def __call__(self, data: dict, non_auctions):
'''Add modeling for the entry describing a physical auction catalog in the PSCP dataset.'''
cno = data['catalog_number']
owner = data['owner_code']
copy = data['copy_number']
rec_num = data['star_record_no']
sale_type = non_auctions.get(cno,
data.get('non_auction_flag', 'Auction'))
keys = [v for v in [cno, owner, copy] if v]
record_uri = self.helper.make_proj_uri('ENTRY', 'PHYS-CAT', *keys)
content = data['star_csv_data']
catalog_label = self.helper.physical_catalog_label(
cno, sale_type, owner, copy)
row_name = f'STAR Entry for Physical {catalog_label}'
row = vocab.EntryTextForm(ident=record_uri,
content=content,
label=row_name)
row.part_of = self.helper.static_instances.get_instance(
'LinguisticObject', 'db-sales_catalogs')
creation = model.Creation(ident='')
creation.carried_out_by = self.helper.static_instances.get_instance(
'Group', 'gpi')
row.created_by = creation
row.identified_by = self.helper.gpi_number_id(rec_num,
vocab.StarNumber)
row.identified_by = vocab.PrimaryName(ident='', content=row_name)
data['_catalog_record'] = add_crom_data({'uri': record_uri}, row)
yield data
def write_obj1(self, id):
'''Writes a Person model object with a label and a PrimaryName'''
p1 = vocab.Person(ident=f'urn:{id}', label='Greg')
p1.identified_by = vocab.PrimaryName(content='Gregory Williams')
w = MergingFileWriter(directory=self.path, model='test-model')
d = self.obj_to_dict(p1, id)
self.writer(d)
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
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)
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}')
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}')
instances = defaultdict(dict)
instances.update({
'Group': {
'gci': gci,
'gri': gri
},
'Person': {
'lugt': lugt
}
})
return instances
def static_db_instance(self, *keys, **kwargs):
uri = self.helper.make_shared_uri('DB', *keys)
label = ' '.join(keys)
name = kwargs.get('name', f'STAR {label} Database')
db = vocab.Database(ident=uri, label=name)
db.identified_by = vocab.PrimaryName(ident='', content=name)
creator = kwargs.get('creator')
if creator:
creation = model.Creation(ident='')
creation.carried_out_by = creator
db.created_by = creation
return db
def add_title(self, data):
'''
Special handling is given to modeling of the title (PrimaryName) of the article.
If we can detect that it is written in a language that matches one of the
languages asserted for either the document or summaries, then we assert that
as the language of the title Linguistic Object.
'''
if data.get('title'):
title = data['title']
restrict = data['_declared_languages']
lang = self.helper.validated_string_language(title, restrict)
pn = vocab.PrimaryName(ident='', content=title)
if lang:
pn.language = lang
data['identifiers'].append(pn)
def model_series_group(self, record, data):
if not data:
return
record.setdefault('identifiers', [])
record.setdefault('referred_to_by', [])
record.setdefault('language', [])
title = data.get('title')
title_translated = data.get('title_translated')
variant_titles = _as_list(data.get('variant_title'))
related_titles = _as_list(data.get('related_title'))
lang_docs = _as_list(data['lang_doc'])
frequency = data.get('frequency')
start_year = data.get('start_year')
cease_year = data.get('cease_year')
issn = data.get('issn')
coden = data.get('coden')
if title:
record['label'] = title
record['identifiers'].append(
vocab.PrimaryName(ident='', content=title))
if title_translated:
record['identifiers'].append(
vocab.TranslatedTitle(ident='', content=title))
for vtitle in variant_titles:
record['identifiers'].append(vocab.Title(ident='', content=vtitle))
for lang in lang_docs:
l = self.helper.language_object_from_code(lang)
if l:
record['language'].append(l)
if frequency:
record['referred_to_by'].append(
vocab.Note(ident='', content=frequency))
if start_year:
record['_publishing_start_year'] = start_year
if cease_year:
record['_publishing_cease_year'] = cease_year
if issn:
record['identifiers'].append(
vocab.IssnIdentifier(ident='', content=issn))
if coden:
record['identifiers'].append(
vocab.CodenIdentifier(ident='', content=coden))
def __call__(self, data: dict):
'''Add modeling for the entry describing a person/group in the PSCP PEOPLE dataset.'''
recno = data['star_record_no']
auth_name = data.get('auth_name')
record_uri = self.helper.make_proj_uri('ENTRY', 'PEOPLE', recno)
content = data['star_csv_data']
record = vocab.EntryTextForm(
ident=record_uri,
label=f'Entry recorded in PSCP PEOPLE dataset for {auth_name}',
content=content)
creation = model.Creation(ident='')
creation.carried_out_by = self.helper.static_instances.get_instance(
'Group', 'gpi')
record.created_by = creation
record.identified_by = self.helper.gpi_number_id(
recno, vocab.StarNumber)
record.identified_by = vocab.PrimaryName(
ident='', content=f'STAR Person Authority Entry {recno}')
record.part_of = self.helper.static_instances.get_instance(
'LinguisticObject', 'db-people')
data['_entry_record'] = add_crom_data({'uri': record_uri}, record)
yield data
def __call__(self, data: dict, post_sale_map, unique_catalogs,
subject_genre, destruction_types_map):
'''Add modeling for an object described by a sales record'''
hmo = get_crom_object(data)
parent = data['parent_data']
auction_data = parent.get('auction_of_lot')
if auction_data:
lno = str(auction_data['lot_number'])
data.setdefault('identifiers', [])
if not lno:
warnings.warn(f'Setting empty identifier on {hmo.id}')
data['identifiers'].append(vocab.LotNumber(ident='', content=lno))
else:
warnings.warn(f'***** NO AUCTION DATA FOUND IN populate_object')
cno = auction_data['catalog_number']
lno = auction_data['lot_number']
date = implode_date(auction_data, 'lot_sale_')
lot = self.helper.shared_lot_number_from_lno(
lno
) # the current key for this object; may be associated later with prev and post object keys
now_key = (cno, lno, date)
data['_locations'] = []
data['_events'] = []
record = self._populate_object_catalog_record(data, parent, lot, cno,
parent['pi_record_no'])
self._populate_object_visual_item(data, subject_genre)
self._populate_object_destruction(data, parent, destruction_types_map)
self.populate_object_statements(data)
self._populate_object_present_location(data, now_key,
destruction_types_map)
self._populate_object_notes(data, parent, unique_catalogs)
self._populate_object_prev_post_sales(data, now_key, post_sale_map)
for p in data.get('portal', []):
url = p['portal_url']
hmo.referred_to_by = vocab.WebPage(ident=url, label=url)
if 'title' in data:
title = data['title']
if not hasattr(hmo, '_label'):
typestring = data.get('object_type', 'Object')
hmo._label = f'{typestring}: “{title}”'
del data['title']
shorter = truncate_with_ellipsis(title, 100)
if shorter:
description = vocab.Description(ident='', content=title)
description.referred_to_by = record
hmo.referred_to_by = description
title = shorter
t = vocab.PrimaryName(ident='', content=title)
t.classified_as = model.Type(
ident='http://vocab.getty.edu/aat/300417193', label='Title')
t.referred_to_by = record
data['identifiers'].append(t)
for d in data.get('other_titles', []):
title = d['title']
t = vocab.Name(ident='', content=title)
data['identifiers'].append(t)
return data
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 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_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))
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}')
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}')
materials = {
aat: model.Material(ident=f'http://vocab.getty.edu/aat/{aat}',
label=label)
for aat, label in self.services['materials'].items()
}
instances = defaultdict(dict)
instances.update({
'Group': {
'gci': gci,
'gri': gri,
'knoedler': knoedler
},
'Person': {
'lugt': lugt
},
'Material': materials,
'Place': self._static_place_instances()
})
return instances
def model_issue_group(self, record, data, seq):
record.setdefault('^part', [])
issue_id = data['issue_id']
title = data.get('title')
title_translated = data.get('title_translated')
date = data.get('date')
# issue_group/date/display_date
# issue_group/date/sort_year
volume = data.get('volume')
number = data.get('number')
note = data.get('note')
journal_label = record['label']
issue_label = f'Issue of {journal_label}'
if title:
issue_label = f'{journal_label}: “{title}”'
if volume and number:
issue_label = f'{issue_label} (v. {volume}, n. {number})'
elif volume and number:
issue_label = f'{journal_label} (v. {volume}, n. {number})'
jid = record['record_desc_group']['record_id']
issue = {
'uri': self.helper.issue_uri(jid, issue_id),
'label': issue_label,
'object_type': vocab.IssueText,
'identifiers': [self.helper.gci_number_id(issue_id)],
'referred_to_by': [],
'used_for': [],
}
if title:
issue['identifiers'].append(
vocab.PrimaryName(ident='', content=title))
if title_translated:
issue['identifiers'].append(
vocab.TranslatedTitle(ident='', content=title_translated))
if date:
display_date = date.get('display_date')
sort_year = date.get('sort_year')
if display_date or sort_year:
a_uri = issue['uri'] + f'-pub'
a = vocab.Publishing(ident=a_uri,
label=f'Publishing of {issue_label}')
ts = model.TimeSpan(ident='')
if display_date:
ts._label = display_date
ts.identified_by = vocab.DisplayName(ident='',
content=display_date)
if sort_year:
try:
year = int(sort_year)
ts.begin_of_the_begin = '%04d-01-01:00:00:00Z' % (
year, )
ts.end_of_the_end = '%04d-01-01:00:00:00Z' % (year +
1, )
except:
pass
a.timespan = ts
issue['used_for'].append(a)
# TODO:
# volume
# number
if note:
issue['referred_to_by'].append(vocab.Note(ident='', content=note))
mlalo = MakeLinkedArtLinguisticObject()
mlalo(issue)
i = get_crom_object(issue)
for a in issue.get('used_for', []):
i.used_for = a
record['^part'].append(issue)
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
def add_names(self,
data: dict,
referrer=None,
role=None,
group=False,
**kwargs):
'''
Based on the presence of `auth_name` and/or `name` fields in `data`, sets the
`label`, `names`, and `identifier` keys to appropriate strings/`model.Identifier`
values.
If the `role` string is given (e.g. 'artist'), also sets the `role_label` key
to a value (e.g. 'artist “RUBENS, PETER PAUL”').
'''
data.setdefault('identifiers', [])
auth_name = data.get('auth_name', '')
disp_name = data.get('auth_display_name')
name_types = [vocab.PrimaryName]
personalNameType = vocab.CorporateName if group else vocab.PersonalName
if disp_name:
if auth_name:
data['identifiers'].append(
vocab.PrimaryName(ident='', content=auth_name))
data['label'] = auth_name
auth_name = disp_name
name_types = [personalNameType, vocab.DisplayName]
role_label = None
if self.acceptable_person_auth_name(auth_name):
if role:
role_label = f'{role} “{auth_name}”'
data.setdefault('label', auth_name)
pname = vocab.make_multitype_obj(
*name_types, ident='', content=auth_name
) # NOTE: most of these are also vocab.SortName, but not 100%, so witholding that assertion for now
if referrer:
pname.referred_to_by = referrer
data['identifiers'].append(pname)
data.setdefault('names', [])
names = []
name = data.get('name')
if name:
del data[
'name'] # this will be captured in the 'names' array, so remove it here so the output isn't duplicated
names.append(name)
variant_names = data.get('variant_names')
if variant_names:
names += [n.strip() for n in variant_names.split(';')]
for name in names:
if role and not role_label:
role_label = f'{role} “{name}”'
name_kwargs = {}
# name_kwargs['classified_as'] = model.Type(ident='http://vocab.getty.edu/aat/300266386', label='Personal Name')
name_kwargs['classified_as'] = personalNameType
if referrer:
name_kwargs['referred_to_by'] = [referrer]
data['names'].append((name, name_kwargs))
data.setdefault('label', name)
data.setdefault('label', '(Anonymous)')
if role and not role_label:
role_label = f'anonymous {role}'
if role:
data['role_label'] = role_label