')
citation = citationP.rstrip(')')
else:
area_name = l
citation = None
d = (level, area_name, citation, None)
#print(d)
data.append(d)
results = async_getter(sgv.findByTerm, [(d[1], ) for d in data])
#results = [None] * len(data)
curies = [[r['curie'] for r in _ if 'UBERON' in r['curie']] if _ else []
for _ in results]
output = [_[0] if _ else None for _ in curies]
header = ['Depth', 'Name', 'Citation', 'NextSyn', 'Uberon']
zoop = [header] + [r for r in zip(*zip(*data), output)] + \
[(0, 'Appendix END None', None, None, None)] # needed to add last appendix
class SP(rowParse):
def __init__(self):
self.nodes = defaultdict(dict)
self._appendix = 0
self.appendicies = {}
self._last_at_level = {}
self.names = defaultdict(set)
self.children = defaultdict(set)
self.parents = defaultdict(set)
self.next_syn = False
super().__init__(zoop)
def Depth(self, value):
if self.next_syn:
self.synonym = self.next_syn
else:
self.synonym = False
self.depth = value
def Name(self, value):
self.name = value
def Citation(self, value):
self.citation = value
def NextSyn(self, value):
if value:
self.next_syn = self._rowind
else:
self.next_syn = False
def Uberon(self, value):
self.uberon = value
def _row_post(self):
# check if we are in the next appendix
# may want to xref ids between appendicies as well...
if self.depth == 0:
if self.name.startswith('Appendix'):
if self._appendix:
self.appendicies[self._appendix]['children'] = dict(
self.children)
self.appendicies[self._appendix]['parents'] = dict(
self.parents)
self._last_at_level = {}
self.children = defaultdict(set)
self.parents = defaultdict(set)
_, num, apname = self.name.split(' ', 2)
if num == 'END':
return
self._appendix = int(num)
self.appendicies[self._appendix] = {
'name':
apname.capitalize(),
'type':
self.citation.capitalize() if self.citation else None
}
return
else:
if ' [' in self.name:
name, taxonB = self.name.split(' [')
self.name = name
self.appendicies[self._appendix][
'taxon'] = taxonB.rstrip(']').capitalize()
else: # top level is animalia
self.appendicies[
self._appendix]['taxon'] = 'ANIMALIA'.capitalize()
self.name = self.name.capitalize()
self.citation = self.citation.capitalize()
# nodes
if self.synonym:
self.nodes[self.synonym]['synonym'] = self.name
self.nodes[self.synonym]['syn-cite'] = self.citation
self.nodes[self.synonym]['syn-uberon'] = self.uberon
return
else:
if self.citation: # Transverse Longitudinal etc all @ lvl4
self.names[self.name + ' ' + self.citation].add(
self._rowind)
else:
self.name += str(self._appendix) + self.nodes[
self._last_at_level[self.depth - 1]]['label']
#print(level, self.name)
# can't return here because they are their own level
# replace with actually doing something...
self.nodes[self._rowind]['label'] = self.name
self.nodes[self._rowind]['citation'] = self.citation
self.nodes[self._rowind]['uberon'] = self.uberon
# edges
self._last_at_level[self.depth] = self._rowind
# TODO will need something to deal with the Lateral/
if self.depth > 0:
try:
parent = self._last_at_level[self.depth - 1]
except:
embed()
self.children[parent].add(self._rowind)
self.parents[self._rowind].add(parent)
def _end(self):
replace = {}
for asdf in [
sorted(n) for k, n in self.names.items() if len(n) > 1
]:
replace_with, to_replace = asdf[0], asdf[1:]
for r in to_replace:
replace[r] = replace_with
for r, rw in replace.items():
#print(self.nodes[rw])
o = self.nodes.pop(r)
#print(o)
for vals in self.appendicies.values():
children = vals['children']
parents = vals['parents']
# need reversed so children are corrected before swap
for r, rw in reversed(sorted(replace.items())):
if r in parents:
child = r
new_child = rw
parent = parents.pop(child)
parents[new_child] = parent
parent = list(parent)[0]
children[parent].remove(child)
children[parent].add(new_child)
if r in children:
parent = r
new_parent = rw
childs = children.pop(parent)
children[new_parent] = childs
for child in childs:
parents[child] = {new_parent}
self.nodes = dict(self.nodes)
sp = SP()
tp = [
_ for _ in sorted([
'{: <50}'.format(n['label']) +
n['uberon'] if n['uberon'] else n['label']
for n in sp.nodes.values()
])
]
#print('\n'.join(tp))
#print(sp.appendicies[1].keys())
#print(sp.nodes[1].keys())
nbase = PREFIXES['SWAN'] + '%s'
json_ = {'nodes': [], 'edges': []}
parent = ILXREPLACE('swansonBrainRegionConcept')
for node, anns in sp.nodes.items():
nid = nbase % node
new_graph.add_class(nid, parent, label=anns['label'])
new_graph.add_trip(nid, 'NIFRID:definingCitation', anns['citation'])
json_['nodes'].append({'lbl': anns['label'], 'id': 'SWA:' + str(node)})
#if anns['uberon']:
#new_graph.add_trip(nid, owl.equivalentClass, anns['uberon']) # issues arrise here...
for appendix, data in sp.appendicies.items():
aid = PREFIXES['SWAA'] + str(appendix)
new_graph.add_class(aid, label=data['name'].capitalize())
new_graph.add_trip(
aid, 'ilx:hasTaxonRank',
data['taxon']) # FIXME appendix is the data artifact...
children = data['children']
ahp = HASPART + str(appendix)
apo = PARTOF + str(appendix)
new_graph.add_op(ahp, transitive=True)
new_graph.add_op(apo, inverse=ahp, transitive=True)
for parent, childs in children.items(
): # FIXME does this give complete coverage?
pid = nbase % parent
for child in childs:
cid = nbase % child
new_graph.add_hierarchy(
cid, ahp, pid) # note hierarhcy inverts direction
new_graph.add_hierarchy(pid, apo, cid)
json_['edges'].append({
'sub': 'SWA:' + str(child),
'pred': apo,
'obj': 'SWA:' + str(parent)
})
new_graph.write()
if False:
Query = namedtuple('Query',
['root', 'relationshipType', 'direction', 'depth'])
mapping = (1, 1, 1, 1, 30, 83, 69, 70, 74, 1) # should generate?
for i, n in enumerate(mapping):
a, b = creatTree(*Query('SWA:' + str(n), 'ilx:partOf' + str(i + 1),
'INCOMING', 10),
json=json_)
print(a)
return ontid, None
def make_nifga_graph(_doprint=False):
# use equivalent class mappings to build a replacement mapping
g = rdflib.Graph()
g.parse(nifga_path, format='turtle')
getQname = g.namespace_manager.qname
classes = sorted([
getQname(_) for _ in g.subjects(RDF.type, OWL.Class)
if type(_) is URIRef
])
curies = ['NIFGA:' + n for n in classes if ':' not in n]
matches = async_getter(sgv.findById, [(c, ) for c in curies])
replaced_by = {}
exact = {}
internal_equivs = {}
irbcs = {}
def equiv(curie, label):
if curie in manual:
replaced_by[curie] = manual[curie]
return manual[curie]
ec = sgg.getNeighbors(curie, relationshipType='equivalentClass')
nodes = [n for n in ec['nodes'] if n['id'] != curie]
if len(nodes) > 1:
#print('wtf node', [n['id'] for n in nodes], curie)
for node in nodes:
id_ = node['id']
label_ = node['lbl']
if id_.startswith('UBERON'):
if curie in replaced_by:
one = replaced_by[curie]
replaced_by[curie] = one, id_
print('WE GOT DUPES', curie, label, one, id_) # TODO
else:
replaced_by[curie] = id_
else:
internal_equivs[curie] = id_
elif not nodes:
node = sgg.getNode(curie)['nodes'][0]
if OWL.deprecated.toPython() in node['meta']:
print('THIS CLASS IS DEPRECATED', curie)
lbl = node['lbl']
if lbl.startswith(
'Predominantly white regional') or lbl.startswith(
'Predominantly gray regional'):
print('\tHE\'S DEAD JIM!', lbl, node['id'])
replaced_by[curie] = 'NOREP'
if IRBC in node['meta']:
existing_replaced = node['meta'][IRBC][0]
ec2 = sgg.getNeighbors(existing_replaced,
relationshipType='equivalentClass')
print('\tFOUND ONE', existing_replaced)
#scigPrint.pprint_node(sgg.getNode(existing_replaced))
if ec2['edges']: # pass the buck if we can
print('\t', end='')
scigPrint.pprint_edge(ec2['edges'][0])
rb = ec2['edges'][0]['obj']
print('\tPASSING BUCK : (%s -> %s -> %s)' %
(curie, existing_replaced, rb))
irbcs[curie] = (existing_replaced, rb)
replaced_by[curie] = rb
return nodes
else:
er_node = sgv.findById(existing_replaced)
if not er_node['deprecated']:
if not er_node['curie'].startswith('NIFGA:'):
print('\tPASSING BUCK : (%s -> %s)' %
(curie, er_node['curie']))
return nodes
print(
'\tERROR: could not pass buck, we are at a dead end at',
er_node) # TODO
print()
moar = [
t for t in sgv.findByTerm(label)
if t['curie'].startswith('UBERON')
]
if moar:
#print(moar)
#replaced_by[curie] = moar[0]['curie']
if len(moar) > 1:
print('WARNING', curie, label,
[(m['curie'], m['labels'][0]) for m in moar])
for node in moar:
#if node['curie'] in uberon_obsolete: # node['deprecated']?
#continue
ns = sgg.getNode(node['curie'])
assert len(
ns['nodes']) == 1, "WTF IS GOING ON %s" % node['curie']
ns = ns['nodes'][0]
if _doprint:
print(
'Found putative replacement in moar: (%s -> %s)' %
(curie, ns['id']))
if DBX in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][DBX],
node['labels'][0], node['synonyms'])
if AID in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][AID],
node['labels'][0], node['synonyms'])
if CON in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][CON],
node['labels'][0], node['synonyms'])
replaced_by[curie] = ns['id']
else:
replaced_by[curie] = None
if False: # review
print('NO FORWARD EQUIV', tc.red(curie), label) # TODO
for k, v in sorted(
sgg.getNode(curie)['nodes'][0]['meta'].items()):
if type(v) == iter:
print(' ' * 4, k)
for _ in v:
print(' ' * 8, _)
else:
print(' ' * 4, k, v)
else:
node = nodes[0]
replaced_by[curie] = node['id']
exact[curie] = node['id']
return nodes
equivs = [equiv(c['curie'], c['labels'][0])
for c in matches] # async causes print issues :/
return g, matches, exact, internal_equivs, irbcs, replaced_by
curies = c.getCuries()
curies.pop('') # don't want NIFSTD uris just yet
with open(os.path.expanduser('~/git/nlxeol/neurolex_full.csv'), 'rt') as f:
rows = [r for r in csv.reader(f)]
Id = rows[0].index('Id')
ids = [(r[Id],) for r in rows if r[Id] and r[Id] != 'Id' and 'Resource:' not in r[0]]
items = tuple(curies.items())
findById = v.findById
def async_func(id_):
if ':' in id_:
out = findById(id_)
if out:
return id_, out['curie']
for prefix, uri in items:
curie = prefix + ':' + id_
out = findById(curie)
if out:
return id_, out['curie']
return id_, 'NLXONLY'
id_curie = async_getter(async_func, ids)
j = {id_:curie for id_, curie in id_curie}
with open('/tmp/total_curie_fragment.json', 'wt') as f:
json.dump(j, f, sort_keys=True, indent=4)
with open(os.path.expanduser('~/git/nlxeol/neurolex_full.csv'), 'rt') as f:
rows = [r for r in csv.reader(f)]
Id = rows[0].index('Id')
ids = [(r[Id], ) for r in rows
if r[Id] and r[Id] != 'Id' and 'Resource:' not in r[0]]
items = tuple(curies.items())
findById = v.findById
def async_func(id_):
if ':' in id_:
out = findById(id_)
if out:
return id_, out['curie']
for prefix, uri in items:
curie = prefix + ':' + id_
out = findById(curie)
if out:
return id_, out['curie']
return id_, 'NLXONLY'
id_curie = async_getter(async_func, ids)
j = {id_: curie for id_, curie in id_curie}
with open('/tmp/total_curie_fragment.json', 'wt') as f:
json.dump(j, f, sort_keys=True, indent=4)
def swanson():
""" not really a parcellation scheme """
ONT_PATH = 'http://ontology.neuinfo.org/NIF/ttl/generated/'
filename = 'swanson_hierarchies'
ontid = ONT_PATH + filename + '.ttl'
PREFIXES = makePrefixes('ilx', 'owl', 'OBOANN', 'UBERON')
PREFIXES.update({
'':ontid + '/', # looking for better options
'SWAN':'http://swanson.org/node/',
'SWAA':'http://swanson.org/appendix/',
})
new_graph = makeGraph(filename, PREFIXES, writeloc='/tmp/parc/')
new_graph.add_ont(ontid,
'Swanson brain partomies',
'Swanson 2014 Partonomies',
'This file is automatically generated from....',
TODAY)
with open('resources/swanson_aligned.txt', 'rt') as f:
lines = [l.strip() for l in f.readlines()]
# join header on page 794
lines[635] += ' ' + lines.pop(636)
#fix for capitalization since this header is reused
fixed = ' or '.join([' ('.join([n.capitalize() for n in _.split(' (')]) for _ in lines[635].lower().split(' or ')]).replace('human','HUMAN')
lines[635] = fixed
data = []
for l in lines:
if not l.startswith('#'):
level = l.count('.'*5)
l = l.strip('.')
if ' (' in l:
if ') or' in l:
n1, l = l.split(') or')
area_name, citationP = n1.strip().split(' (')
citation = citationP.rstrip(')')
d = (level, area_name, citation, 'NEXT SYN')
data.append(d)
#print(tc.red(tc.bold(repr(d))))
area_name, citationP = l.strip().split(' (')
citation = citationP.rstrip(')')
else:
area_name = l
citation = None
d = (level, area_name, citation, None)
#print(d)
data.append(d)
results = async_getter(sgv.findByTerm, [(d[1],) for d in data])
#results = [None] * len(data)
curies = [[r['curie'] for r in _ if 'UBERON' in r['curie']] if _ else [] for _ in results]
output = [_[0] if _ else None for _ in curies]
header = ['Depth', 'Name', 'Citation', 'NextSyn', 'Uberon']
zoop = [header] + [r for r in zip(*zip(*data), output)] + \
[(0, 'Appendix END None', None, None, None)] # needed to add last appendix
class SP(rowParse):
def __init__(self):
self.nodes = defaultdict(dict)
self._appendix = 0
self.appendicies = {}
self._last_at_level = {}
self.names = defaultdict(set)
self.children = defaultdict(set)
self.parents = defaultdict(set)
self.next_syn = False
super().__init__(zoop)
def Depth(self, value):
if self.next_syn:
self.synonym = self.next_syn
else:
self.synonym = False
self.depth = value
def Name(self, value):
self.name = value
def Citation(self, value):
self.citation = value
def NextSyn(self, value):
if value:
self.next_syn = self._rowind
else:
self.next_syn = False
def Uberon(self, value):
self.uberon = value
def _row_post(self):
# check if we are in the next appendix
# may want to xref ids between appendicies as well...
if self.depth == 0:
if self.name.startswith('Appendix'):
if self._appendix:
self.appendicies[self._appendix]['children'] = dict(self.children)
self.appendicies[self._appendix]['parents'] = dict(self.parents)
self._last_at_level = {}
self.children = defaultdict(set)
self.parents = defaultdict(set)
_, num, apname = self.name.split(' ', 2)
if num == 'END':
return
self._appendix = int(num)
self.appendicies[self._appendix] = {
'name':apname.capitalize(),
'type':self.citation.capitalize() if self.citation else None}
return
else:
if ' [' in self.name:
name, taxonB = self.name.split(' [')
self.name = name
self.appendicies[self._appendix]['taxon'] = taxonB.rstrip(']').capitalize()
else: # top level is animalia
self.appendicies[self._appendix]['taxon'] = 'ANIMALIA'.capitalize()
self.name = self.name.capitalize()
self.citation = self.citation.capitalize()
# nodes
if self.synonym:
self.nodes[self.synonym]['synonym'] = self.name
self.nodes[self.synonym]['syn-cite'] = self.citation
self.nodes[self.synonym]['syn-uberon'] = self.uberon
return
else:
if self.citation: # Transverse Longitudinal etc all @ lvl4
self.names[self.name + ' ' + self.citation].add(self._rowind)
else:
self.name += str(self._appendix) + self.nodes[self._last_at_level[self.depth - 1]]['label']
#print(level, self.name)
# can't return here because they are their own level
# replace with actually doing something...
self.nodes[self._rowind]['label'] = self.name
self.nodes[self._rowind]['citation'] = self.citation
self.nodes[self._rowind]['uberon'] = self.uberon
# edges
self._last_at_level[self.depth] = self._rowind
# TODO will need something to deal with the Lateral/
if self.depth > 0:
try:
parent = self._last_at_level[self.depth - 1]
except:
embed()
self.children[parent].add(self._rowind)
self.parents[self._rowind].add(parent)
def _end(self):
replace = {}
for asdf in [sorted(n) for k,n in self.names.items() if len(n) > 1]:
replace_with, to_replace = asdf[0], asdf[1:]
for r in to_replace:
replace[r] = replace_with
for r, rw in replace.items():
#print(self.nodes[rw])
o = self.nodes.pop(r)
#print(o)
for vals in self.appendicies.values():
children = vals['children']
parents = vals['parents']
# need reversed so children are corrected before swap
for r, rw in reversed(sorted(replace.items())):
if r in parents:
child = r
new_child = rw
parent = parents.pop(child)
parents[new_child] = parent
parent = list(parent)[0]
children[parent].remove(child)
children[parent].add(new_child)
if r in children:
parent = r
new_parent = rw
childs = children.pop(parent)
children[new_parent] = childs
for child in childs:
parents[child] = {new_parent}
self.nodes = dict(self.nodes)
sp = SP()
tp = [_ for _ in sorted(['{: <50}'.format(n['label']) + n['uberon'] if n['uberon'] else n['label'] for n in sp.nodes.values()])]
#print('\n'.join(tp))
#print(sp.appendicies[1].keys())
#print(sp.nodes[1].keys())
nbase = 'http://swanson.org/node/%s'
json_ = {'nodes':[],'edges':[]}
for node, anns in sp.nodes.items():
nid = nbase % node
new_graph.add_class(nid, 'ilx:swansonBrainRegionConcept', label=anns['label'])
new_graph.add_node(nid, 'OBOANN:definingCitation', anns['citation'])
json_['nodes'].append({'lbl':anns['label'],'id':'SWA:' + str(node)})
#if anns['uberon']:
#new_graph.add_node(nid, rdflib.OWL.equivalentClass, anns['uberon']) # issues arrise here...
for appendix, data in sp.appendicies.items():
aid = 'http://swanson.org/appendix/%s' % appendix
new_graph.add_class(aid, label=data['name'].capitalize())
new_graph.add_node(aid, 'ilx:hasTaxonRank', data['taxon']) # FIXME appendix is the data artifact...
children = data['children']
ahp = HASPART + str(appendix)
apo = PARTOF + str(appendix)
new_graph.add_op(ahp, transitive=True)
new_graph.add_op(apo, inverse=ahp, transitive=True)
for parent, childs in children.items(): # FIXME does this give complete coverage?
pid = nbase % parent
for child in childs:
cid = nbase % child
new_graph.add_hierarchy(cid, ahp, pid) # note hierarhcy inverts direction
new_graph.add_hierarchy(pid, apo, cid)
json_['edges'].append({'sub':'SWA:' + str(child),'pred':apo,'obj':'SWA:' + str(parent)})
new_graph.write(convert=False)
if False:
Query = namedtuple('Query', ['root','relationshipType','direction','depth'])
mapping = (1, 1, 1, 1, 30, 83, 69, 70, 74, 1) # should generate?
for i, n in enumerate(mapping):
a, b = creatTree(*Query('SWA:' + str(n), 'ilx:partOf' + str(i + 1), 'INCOMING', 10), json=json_)
print(a)
return ontid, None
def make_nifga_graph(_doprint=False):
# use equivalent class mappings to build a replacement mapping
g = rdflib.Graph()
g.parse(nifga_path, format='turtle')
getQname = g.namespace_manager.qname
classes = sorted([getQname(_) for _ in g.subjects(RDF.type, OWL.Class) if type(_) is URIRef])
curies = ['NIFGA:' + n for n in classes if ':' not in n]
matches = async_getter(sgv.findById, [(c,) for c in curies])
replaced_by = {}
exact = {}
internal_equivs = {}
irbcs = {}
def equiv(curie, label):
if curie in manual:
replaced_by[curie] = manual[curie]
return manual[curie]
ec = sgg.getNeighbors(curie, relationshipType='equivalentClass')
nodes = [n for n in ec['nodes'] if n['id'] != curie]
if len(nodes) > 1:
#print('wtf node', [n['id'] for n in nodes], curie)
for node in nodes:
id_ = node['id']
label_ = node['lbl']
if id_.startswith('UBERON'):
if curie in replaced_by:
one = replaced_by[curie]
replaced_by[curie] = one, id_
print('WE GOT DUPES', curie, label, one, id_) # TODO
else:
replaced_by[curie] = id_
else:
internal_equivs[curie] = id_
elif not nodes:
node = sgg.getNode(curie)['nodes'][0]
if OWL.deprecated.toPython() in node['meta']:
print('THIS CLASS IS DEPRECATED', curie)
lbl = node['lbl']
if lbl.startswith('Predominantly white regional') or lbl.startswith('Predominantly gray regional'):
print('\tHE\'S DEAD JIM!', lbl, node['id'])
replaced_by[curie] = 'NOREP'
if IRBC in node['meta']:
existing_replaced = node['meta'][IRBC][0]
ec2 = sgg.getNeighbors(existing_replaced, relationshipType='equivalentClass')
print('\tFOUND ONE', existing_replaced)
#scigPrint.pprint_node(sgg.getNode(existing_replaced))
if ec2['edges']: # pass the buck if we can
print('\t',end='')
scigPrint.pprint_edge(ec2['edges'][0])
rb = ec2['edges'][0]['obj']
print('\tPASSING BUCK : (%s -> %s -> %s)' % (curie, existing_replaced, rb))
irbcs[curie] = (existing_replaced, rb)
replaced_by[curie] = rb
return nodes
else:
er_node = sgv.findById(existing_replaced)
if not er_node['deprecated']:
if not er_node['curie'].startswith('NIFGA:'):
print('\tPASSING BUCK : (%s -> %s)' % (curie, er_node['curie']))
return nodes
print('\tERROR: could not pass buck, we are at a dead end at', er_node) # TODO
print()
moar = [t for t in sgv.findByTerm(label) if t['curie'].startswith('UBERON')]
if moar:
#print(moar)
#replaced_by[curie] = moar[0]['curie']
if len(moar) > 1:
print('WARNING', curie, label, [(m['curie'], m['labels'][0]) for m in moar])
for node in moar:
#if node['curie'] in uberon_obsolete: # node['deprecated']?
#continue
ns = sgg.getNode(node['curie'])
assert len(ns['nodes']) == 1, "WTF IS GOING ON %s" % node['curie']
ns = ns['nodes'][0]
if _doprint:
print('Found putative replacement in moar: (%s -> %s)' % (curie, ns['id']))
if DBX in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][DBX],
node['labels'][0], node['synonyms'])
if AID in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][AID],
node['labels'][0], node['synonyms'])
if CON in ns['meta']:
print(' ' * 8, node['curie'], ns['meta'][CON],
node['labels'][0], node['synonyms'])
replaced_by[curie] = ns['id']
else:
replaced_by[curie] = None
if False: # review
print('NO FORWARD EQUIV', tc.red(curie), label) # TODO
for k,v in sorted(sgg.getNode(curie)['nodes'][0]['meta'].items()):
if type(v) == iter:
print(' ' * 4, k)
for _ in v:
print(' ' * 8, _)
else:
print(' ' * 4, k, v)
else:
node = nodes[0]
replaced_by[curie] = node['id']
exact[curie] = node['id']
return nodes
equivs = [equiv(c['curie'], c['labels'][0]) for c in matches] # async causes print issues :/
return g, matches, exact, internal_equivs, irbcs, replaced_by
