def test_version_f1000(self):
# this doesn't exist
#d = idlib.Doi('https://doi.org/10.12688/f1000research.6555')
#d = d.identifier_bound_version_metadata
d1 = idlib.Doi('https://doi.org/10.12688/f1000research.6555.1')
d1.identifier_bound_version_metadata
# of course there is no linke between the two >_<
d2 = idlib.Doi('https://doi.org/10.12688/f1000research.6555.2')
d2.identifier_bound_version_metadata
def doi(doi_string): # FIXME massive network sandbox violation here
""" check if a doi string resolves, if it does, return it """
if doi_string is None:
raise TypeError('WHAT HAVE YOU DONE!?')
doi = idlib.Doi(doi_string)
try:
metadata = doi.metadata() # FIXME network sandbox violation
if metadata is not None:
return doi
except idlib.exceptions.RemoteError:
# sometimes a doi is present on the platform but does not resolve
# in which case we don't add it as metadata because it has not
# been officially published, just reserved, this check is more
# correct than checkin the status on the platform
# FIXME HOWEVER it violates the network sandbox, so we probably
# need an extra step during the data retrieval phase which attempts
# to fetch all the doi metadata
pass
except Exception as e:
# XXX random errors need to be ignored here for now
# since this really should not be run at this step
# due to the network dependency, we need a post-network
# step where we can strip out all the things that fail
log.exception(e)
def triples(self):
crossref_doi_pred = rdflib.term.URIRef('http://prismstandard.org/namespaces/basic/2.1/doi')
for blob in self.data['identifier_metadata']:
id = blob['id']
if not isinstance(id, idlib.Stream):
id = idlib.Auto(id)
if not hasattr(id, 'asUri'):
breakpoint()
s = id.asUri(rdflib.URIRef)
if 'source' in blob:
source = blob['source'] # FIXME we need to wrap this in our normalized representation
if source == 'Crossref': # FIXME CrossrefConvertor etc. OR put it in idlib as a an alternate ttl
pos = (
(rdf.type, owl.NamedIndividual),
(rdf.type, TEMP[blob['type']]),
(dc.publisher, blob['publisher']),
#(dc.type, blob['type']), # FIXME semantify
(dc.title, blob['title']),
(dc.date, self.published_online(blob)), # FIXME .... dangerzone
)
g = OntGraph()
doi = idlib.Doi(id) if not isinstance(id, idlib.Doi) else id # FIXME idlib streams need to recognize their own type in __new__
data = doi.ttl()
if data is None: # blackfynn has some bad settings on their doi records ...
return
try:
g.parse(data=data, format='ttl') # FIXME network bad
except BaseException as e:
loge.exception(e)
_tr = [s for s, p, o in g if p == crossref_doi_pred]
if _tr:
_their_record_s = _tr[0]
yield s, owl.sameAs, _their_record_s
yield from g
else:
g.debug()
log.critical('No crossref doi section in graph!')
else:
msg = f'dont know what to do with {source}'
log.error(msg)
#raise NotImplementedError(msg)
return
else:
msg = f'dont know what to do with {blob} for {id.identifier}'
log.error(msg)
#raise NotImplementedError(msg)
return
for p, oraw in pos:
if oraw is not None:
o = rdflib.Literal(oraw) if not isinstance(oraw, rdflib.URIRef) else oraw
yield s, p, o
def originating_article_doi(self, value):
self._error_on_na(value)
#self._error_on_tbd(value) # TODO?
value, _j = self._deatag(value)
for val in value.split(','):
v = val.strip()
if v:
try:
yield idlib.Doi(v)
except idlib.exceptions.MalformedIdentifierError as e:
logd.exception(e)
def _protocol_url_or_doi(self, value):
doi = False
if 'doi' in value:
doi = True
elif value.startswith('10.'):
value = 'doi:' + value
doi = True
if doi:
value = idlib.Doi(value)
else:
value = idlib.Pio(value)
return value
def _protocol_url_or_doi(self, value):
doi = False
if 'doi' in value:
doi = True
elif value.startswith('10.'):
value = 'doi:' + value
doi = True
if doi:
value = idlib.Doi(value) # XXX possible encapsulation issue
else:
value = idlib.Pio(value) # XXX possible encapsulation issue
return value
def triples(self):
for blob in self.data['identifier_metadata']:
id = blob['id']
if not isinstance(id, idlib.Stream):
id = idlib.Auto(id)
s = id.asType(rdflib.URIRef)
if 'source' in blob:
source = blob[
'source'] # FIXME we need to wrap this in our normalized representation
if source == 'Crossref': # FIXME CrossrefConvertor etc. OR put it in idlib as a an alternate ttl
pos = (
(rdf.type, owl.NamedIndividual),
(rdf.type, TEMP[blob['type']]),
(dc.publisher, blob['publisher']),
#(dc.type, blob['type']), # FIXME semantify
(dc.title, blob['title']),
(dc.date,
self.published_online(blob)), # FIXME .... dangerzone
)
g = OntGraph()
doi = idlib.Doi(id) if not isinstance(
id, idlib.Doi
) else id # FIXME idlib streams need to recognize their own type in __new__
g.parse(data=doi.ttl(), format='ttl') # FIXME network bad
_their_record_s = [
s for s, p, o in g if p == rdflib.term.URIRef(
'http://prismstandard.org/namespaces/basic/2.1/doi'
)
][0]
yield s, owl.sameAs, _their_record_s
yield from g
else:
msg = f'dont know what to do with {source}'
log.error(msg)
#raise NotImplementedError(msg)
return
else:
msg = f'dont know what to do with {blob} for {id.identifier}'
log.error(msg)
#raise NotImplementedError(msg)
return
for p, oraw in pos:
if oraw is not None:
o = rdflib.Literal(oraw) if not isinstance(
oraw, rdflib.URIRef) else oraw
yield s, p, o
def doi(doi_string): # FIXME massive network sandbox violation here
""" check if a doi string resolves, if it does, return it """
doi = idlib.Doi(doi_string)
try:
metadata = doi.metadata() # FIXME network sandbox violation
if metadata is not None:
return doi
except idlib.exceptions.RemoteError:
# sometimes a doi is present on the platform but does not resolve
# in which case we don't add it as metadata because it has not
# been officially published, just reserved, this check is more
# correct than checkin the status on the platform
# FIXME HOWEVER it violates the network sandbox, so we probably
# need an extra step during the data retrieval phase which attempts
# to fetch all the doi metadata
pass
def mkval(cell):
hl = cell.hyperlink
cv = cell.value
if hl is None:
hl = cv if cv else None
if hl is not None:
try:
return idlib.Pio(hl)
except idlib.exc.IdlibError as e:
try:
return idlib.Doi(hl)
except idlib.exc.IdlibError as e:
pass
logd.warning(f'unhandled value {cell.value}')
return cv
def __new__(cls, something):
if '10.' in something:
if 'http' in something and 'doi.org' not in something:
pass # probably a publisher uri that uses the handle
else:
return idlib.Doi(something)
if 'orcid' in something:
return idlib.Orcid(something)
if '/ror.org/' in something or something.startswith('ror:'):
return idlib.Ror(something)
if 'protocols.io' in something:
return idlib.Pio(something)
return oq.OntId(something)
return OntTerm(something) # use the better local version of OntTerm
def doi(self):
blob = self.bfobject.doi
print(blob)
if blob:
return idlib.Doi(blob['doi'])
def doi(self):
data = self.data()
if data:
doi = data['doi']
if doi:
return idlib.Doi(doi)
