def getCreator(self, creator):
"""
Find the painter with the name in creator
First check if the name is already in the self.creators cache
Second, do a search
If a hit is found, update the cache in self.creators
"""
# First use the cache
if creator in self.creators:
return self.creators[creator]
# Search Wikidata for a suitable candidate, tell the search to only return humans
searchstring = u'%s haswbstatement:P31=Q5' % (creator, )
creategen = pagegenerators.PreloadingItemGenerator(
pagegenerators.WikibaseItemGenerator(
pagegenerators.SearchPageGenerator(searchstring,
step=None,
total=50,
namespaces=[0],
site=self.repo)))
for creatoritem in creategen:
if creatoritem.isRedirectPage():
creatoritem = creatoritem.getRedirectTarget()
# See if the label or one of the aliases of the creatoritem matches the string we have. Only label is case insensitive.
if (creatoritem.get().get('labels').get('en')
and creatoritem.get().get('labels').get('en').lower()
== creator.lower()) or (
creatoritem.get().get('aliases').get('en') and creator
in creatoritem.get().get('aliases').get('en')):
if u'P106' in creatoritem.get().get('claims'):
existing_claims = creatoritem.get().get('claims').get(
'P106')
for existing_claim in existing_claims:
if existing_claim.target_equals(u'Q1028181'):
self.creators[creator] = creatoritem
return creatoritem
# Regex that should match all the anonymous work stuff that isn't covered by the list
anonymousRegex = '^(Workshop of|Follower of|Circle of|Manner of|Forgery after|School of|After|Unidentified Artist|School of)\s.*$'
anonymousMatch = re.match(anonymousRegex, creator, flags=re.I)
if anonymousMatch:
self.creators[creator] = self.creators.get('anonymous')
return self.creators.get('anonymous')
# We don't want to do the same search over and over again
self.creators[creator] = None
return None
def subgenerator(self):
limit = self.getOption('limit')
for ident in self.item_ids:
from_item = pywikibot.ItemPage(self.repo, ident)
for item in pagegenerators.WikibaseItemGenerator(
from_item.backlinks(total=limit,
filterRedirects=False,
namespaces=[0])):
yield item
if limit is not None:
limit -= 1
if limit == 0:
return
for prop in self.url_props:
ok = True
while ok and limit != 0:
ok = False
query = self.store.build_query(
'fake_references_url',
limit=500 if limit is None else min(500, limit),
prop=prop)
for item in pagegenerators.WikidataSPARQLPageGenerator(
query, site=self.repo):
ok = True
yield item
if limit is not None:
limit -= 1
for prop in self.ref_props:
ok = True
while ok and limit != 0:
ok = False
query = self.store.build_query(
'fake_references',
limit=100 if limit is None else min(100, limit),
prop=prop)
for item in pagegenerators.WikidataSPARQLPageGenerator(
query, site=self.repo):
ok = True
yield item
if limit is not None:
limit -= 1
def findPerson(self, name, yob, yod):
"""
Find a person.
:param name: Name of the person
:param yob: Year of birth of the person
:param yod: Year of death of the person
:return: ItemPage if a person is found
"""
# Search Wikidata for a suitable candidate, tell the search to only return humans
searchstring = u'"%s" haswbstatement:P31=Q5' % (name, )
persongen = pagegenerators.PreloadingItemGenerator(
pagegenerators.WikibaseItemGenerator(
pagegenerators.SearchPageGenerator(searchstring,
step=None,
total=50,
namespaces=[0],
site=self.repo)))
foundperson = False
for personitem in persongen:
#print (u'Possible match %s' % (personitem.title(),))
if personitem.isRedirectPage():
personitem = personitem.getRedirectTarget()
# See if the label or one of the aliases of the creatoritem matches the string we have. Only label is case insensitive.
#if (personitem.get().get('labels').get('en') and personitem.get().get('labels').get('en').lower() == name.lower()) or (personitem.get().get('aliases').get('en') and name in personitem.get().get('aliases').get('en')):
# print (u'Label match for %s' % (personitem.title(),))
# # Check of year of birth and year of death match
if u'P569' in personitem.get().get(
'claims') and u'P570' in personitem.get().get('claims'):
dob = personitem.get().get('claims').get('P569')[0].getTarget()
dod = personitem.get().get('claims').get('P570')[0].getTarget()
foundperson = True
if dob and dod:
#print (u'Date found dob "%s" "%s" "%s"' % (dob, dob.year, yob))
#print (u'Date found dod "%s" "%s" "%s"' % (dod, dod.year, yod))
if int(dob.year) == int(yob) and int(dod.year) == int(yod):
#print (u'maaaaaaaaaaaaaaaaaaaaaaaatcchhhhh')
return personitem
return foundperson
def match_name_off_labs(name, types, wd, limit):
"""
Check if there is an item matching the name using API search.
Less good than match_name_on_labs() but works from anywhere.
@param name: The name to search for
@type name: basestring
@param types: The Q-values which are allowed for INSTANCE_OF_P
@type types: tuple of basestring
@param wd: The running WikidataStuff instance
@type wd: WikidataStuff (WD)
@return: Any matching items
@rtype: list (of pywikibot.ItemPage)
"""
matches = []
objgen = pagegenerators.PreloadingItemGenerator(
pagegenerators.WikibaseItemGenerator(
pagegenerators.SearchPageGenerator(name,
step=None,
total=10,
namespaces=[0],
site=wd.repo)))
# check if P31 and then if any of prop[typ] in P31
i = 0
for obj in objgen:
obj = wd.bypassRedirect(obj)
i += 1
if i > limit:
# better to skip than to crash when search times out
# remove any matches (since incomplete) and exit loop
return [] # avoids keeping a partial list
if (name in obj.get().get('labels').values()
or name in obj.get().get('aliases').values()):
filter_on_types(obj, types, matches)
return matches