def flag_dep(json_):
for node in json_['nodes']:
if DEP in node['meta']:
curie = node['id']
label = node['lbl']
node['id'] = tc.red(curie)
node['lbl'] = tc.red(label)
for edge in json_['edges']:
if edge['sub'] == curie:
edge['sub'] = tc.red(curie)
elif edge['obj'] == curie:
edge['obj'] = tc.red(curie)
def flag_dep(json_):
for node in json_['nodes']:
if DEP in node['meta']:
curie = node['id']
label = node['lbl']
node['id'] = tc.red(curie)
node['lbl'] = tc.red(label)
for edge in json_['edges']:
if edge['sub'] == curie:
edge['sub'] = tc.red(curie)
elif edge['obj'] == curie:
edge['obj'] = tc.red(curie)
def spell(filenames, debug=False):
if hunspell is None:
raise ImportError('hunspell is not installed on your system. If you want '
'to run `ontutils spell` please run pipenv install --dev --skip-lock. '
'You will need the development libs for hunspell on your system.')
spell_objects = (u for r in Parallel(n_jobs=9)(delayed(get_spells)(f) for f in filenames) for u in r)
hobj = hunspell.HunSpell('/usr/share/hunspell/en_US.dic', '/usr/share/hunspell/en_US.aff')
#nobj = hunspell.HunSpell(os.path.expanduser('~/git/domain_wordlists/neuroscience-en.dic'), '/usr/share/hunspell/en_US.aff') # segfaults without aff :x
collect = set()
for filename, s, p, o in spell_objects:
missed = False
no = []
for line in o.split('\n'):
nline = []
for tok in line.split(' '):
prefix, tok, suffix = tokstrip(tok)
#print((prefix, tok, suffix))
if not hobj.spell(tok):# and not nobj.spell(tok):
missed = True
collect.add(tok)
nline.append(prefix + tc.red(tok) + suffix)
else:
nline.append(prefix + tok + suffix)
line = ' '.join(nline)
no.append(line)
o = '\n'.join(no)
if missed:
#print(filename, s, o)
print('>>>', o)
if debug:
[print(_) for _ in sorted(collect)]
breakpoint()
def _ontology_local_repo(self):
try:
stated_repo = Path(self.config['ontology_local_repo'])
except (KeyError, TypeError, FileNotFoundError) as e:
stated_repo = Path('/dev/null/does-not-exist')
maybe_repo = self._maybe_repo
if stated_repo.exists():
return stated_repo
elif maybe_repo.exists():
return maybe_repo
else:
maybe_start = Path(__file__).parent.parent.parent.absolute()
maybe_base = maybe_start
fsroot = Path('/')
while maybe_base != fsroot:
maybe_repo = maybe_base / self.ontology_repo
if maybe_repo.exists():
log.info(
tc.blue('INFO:') +
f'Ontology repository found at {maybe_repo}')
return maybe_repo
else:
maybe_base = maybe_base.parent
else:
log.warning(
tc.red('WARNING:') +
f'No repository found in any parent directory of {maybe_start}'
)
return Path('/dev/null/does-not-exist') # seems reaonsable ...
def parents(self):
if not self.orphaned:
for aid in self.anno.references:
try:
yield self._getAATById(aid)
except KeyError:
self.orphaned = True
print(tc.red('WARNING:'), f'parent annotation was deleted from {self.id}')
def fetch_and_save(url, loc):
resp = requests.get(url)
saveloc = (path / loc).as_posix()
if resp.ok:
with open(saveloc, 'wb') as f:
f.write(resp.content)
print(tc.blue('INFO:'), f'{url:<60} written to {loc}')
else:
print(tc.red('WARNING:'), f'failed to fetch {url}')
def sv(asdf, start, ind, warn=False):
if type(asdf) not in (bool, int) and asdf.startswith('http'):
for iri, short in scigPrint.shorten.items():
if iri in asdf:
return scigPrint.wrap(asdf.replace(iri, short + ':'),
start, ind)
print(tc.red('WARNING:'), 'Shorten failed for', tc.ltyellow(asdf))
return scigPrint.wrap(repr(asdf), start, ind)
else:
return scigPrint.wrap(repr(asdf), start, ind)
def import_tree(graph, ontologies, **kwargs):
for ontology in ontologies:
thisfile = Path(ontology).name
print(thisfile)
OntCuries.populate(graph)
j = graph.asOboGraph('owl:imports', restriction=False)
try:
t, te = creatTree(*Query(f'NIFTTL:{thisfile}', 'owl:imports', 'OUTGOING', 30), json=j, prefixes=dict(graph.namespace_manager), **kwargs)
#print(t)
yield t, te
except KeyError:
print(tc.red('WARNING:'), 'could not find', ontology, 'in import chain') # TODO zap onts w/o imports
def build_transgenic_lines(self):
"""
init class | "transgenic_line_source_name":"stock_number" a Class
add superClass | rdfs:subClassOf ilxtr:transgenicLine
add *order* | ilxtr:useObjectProperty ilxtr:<order>
add name | rdfs:label "name"
add def | definition: "description"
add transtype | rdfs:hasTransgenicType "transgenic_line_type_name"
"""
triples = []
for cell_specimen in self.neuron_data:
for tl in cell_specimen['donor']['transgenic_lines']:
_id = tl['stock_number'] if tl['stock_number'] else tl['id']
prefix = tl['transgenic_line_source_name']
line_type = tl['transgenic_line_type_name']
if line_type == 'driver' and 'CreERT2' in tl['name']:
line_type = 'inducibleDriver'
if prefix not in ['JAX', 'MMRRC', 'AIBS']:
print(tc.red('WARNING:'), 'unknown prefix', prefix,
json.dumps(tl, indent=4))
continue
elif prefix == 'AIBS':
prefix = 'AllenTL'
_class = self.ns[prefix][str(_id)]
triples.append((_class, rdf.type, owl.Class))
triples.append(
(_class, rdfs.label, rdflib.Literal(tl['name'])))
triples.append(
(_class, definition, rdflib.Literal(tl['description'])))
triples.append((_class, rdfs.subClassOf, ilxtr.transgenicLine))
triples.append((_class, ilxtr.hasTransgenicType,
ilxtr[line_type + 'Line']))
# TODO aspects.ttl?
transgenic_lines = simpleOnt(
filename='allen-transgenic-lines',
local_base=graphBase.local_base,
path='ttl/generated/',
prefixes=self.prefixes,
triples=triples,
comment='Allen transgenic lines for cell types',
branch=self.branch,
calling__file__=__file__,
)
transgenic_lines._graph.write()
def spell(filenames, debug=False):
if hunspell is None:
raise ImportError(
'hunspell is not installed on your system. If you want '
'to run `ontutils spell` please run pipenv install --dev --skip-lock. '
'You will need the development libs for hunspell on your system.')
hobj = hunspell.HunSpell('/usr/share/hunspell/en_US.dic',
'/usr/share/hunspell/en_US.aff')
#nobj = hunspell.HunSpell(os.path.expanduser('~/git/domain_wordlists/neuroscience-en.dic'), '/usr/share/hunspell/en_US.aff') # segfaults without aff :x
collect = set()
for filename in filenames:
missed = False
no = []
with open(filename, 'rt') as f:
for line_ in f.readlines():
line = line_.rstrip()
nline = []
#print(tc.blue(line))
for pattern in _bads:
line = line.replace(pattern, ' ' * len(pattern))
#print(line)
for tok in line.split(' '):
prefix, tok, suffix = tokstrip(tok)
#print((prefix, tok, suffix))
if not hobj.spell(tok): # and not nobj.spell(tok):
missed = True
collect.add(tok)
nline.append(prefix + tc.red(tok) + suffix)
else:
nline.append(prefix + tok + suffix)
line = ' '.join(nline)
no.append(line)
o = '\n'.join(no)
if missed:
#print(filename, s, o)
print('>>>', o)
pass
if debug:
[print(_) for _ in sorted(collect)]
breakpoint()
def import_tree(graph, ontologies, **kwargs):
for ontology in ontologies:
thisfile = Path(ontology).name
print(thisfile)
mg = makeGraph('', graph=graph)
mg.add_known_namespaces('owl', 'obo', 'dc', 'dcterms', 'dctypes',
'skos', 'NIFTTL')
j = mg.make_scigraph_json('owl:imports', direct=True)
try:
t, te = creatTree(*Query(f'NIFTTL:{thisfile}', 'owl:imports',
'OUTGOING', 30),
json=j,
prefixes=mg.namespaces,
**kwargs)
#print(t)
yield t, te
except KeyError:
print(tc.red('WARNING:'), 'could not find', ontology,
'in import chain') # TODO zap onts w/o imports
def print_trees(graph, bridge):
PPO = 'ro:proper_part_of'
HPP = 'ro:has_proper_part'
hpp = HPP.replace('ro:', graph.namespaces['ro'])
ppo = PPO.replace('ro:', graph.namespaces['ro'])
a, b = creatTree(
*Query(
tc.red('birnlex_796'), HPP, 'OUTGOING', 10
), # FIXME seems to be a last one wins bug here with birnlex_796 vs NIFGA:birnlex_796 depending on the has seed...
json=graph.make_scigraph_json(HPP))
c, d = creatTree(*Query('NIFGA:birnlex_796', hpp, 'OUTGOING', 10),
graph=sgg)
j = bridge.make_scigraph_json(
HPP) # issue https://github.com/RDFLib/rdflib/pull/661
e, f = creatTree(*Query('UBERON:0000955', HPP, 'OUTGOING', 10), json=j)
k_, l_ = creatTree(*Query('NIFGA:nlx_anat_101177', ppo, 'INCOMING', 10),
graph=sgg)
merge = dict(d[-1]) # full tree with ppo converted to hpp
merge['nodes'].extend(l_[-1]['nodes'])
merge['edges'].extend([{
'sub': e['obj'],
'pred': hpp,
'obj': e['sub']
} for e in l_[-1]['edges']])
m_, n_ = creatTree(*Query('NIFGA:birnlex_796', hpp, 'OUTGOING', 10),
json=merge)
print('nifga dep')
print(a)
print('nifga live')
print(c)
print('new bridge')
print(e)
print('nifga total (both directions)')
print(m_)
print('nifga white matter')
print(k_)
return a, b, c, d, e, f, k_, l_, m_, n_
def main():
from docopt import docopt, parse_defaults
args = docopt(__doc__, version='ontutils 0.0.1')
defaults = {
o.name: o.value if o.argcount else None
for o in parse_defaults(__doc__)
}
verbose = args['--verbose']
debug = args['--debug']
repo_name = args['<repo>']
git_local = os.path.expanduser(args['--git-local'])
epoch = args['--epoch']
curies_location = args['--curies']
curies = getCuries(curies_location)
curie_prefixes = set(curies.values())
filenames = args['<file>']
filenames.sort(key=lambda f: os.path.getsize(f),
reverse=True) # make sure the big boys go first
refactor_skip = ('nif.ttl', 'resources.ttl', 'generated/chebislim.ttl',
'unused/ro_bfo_bridge.ttl', 'generated/ncbigeneslim.ttl',
'generated/NIF-NIFSTD-mapping.ttl')
rfilenames = [f for f in filenames if f not in refactor_skip]
if args['set']:
from pyontutils.config import auth
uc = auth.user_config
def set_uc(var, value):
with open(uc._path, 'rt') as f:
text = f.read()
if '#' in text:
msg = f'Comments detected! Not writing config! {uc._path}'
raise ValueError(msg)
blob = uc.load()
# XXX NEVER DUMP A CONFIG THIS YOU _WILL_ KLOBBER IT
# BY ACCIDENT AT SOME POINT AND WILL ERASE ANY/ALL COMMENTS
# THERE IS NO SAFETY WITH THIS IMPLEMENTATION
# USERS SHOULD EDIT THEIR CONFIGS DIRECTLY
# except that it makes giving instructions for
# setting values a bit more complicated
blob['auth-variables'][var] = value
uc.dump(blob)
if args['ontology-local-repo']:
var = 'ontology-local-repo'
olr = Path(args['<path>']).expanduser().resolve()
olr_string = olr.as_posix()
set_uc(var, olr_string)
value2 = auth.get_path(var)
if not value2.exists():
msg = f'{var} path does not exist! {value2}'
print(tc.red('WARNING'), msg)
msg = f'{var} path {value2} written to {uc._path}'
print(msg)
assert olr == value2
elif args['scigraph-api-key']:
# FIXME this is a hack on top of orthauth, which will not
#
# check the secrets path first to make sure it is ok
# be implementing programmtic modification of user config
# files any time soon, though it might make sense to have a
# "machine config path" in addition to auth and user config
path = ['scigraph', 'api', 'key']
spath = auth._pathit(uc.get_blob('auth-stores', 'secrets')['path'])
if not spath.parent.exists():
spath.parent.mkdir(parents=True)
spath.parent.chmod(0o0700)
if spath.suffix != '.yaml':
msg = f"Can't write secrets file of type {spath.suffix}"
args = None
raise NotImplementedError(msg)
v = None
try:
s = uc.secrets
v = s(*path)
except:
pass
if v is not None:
v = None
raise ValueError(f'Path already in secrets! {path} in {spath}')
# safely append to the secrets file
key = args['<key>']
path_key = f'\nscigraph:\n api:\n key: {key}'
if not spath.exists():
spath.touch()
spath.chmod(0o0600)
with open(spath, 'a+') as f:
f.write(path_key)
# set the config var
var = 'scigraph-api-key'
value = {'path': ' '.join(path)}
set_uc(var, value) # set the path
# XXX NOTE yes, it is correct to do this only after secrets succeeds
# otherwise it is possible to get into a state where secrets does
# not exist but there is a path pointing to it, so load this
# ontutils file will fail during import time
# test that we got the value we expected
value2 = auth.get(var)
msg = (f'Key written to secrets. {spath} and path to '
f'key was written to config {uc._path}')
print(msg)
assert key == value2, 'Key retrieved does not match key set!'
elif args['devconfig']:
if args['--write']:
file = devconfig.write(args['--output-file'])
print(f'config written to {file}')
elif args['<field>']:
for f in args['<field>']:
print(getattr(devconfig, f, ''))
else:
print(devconfig)
elif args['catalog-extras']:
catalog_extras(args['--fetch'])
elif args['version-iri']:
version_iris(*filenames, epoch=epoch)
elif args['scigraph-stress']:
scigraph_stress(int(args['--rate']), int(args['--timeout']), verbose,
debug)
elif args['deadlinks']:
deadlinks(filenames, int(args['--rate']), int(args['--timeout']),
verbose, debug)
elif args['spell']:
spell(filenames, debug)
elif args['iri-commit']:
make_git_commit_command(git_local, repo_name)
elif args['uri-switch']:
uri_switch(rfilenames, uri_switch_values)
elif args['backend-refactor']:
backend_refactor(rfilenames, backend_refactor_values)
elif args['todo']:
graph = loadall(git_local, repo_name, local=True)
graph_todo(graph, curie_prefixes, uri_switch_values)
breakpoint()
elif args['expand']:
curies['NLXWIKI'] = 'http://legacy.neurolex.org/wiki/'
for curie in args['<curie>']:
prefix, suffix = curie.split(':')
print(curies[prefix] + suffix)
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
def main():
for filename in ('mbaslim', 'hbaslim', 'paxinos-rat-labels',
'waxholm-rat-labels'):
filepath = gitf / 'NIF-Ontology/ttl/generated/parcellation' / (
filename + '.ttl')
dir_ = filepath.parent.as_posix()
print(dir_)
file_commit = subprocess.check_output(
[
'git', 'log', '-n', '1', '--pretty=format:%H', '--',
filepath.name
],
cwd=dir_,
stderr=subprocess.DEVNULL).decode().rstrip()
graph = rdflib.Graph().parse(filepath.as_posix(), format='ttl')
g = makeGraph('', graph=graph)
annos = defaultdict(set)
anno_trips = defaultdict(set)
for triple, predicate_objects in annotation.parse(graph=graph):
for a_p, a_o in predicate_objects:
annos[a_p, a_o].add(triple)
anno_trips[triple].add((a_p, a_o))
anno_trips = {k: v for k, v in anno_trips.items()}
for lifted_triple in restriction.parse(graph=graph):
graph.add(lifted_triple)
out_header = 'label|abbrev|curie|superPart curie\n'
out = []
editions_header = 'edition|label|abbrev|curie\n'
editions = []
for s in graph.subjects(rdf.type, owl.Class):
rdfsLabel = next(graph.objects(s, rdfs.label))
try:
prefLabel = next(graph.objects(s, skos.prefLabel))
except StopIteration:
print(tc.red('WARNING:'),
f'skipping {s} {rdfsLabel} since it has no prefLabel')
continue
syns = sorted(
graph.objects(s, NIFRID.synonym)
) # TODO are there cases where we need to recaptulate what we are doing for for abbrevs?
abbrevs = sorted(graph.objects(
s, NIFRID.abbrev)) # FIXME paxinos has more than one
try:
if annos:
if len(abbrevs) > 1:
print(tc.blue('INFO:'), g.qname(s),
repr(prefLabel.value), 'has multiple abbrevs',
[a.value for a in abbrevs])
# prefer latest
current_edition = ''
for a in abbrevs:
for a_p, edition in anno_trips[s, NIFRID.abbrev, a]:
if a_p == ilxtr.literalUsedBy:
if current_edition < edition:
current_edition = edition
abbrev = a
else:
abbrev = abbrevs[0]
except IndexError:
abbrev = ''
try:
superPart = next(graph.objects(s, ilxtr.labelPartOf))
except StopIteration:
superPart = ''
out.append(
f'{prefLabel}|{abbrev}|{g.qname(s)}|{g.qname(superPart)}')
if annos:
#asdf = {'ed':{'label':,'abbrev':,'curie':}}
asdf = defaultdict(dict)
triple = s, skos.prefLabel, prefLabel
eds = anno_trips[triple]
for a_p, a_o in eds:
asdf[a_o]['curie'] = g.qname(s)
asdf[a_o]['label'] = prefLabel
for syn in graph.objects(s, NIFRID.synonym):
triple = s, NIFRID.synonym, syn
eds = anno_trips[triple]
for a_p, a_o in eds:
asdf[a_o]['curie'] = g.qname(s)
if 'label' in asdf[a_o]:
print(
tc.red('WARNING:'),
f'{a_o} already has a label "{asdf[a_o]["label"]}" for "{syn}"'
)
asdf[a_o]['label'] = syn
for abbrev in graph.objects(s, NIFRID.abbrev):
triple = s, NIFRID.abbrev, abbrev
eds = anno_trips[triple]
#print('aaaaaaaaaaa', g.qname(s), )
for a_p, a_o in eds:
asdf[a_o]['curie'] = g.qname(s)
if 'abbrev' in asdf[a_o]:
print(
tc.red('WARNING:'),
f'{a_o} already has a abbrev "{asdf[a_o]["abbrev"]}" for "{abbrev}"'
)
asdf[a_o]['abbrev'] = abbrev
#print(asdf)
for ed, kwargs in sorted(asdf.items()):
if 'abbrev' not in kwargs:
print('Skipping', ed, 'for\n', kwargs)
continue
editions.append('{ed}|{label}|{abbrev}|{curie}'.format(
ed=g.qname(ed), **kwargs))
with open('/tmp/' + filename + f'-{file_commit[:8]}.psv', 'wt') as f:
f.write(out_header + '\n'.join(sorted(out, key=labelkey)))
if editions:
with open('/tmp/' + filename + f'-editions-{file_commit[:8]}.psv',
'wt') as f:
f.write(editions_header +
'\n'.join(sorted(editions, key=edkey)))
def printer(*args, **kwargs):
printe(*(tc.red(repr(a)) for a in args), **kwargs)
def would_you_like_to_know_more_question_mark():
# resolving differences between classes
more_ids = set((
'http://uri.neuinfo.org/nif/nifstd/readable/ChEBIid',
'http://uri.neuinfo.org/nif/nifstd/readable/GOid',
'http://uri.neuinfo.org/nif/nifstd/readable/MeshUid',
'http://uri.neuinfo.org/nif/nifstd/readable/PMID',
'http://uri.neuinfo.org/nif/nifstd/readable/UmlsCui',
'http://uri.neuinfo.org/nif/nifstd/readable/bamsID',
'http://uri.neuinfo.org/nif/nifstd/readable/bonfireID',
'http://uri.neuinfo.org/nif/nifstd/readable/cell_ontology_ID',
'http://uri.neuinfo.org/nif/nifstd/readable/definingCitationID',
'http://uri.neuinfo.org/nif/nifstd/readable/definingCitationURI',
'http://uri.neuinfo.org/nif/nifstd/readable/emapMouseStageDataID',
'http://uri.neuinfo.org/nif/nifstd/readable/emapMouseStageDiagramID',
'http://uri.neuinfo.org/nif/nifstd/readable/externalSourceId',
'http://uri.neuinfo.org/nif/nifstd/readable/externalSourceURI',
'http://uri.neuinfo.org/nif/nifstd/readable/gbifID',
'http://uri.neuinfo.org/nif/nifstd/readable/gbifTaxonKeyID',
'http://uri.neuinfo.org/nif/nifstd/readable/gene_Ontology_ID',
#'http://uri.neuinfo.org/nif/nifstd/readable/hasExternalSource',
'http://uri.neuinfo.org/nif/nifstd/readable/hasGenbankAccessionNumber',
'http://uri.neuinfo.org/nif/nifstd/readable/imsrStandardStrainName',
'http://uri.neuinfo.org/nif/nifstd/readable/isReplacedByClass',
'http://uri.neuinfo.org/nif/nifstd/readable/jaxMiceID',
'http://uri.neuinfo.org/nif/nifstd/readable/ncbiTaxID',
'http://uri.neuinfo.org/nif/nifstd/readable/neuronamesID',
'http://uri.neuinfo.org/nif/nifstd/readable/nifID',
'http://uri.neuinfo.org/nif/nifstd/readable/sao_ID',
'http://uri.neuinfo.org/nif/nifstd/readable/umls_ID',
'http://www.geneontology.org/formats/oboInOwl#id',
))
outside = []
eee = {}
resolver_not_ilx_only_but_not_in_scigraph = set() # resources.ttl
_res = Graph().parse((gitf / 'NIF-Ontology/ttl/resources.ttl').as_posix(), format='turtle')
reslookup = {uri:[l] for uri, l in _res.subject_objects(rdfs.label)}
for uri in chain(h_uris, resolver_not_ilx_only):
if 'uri.neuinfo.org' in uri:
try:
meta = sgg.getNode(uri.toPython())['nodes'][0]['meta']
asdf = {hng.qname(k):v for k, v in meta.items() if k in more_ids}
except TypeError:
resolver_not_ilx_only_but_not_in_scigraph.add(uri) # resources.ttl ;)
if uri in reslookup: # no differentia
asdf = False
else:
asdf = False
print('WTF', uri)
if asdf:
#print(uri, asdf)
eee[uri] = asdf
for l in asdf.values():
for e in l:
outside.append(e)
outside_dupes = [v for v, c in Counter(outside).most_common() if c > 1]
eee_dupes = {k:v for k, v in eee.items() if anyMembers(outside_dupes, *(e for l in v.values() for e in l))}
#for uri, meta in sorted(eee_dupes.items(), key=lambda a:sorted(a[1].values())):
#print(uri.toPython(), sorted((e.replace('PMID: ', 'PMID:'), k) for k, l in meta.items() for e in l))
# attempt to deal with label mappings
iexisting = defaultdict(set)
iiexisting = {}
for i, existing in zip(datal('ilx'), datal('iri')):
#if 'uri.neuinfo.org' in existing:
if 'interlex.org' not in existing and 'neurolex.org' not in existing:
iexisting[i].add(URIRef(existing))
iiexisting[URIRef(existing)] = i
iexisting = {**iexisting}
_ilabs = {k:l for k, l in zip(datal('ilx'), datal('label'))}
def inner(iri):
resp = sgv.findById(iri)
if resp is not None:
l = resp['labels']
else:
l = [] #_ilabs[iiexisting[iri]] + '** already in ilx **']
#print('trouble?', iri) # ilx only
return iri, l
#labs = {k:v[0] if v else '<--NO-LABEL-->' for k, v in Async()(deferred(inner)(id_) for id_ in chain(h_uris, (e for s in iexisting.values() for e in s)))}
labs = {k:v[0] if v else '<--NO-LABEL-->' for k, v in Async()(deferred(inner)(id_) for id_ in h_uris)}
ilabs = {k:l.lower() for k, l in zip(datal('ilx'), datal('label'))}
iilabs = {v:k for k, v in ilabs.items()}
assert len(ilabs) == len(iilabs)
missing_map = {k:iilabs[v.lower()] for k, v in labs.items() if v and v.lower() in iilabs} # XXX this is not valid
missing_existing = {i:[m, *iexisting[i]] for m, i in missing_map.items() if i in iexisting}
missing_equivs = {next(iter(iexisting[i])):i for m, i in missing_map.items() if i in iexisting}
eid = NIFRID.externalSourceId.toPython()
ded = owl.deprecated.toPython()
# SP: -> swissprot vs uniprot
mmr = []
proto_mmr_1_to_1 = {}
arrr = defaultdict(set)
uniprot_iuphar = set()
for uri, ilx_frag in {**missing_equivs, **missing_map}.items():
uri = URIRef(uri)
try:
meta = sgg.getNode(uri.toPython())['nodes'][0]['meta']
except TypeError:
# just ignore these, they are ilx only :/
meta = {}
if eid in meta:
src = meta[eid][0]
if src.startswith('SP:'):
src = tc.yellow(src.replace('SP:', 'http://www.uniprot.org/uniprot/'))
#elif src.startswith('IUPHAR:'):
#pass
#else:
#src = 'TODO'
elif ded in meta and meta[ded]:
src = tc.red('ded ')
else:
src = 'TODO'
val = labs[uri] if uri in labs else _ilabs[ilx_frag] + ' **'
if uri in eee:
differentia = str(eee[uri])
for v in eee[uri].values():
for e in v:
arrr[e].add(uri)
if 'SP:' in e or 'IUPHAR:' in e:
uniprot_iuphar.add(uri)
else:
differentia = ''
if uri in _ilx and uri in all_uris:
ruri = SGG[hng.qname(uri)]
ruri = tc.blue(f'{ruri:<60}')
else:
ruri = uri
ruri = f'{ruri:<60}'
v = ' '.join((f'{val:<60}',
src,
ruri,
ilxb[ilx_frag],
differentia))
mmr.append(v)
proto_mmr_1_to_1[uri] = v
src = None
arrr = {**arrr}
arrr_not_1_to_1 = {k:v for k, v in arrr.items() if len(v) > 1}
#arrr_n11_uris = set((u.toPython() for v in arrr_not_1_to_1.values() for u in v))
arrr_n11_uris = set.union(*arrr_not_1_to_1.values())
mmr_1_to_1 = {k:v for k, v in proto_mmr_1_to_1.items() if k not in arrr_n11_uris}
no_uniprot = {k:v for k, v in proto_mmr_1_to_1.items() if k not in uniprot_iuphar}
arrr_n11_text = '\n'.join(f'{k:<15} {sorted(_.toPython() for _ in v)}' for k, v in arrr_not_1_to_1.items())
mmr.sort()
mmr_text = '\n'.join(mmr)
mmr_1_to_1_text = '\n'.join(sorted(mmr_1_to_1.values()))
no_uniprot_text = '\n'.join(sorted(no_uniprot.values()))
def as_pretty_diff(self,
pathmeta,
othermeta,
pathobject=None,
title=None,
human=False):
if title is None:
if pathobject is not None:
title = pathobject.relative_to(pathobject.cwd()).as_posix()
else:
title = ''
if pathmeta.content_different(othermeta):
title = tc.red(title)
def merge(a, b):
keys = [k for k, _ in a]
okeys = [ok for ok, _ in b]
kind = [(k, okeys.index(k) - i, k) if k in okeys else (None, 0, k)
for i, k in enumerate(keys)]
okind = [(ok, keys.index(ok) - i, ok) if ok in keys else
(None, 0, ok) for i, ok in enumerate(okeys)]
long, short = (kind, okind) if len(kind) > len(okind) else (okind,
kind)
start_align = None
aligned = []
offset = 0
for i, (k, v, av) in enumerate(long):
i -= offset
if k is None and start_align is None:
aligned.append(av)
aligned.append(short[i + v][-1])
elif k is not None:
if v <= 0:
start_align = i
aligned.append(av)
else:
for j in range(v):
aligned.append(short[i + j][-1])
offset += v
aligned.append(av)
elif k is None:
aligned.append(av)
# FIXME what to do if short has a key that long does not?
lkv = len(a[0])
lokv = len(b[0])
kv = {k: [k, v] for k, v in a}
okv = {k: [k, v] for k, v in b}
for k in aligned:
l = kv[k] if k in kv else ([''] * lkv)
r = okv[k] if k in okv else ([''] * lokv)
yield l + r
h = [['Key', 'Value', 'Key Other', 'Value Other']]
rows = h + list(
merge(self.rows(pathmeta, human=human),
self.rows(othermeta, human=human)))
try:
table = AsciiTable(rows, title=title).table
except TypeError as e:
breakpoint()
raise e
return table