def make_place(self, data: dict, base_uri=None):
'''
Given a dictionary representing data about a place, construct a model.Place object,
assign it as the crom data in the dictionary, and return the dictionary.
The dictionary keys used to construct the place object are:
- name
- type (one of: 'City', 'State', 'Province', or 'Country')
- part_of (a recursive place dictionary)
'''
unique_locations = self.unique_locations
TYPES = {
'city': vocab.instances['city'],
'province': vocab.instances['province'],
'state': vocab.instances['province'],
'country': vocab.instances['nation'],
}
if data is None:
return None
type_name = data.get('type', 'place').lower()
name = data.get('name')
label = name
parent_data = data.get('part_of')
place_type = TYPES.get(type_name)
parent = None
if parent_data:
parent_data = self.make_place(parent_data, base_uri=base_uri)
parent = get_crom_object(parent_data)
if label:
label = f'{label}, {parent._label}'
placeargs = {}
if label:
placeargs['label'] = label
if data.get('uri'):
placeargs['ident'] = data['uri']
elif label in unique_locations:
data['uri'] = self.make_proj_uri('PLACE', label)
placeargs['ident'] = data['uri']
elif base_uri:
data['uri'] = base_uri + urllib.parse.quote(label)
placeargs['ident'] = data['uri']
p = model.Place(**placeargs)
if place_type:
p.classified_as = place_type
if name:
p.identified_by = model.Name(ident='', content=name)
else:
warnings.warn(f'Place with missing name on {p.id}')
if parent:
p.part_of = parent
data['part_of'] = parent_data
return add_crom_data(data=data, what=p)
def nation(self, name, ident):
place = vocab.Place(
ident=
'tag:getty.edu,2019:digital:pipeline:provenance:REPLACE-WITH-UUID#PLACE-COUNTRY-'
+ name,
label=name)
nation = model.Place(ident=ident)
place.classified_as = vocab.instances['nation']
place.identified_by = model.Name(ident='', content=name)
return place
def set_properties(self, data, thing):
super().set_properties(data, thing)
with suppress(ValueError, TypeError):
ulan = int(data.get('ulan'))
if ulan:
thing.exact_match = model.BaseResource(
ident=f'http://vocab.getty.edu/ulan/{ulan}')
if 'name' in data:
title_type = model.Type(
ident='http://vocab.getty.edu/aat/300417193', label='Title')
name = data['name']
if name:
if isinstance(name, str):
set_la_name(thing, name, title_type, set_label=True)
elif isinstance(name, (list, tuple)):
value, *properties = name
n = model.Name(ident='', content=value)
n.classified_as = title_type
self.set_lo_properties(n, *properties)
thing.identified_by = n
for uri in data.get('exact_match', []):
thing.exact_match = uri
for sdata in data.get('sojourns', []):
label = sdata.get('label', 'Sojourn activity')
stype = sdata.get('type', model.Activity)
act = stype(ident='', label=label)
ts = get_crom_object(sdata.get('timespan'))
place = get_crom_object(sdata.get('place'))
act.timespan = ts
act.took_place_at = place
thing.carried_out = act
self.set_referred_to_by(sdata, act)
# Locations are names of residence places (P74 -> E53)
# XXX FIXME: Places are their own model
if 'places' in data:
for p in data['places']:
if isinstance(p, model.Place):
pl = p
elif isinstance(p, dict):
pl = get_crom_object(p)
else:
pl = model.Place(ident='', label=p)
#pl._label = p['label']
#nm = model.Name()
#nm.content = p['label']
#pl.identified_by = nm
#for s in p['sources']:
# l = model.LinguisticObject(ident="urn:uuid:%s" % s[1])
# l._label = _row_label(s[2], s[3], s[4])
# pl.referred_to_by = l
thing.residence = pl
def model_publisher_group(self, record, data, seq):
record.setdefault('_publishings', [])
series_label = record['label']
corp_id = data.get('gaia_corp_id')
geog_id = data.get('gaia_geog_id')
a_uri = record['uri'] + f'-pub-{seq}'
cb_label = f' by CB{corp_id}' if corp_id else f' by publisher #{seq}'
a = vocab.Publishing(ident=a_uri,
label=f'Publishing of {series_label}' + cb_label)
if corp_id:
uri = self.helper.corporate_body_uri(corp_id)
a.carried_out_by = model.Group(ident=uri)
if geog_id:
uri = self.helper.place_uri(geog_id)
a.took_place_at = model.Place(ident=uri)
record['_publishings'].append(a)
def make_place(self, data: dict, base_uri=None):
'''
Given a dictionary representing data about a place, construct a model.Place object,
assign it as the crom data in the dictionary, and return the dictionary.
The dictionary keys used to construct the place object are:
- name
- type (one of: 'City', 'State', 'Province', 'Country', or 'Sovereign')
- part_of (a recursive place dictionary)
If the name matches a known unique location (derived from the unique_locations
service data), the normal recursive handling of part_of data is bypassed, using
the
'''
# unique_locations = self.unique_locations
canonical_location_names = self.canonical_location_names
TYPES = {
'city': vocab.instances['city'],
'county': vocab.instances['county'],
'province': vocab.instances['province'],
'state': vocab.instances['province'],
'country': vocab.instances['nation'],
'sovereign': vocab.instances['sovereign'],
}
if data is None:
return None
type_name = data.get('type', 'place').lower()
name = data.get('name')
si = self.static_instances
names = data.get('names', [])
label = name
parent_data = data.get('part_of')
place_type = TYPES.get(type_name)
parent = None
if name.casefold() in canonical_location_names:
name = canonical_location_names.get(name.casefold(), name)
label = name
elif parent_data:
parent_data = self.make_place(parent_data, base_uri=base_uri)
parent = get_crom_object(parent_data)
if label:
label = f'{label}, {parent._label}'
placeargs = {}
p = None
if si:
p = si.get_instance('Place', name)
if not p:
p = si.get_instance('Place', label)
if p:
# this is a static instance. we need to re-thread the part_of relationship
# in the data dictionary, because the serialization depends on the dictionary
# data, not the properties of the modeled object
# from cromulent.model import factory
# print(f'PLACE: {name} => {factory.toString(p, False)}')
add_crom_data(data=data, what=p)
queue = [data]
while queue:
place_data = queue.pop(0)
place = get_crom_object(place_data)
parents = getattr(place, 'part_of', []) or []
if parents:
for parent in parents:
if parent:
if 'part_of' not in place_data:
parent_data = add_crom_data(data={},
what=parent)
place_data['part_of'] = parent_data
else:
parent_data = add_crom_data(
data=place_data['part_of'],
what=parent)
queue.append(parent_data)
elif 'part_of' in place_data:
parent_data = self.make_place(place_data['part_of'],
base_uri=base_uri)
queue.append(parent_data)
if p:
return data
if label:
placeargs['label'] = label
if data.get('uri'):
placeargs['ident'] = data['uri']
# elif label.casefold() in canonical_location_names:
# label = canonical_location_names[label.casefold()]
# data['uri'] = self.make_shared_uri('PLACE', label)
# placeargs['ident'] = data['uri']
elif base_uri:
data['uri'] = base_uri + urllib.parse.quote(label)
placeargs['ident'] = data['uri']
if not p:
p = model.Place(**placeargs)
if place_type:
p.classified_as = place_type
if name:
p.identified_by = vocab.PrimaryName(ident='', content=name)
else:
warnings.warn(f'Place with missing name on {p.id}')
for name in names:
if name:
p.identified_by = model.Name(ident='', content=name)
if parent:
p.part_of = parent
data['part_of'] = parent_data
return add_crom_data(data=data, what=p)
def model_imprint_group(self, record, data):
if not data:
return
record.setdefault('referred_to_by', [])
record.setdefault('used_for', [])
record.setdefault('part_of', [])
record.setdefault('_activities', [])
record.setdefault('_groups', [])
record.setdefault('_places', [])
record.setdefault('identifiers', [])
edition = data.get('edition')
series_number = data.get('series_number')
doi = data.get('doi')
coden = data.get('coden')
website = data.get('website_address')
publishers = _as_list(data.get('publisher'))
distributors = _as_list(data.get('distributor'))
journal = data.get('journal_info')
# imprint_group/journal_info/aata_journal_id
# imprint_group/journal_info/aata_issue_id
degree = data.get('thesis_degree')
tr = data.get('technical_report_number')
if edition:
record['referred_to_by'].append(
vocab.EditionStatement(ident='', content=edition))
if series_number:
record['referred_to_by'].append(
vocab.Note(ident='',
content=series_number)) # TODO: classify this Note
if doi:
record['identifiers'].append(
vocab.DoiIdentifier(ident='', content=doi))
if coden:
record['identifiers'].append(
vocab.CodenIdentifier(ident='', content=coden))
if website:
record['referred_to_by'].append(
vocab.Note(ident='', content=website))
article_label = record['label']
for i, publisher in enumerate(publishers):
corp_id = publisher.get('gaia_corp_id')
geog_id = publisher.get('publisher_location',
{}).get('gaia_geog_id')
a_uri = record['uri'] + f'-pub-{i}'
a = vocab.Publishing(ident=a_uri,
label=f'Publishing of {article_label}')
if corp_id:
uri = self.helper.corporate_body_uri(corp_id)
g = model.Group(ident=uri)
a.carried_out_by = g
record['_groups'].append(add_crom_data({}, g))
if geog_id:
uri = self.helper.place_uri(geog_id)
p = model.Place(ident=uri)
a.took_place_at = p
record['_places'].append(add_crom_data({}, p))
record['used_for'].append(a)
# record['_activities'].append(add_crom_data({}, a))
for i, distributor in enumerate(distributors):
corp_id = distributor.get('gaia_corp_id')
geog_id = distributor.get('distributor_location',
{}).get('gaia_geog_id')
a_uri = record['uri'] + f'-dist-{i}'
a = vocab.Distributing(ident=a_uri,
label=f'Distribution of {article_label}')
if corp_id:
uri = self.helper.corporate_body_uri(corp_id)
g = model.Group(ident=uri)
a.carried_out_by = g
record['_groups'].append(add_crom_data({}, g))
if geog_id:
uri = self.helper.place_uri(geog_id)
p = model.Place(ident=uri)
a.took_place_at = p
record['_places'].append(add_crom_data({}, p))
record['used_for'].append(a)
# record['_activities'].append(add_crom_data({}, a))
if journal:
journal_id = journal.get('aata_journal_id')
issue_id = journal.get('aata_issue_id')
issue_uri = self.helper.issue_uri(journal_id, issue_id)
issue = vocab.IssueText(ident=issue_uri)
record['part_of'].append(add_crom_data({'uri': issue_uri}, issue))
if degree:
record['referred_to_by'].append(
vocab.Note(ident='', content=degree))
if tr:
record['identifiers'].append(model.Identifier(
ident='', content=tr)) # TODO: classify this Identifier
def setup_static_instances(self):
'''
These are instances that are used statically in the code. For example, when we
provide attribution of an identifier to Getty, or use a Lugt number, we need to
serialize the related Group or Person record for that attribution, even if it does
not appear in the source data.
'''
lugt_ulan = 500321736
gri_ulan = 500115990
gci_ulan = 500115991
knoedler_ulan = 500304270
GETTY_PSCP_URI = self.helper.make_shared_uri(
'STATIC', 'ORGANIZATION',
'Project for the Study of Collecting and Provenance')
GETTY_GPI_URI = self.helper.make_shared_uri('STATIC', 'ORGANIZATION',
'Getty Provenance Index')
GETTY_GRI_URI = self.helper.make_proj_uri('ORGANIZATION',
'LOCATION-CODE', 'JPGM')
GETTY_GCI_URI = self.helper.make_shared_uri(
'STATIC', 'ORGANIZATION', 'Getty Conservation Institute')
LUGT_URI = self.helper.make_proj_uri('PERSON', 'ULAN', lugt_ulan)
KNOEDLER_URI = self.helper.make_shared_uri('ORGANIZATION', 'ULAN',
str(knoedler_ulan))
NEWYORK_URI = self.helper.make_shared_uri('PLACE', 'USA', 'NY',
'New York')
gci = model.Group(ident=GETTY_GCI_URI,
label='Getty Conservation Institute')
gci.identified_by = vocab.PrimaryName(
ident='', content='Getty Conservation Institute')
gci.exact_match = model.BaseResource(
ident=f'http://vocab.getty.edu/ulan/{gci_ulan}')
gri = model.Group(ident=GETTY_GRI_URI,
label='Getty Research Institute')
gri.identified_by = vocab.PrimaryName(
ident='', content='Getty Research Institute')
gri.exact_match = model.BaseResource(
ident=f'http://vocab.getty.edu/ulan/{gri_ulan}')
gpi = model.Group(ident=GETTY_GPI_URI, label='Getty Provenance Index')
gpi.identified_by = vocab.PrimaryName(ident='',
content='Getty Provenance Index')
pscp = model.Group(
ident=GETTY_PSCP_URI,
label='Project for the Study of Collecting and Provenance')
pscp.identified_by = vocab.PrimaryName(
ident='',
content='Project for the Study of Collecting and Provenance')
lugt = model.Person(ident=LUGT_URI, label='Frits Lugt')
lugt.identified_by = vocab.PrimaryName(ident='', content='Frits Lugt')
lugt.exact_match = model.BaseResource(
ident=f'http://vocab.getty.edu/ulan/{lugt_ulan}')
knoedler_name = 'M. Knoedler & Co.'
knoedler = model.Group(ident=KNOEDLER_URI, label=knoedler_name)
knoedler.identified_by = vocab.PrimaryName(ident='',
content=knoedler_name)
knoedler.exact_match = model.BaseResource(
ident=f'http://vocab.getty.edu/ulan/{knoedler_ulan}')
newyork_name = 'New York, NY'
newyork = model.Place(ident=NEWYORK_URI, label=newyork_name)
newyork.identified_by = vocab.PrimaryName(ident='',
content=newyork_name)
materials = {}
if 'materials' in self.services:
materials.update({
aat: model.Material(ident=f'http://vocab.getty.edu/aat/{aat}',
label=label)
for aat, label in self.services['materials'].items()
})
places = self._static_place_instances()
places.update({'newyork': newyork})
db_people = self.static_db_instance(
'PEOPLE', name='STAR Person Authority Database', creator=gpi)
db_knoedler = self.static_db_instance('Knoedler',
name='STAR Knoedler Database',
creator=gpi)
db_sales_events = self.static_db_instance(
'Sales',
'Descriptions',
name='STAR Sales Catalogue Database',
creator=gpi)
db_sales_catalogs = self.static_db_instance(
'Sales',
'Catalogue',
name='STAR Physical Sales Catalogue Database',
creator=gpi)
db_sales_contents = self.static_db_instance(
'Sales',
'Contents',
name='STAR Sales Contents Database',
creator=gpi)
instances = defaultdict(dict)
instances.update({
'LinguisticObject': {
'db-people': db_people,
'db-knoedler': db_knoedler,
'db-sales_events': db_sales_events,
'db-sales_catalogs': db_sales_catalogs,
'db-sales_contents': db_sales_contents,
},
'Group': {
'gci': gci,
'pscp': pscp,
'gri': gri,
'gpi': gpi,
'knoedler': knoedler
},
'Person': {
'lugt': lugt
},
'Material': materials,
'Place': places
})
return instances