def test_init_params(self):
p1 = model.Person(ident="urn:uuid:1234")
self.assertEqual(p1.id, "urn:uuid:1234")
p2 = model.Person(ident="http://schema.org/Foo")
self.assertEqual(p2.id, "schema:Foo")
p3 = model.Name(content="Test")
self.assertEqual(p3.content, "Test")
c = model.MonetaryAmount(value=10)
self.assertEqual(c.value, 10)
n = model.Name(value="Rob")
self.assertEqual(n.content, "Rob")
i = model.Identifier(content="xyz123")
self.assertEqual(i.content, "xyz123")
i2 = model.Identifier(value="abc")
self.assertEqual(i2.content, "abc")
def __call__(self, data:dict, event_properties, date_modifiers):
'''Add modeling for an auction event based on properties of the supplied `data` dict.'''
record = get_crom_object(data['_catalog'])
cno = data['catalog_number']
sale_type = data.get('non_auction_flag', 'Auction')
ts, begin, end = timespan_from_bound_components(
data,
date_modifiers,
'sale_begin_', 'begin',
'sale_end_', 'eoe'
)
event_properties['auction_dates'][cno] = (ts, begin, end)
event_properties['auction_date_label'][cno] = ts._label
event_date_label = event_properties['auction_date_label'].get(cno)
auction, uid, uri = self.helper.sale_event_for_catalog_number(cno, sale_type, date_label=event_date_label)
auction.referred_to_by = record
auction.identified_by = model.Name(ident='', content=auction._label)
data['uid'] = uid
data['uri'] = uri
add_crom_data(data=data, what=auction)
catalog = get_crom_object(data['_catalog'])
data['_record'] = data['_catalog']
return data
def model_person_or_group(self,
data: dict,
a: dict,
attribution_group_types,
attribution_group_names,
role='artist',
seq_no=0,
sales_record=None):
if get_crom_object(a):
return a
mods = a['modifiers']
artist = self.helper.add_person(a,
record=sales_record,
relative_id=f'artist-{seq_no+1}',
role=role)
artist_label = a['label']
person = get_crom_object(a)
if mods:
GROUP_TYPES = set(attribution_group_types.values())
GROUP_MODS = {
k
for k, v in attribution_group_types.items() if v in GROUP_TYPES
}
if mods.intersects(GROUP_MODS):
mod_name = list(GROUP_MODS
& mods)[0] # TODO: use all matching types?
clsname = attribution_group_types[mod_name]
cls = getattr(vocab, clsname)
group_name = attribution_group_names[clsname]
group_label = f'{group_name} {artist_label}'
a['label'] = group_label
# The group URI is just the person URI with a suffix. In any case
# where the person is merged, the group should be merged as well.
# For example, when if "RUBENS" is merged, "School of RUBENS" should
# also be merged.
group_id = a['uri'] + f'-{clsname}'
group = cls(ident=group_id, label=group_label)
group.identified_by = model.Name(ident='', content=group_label)
formation = model.Formation(
ident='', label=f'Formation of {group_label}')
formation.influenced_by = person
group.formed_by = formation
pi_record_no = data['pi_record_no']
group_uri_key = ('GROUP', 'PI', pi_record_no, f'{role}Group')
group_data = {
'uri': group_id,
'uri_keys': group_uri_key,
'modifiers': mods,
'label': group_label
}
add_crom_data(group_data, group)
data['_organizations'].append(group_data)
return group_data
add_crom_data(a, artist)
return a
def test_init_params(self):
p1 = model.Person(ident="urn:uuid:1234")
self.assertEqual(p1.id, "urn:uuid:1234")
p2 = model.Person(ident="http://schema.org/Foo")
self.assertEqual(p2.id, "schema:Foo")
p3 = model.Name(content="Test")
self.assertEqual(p3.content, "Test")
def professional_activity(self,
name: str,
century=None,
date_range=None,
classified_as=None,
**kwargs):
'''
Return a vocab.Active object representing the professional activities
of the `name`d person.
If `century` or `date_range` arguments are supplied, they are used to
associate a timespan with the activity.
If a `classified_as` list is supplied, it is used to further classify
the `vocab.Active` object.
'''
if not classified_as:
classified_as = [model.Activity]
args = {'ident': '', 'label': f'Professional activity of {name}'}
if 'ident' in kwargs:
args['ident'] = kwargs['ident']
a = vocab.make_multitype_obj(*classified_as, **args)
ts = self.active_timespan(century=century,
date_range=date_range,
**kwargs)
if ts:
if 'verbatim_active_period' in kwargs:
ts.identified_by = model.Name(
ident='', content=kwargs['verbatim_active_period'])
a.timespan = ts
return 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', 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 populate_destruction_events(self,
data: dict,
note,
*,
type_map,
location=None):
destruction_types_map = type_map
hmo = get_crom_object(data)
title = data.get('title')
short_title = truncate_with_ellipsis(title, 100) or title
r = re.compile(
r'[Dd]estroyed(?: (?:by|during) (\w+))?(?: in (\d{4})[.]?)?')
m = r.search(note)
if m:
method = m.group(1)
year = m.group(2)
# The destruction URI is just the object URI with a suffix. When URIs are
# reconciled during prev/post sale rewriting, this will allow us to also reconcile
# the URIs for the destructions (of which there should only be one per object)
dest_uri = hmo.id + '-Destruction'
d = model.Destruction(ident=dest_uri,
label=f'Destruction of “{short_title}”')
d.referred_to_by = vocab.Note(ident='', content=note)
if year is not None:
begin, end = date_cleaner(year)
ts = timespan_from_outer_bounds(begin, end)
ts.identified_by = model.Name(ident='', content=year)
d.timespan = ts
if method:
with suppress(KeyError, AttributeError):
type_name = destruction_types_map[method.lower()]
otype = vocab.instances[type_name]
event = model.Event(
label=
f'{method.capitalize()} event causing the destruction of “{short_title}”'
)
event.classified_as = otype
d.caused_by = event
data['_events'].append(add_crom_data(data={}, what=event))
if location:
current = parse_location_name(
location, uri_base=self.helper.uid_tag_prefix)
# The place URI used for destruction events is based on the object URI with
# a suffix. When URIs are reconciled during prev/post sale rewriting, this
# will allow us to also reconcile the URIs for the places of destruction
# (of which there should only be one hierarchy per object)
base_uri = hmo.id + '-Destruction-Place,'
place_data = self.helper.make_place(current, base_uri=base_uri)
place = get_crom_object(place_data)
if place:
data['_locations'].append(place_data)
d.took_place_at = place
hmo.destroyed_by = d
def set_properties(self, data, thing):
super().set_properties(data, thing)
with suppress(KeyError):
thing._label = str(data['label'])
for event in data.get('events', []):
thing.carried_out = event
for n in data.get('nationality', []):
thing.classified_as = n
if data.get('formation'):
b = model.Formation()
ts = model.TimeSpan(ident='')
if 'formation_clean' in data and data['formation_clean']:
if data['formation_clean'][0]:
ts.begin_of_the_begin = data['formation_clean'][
0].strftime("%Y-%m-%dT%H:%M:%SZ")
if data['formation_clean'][1]:
ts.end_of_the_end = data['formation_clean'][1].strftime(
"%Y-%m-%dT%H:%M:%SZ")
verbatim = data['formation']
ts._label = verbatim
ts.identified_by = model.Name(ident='', content=verbatim)
b.timespan = ts
b._label = "Formation of %s" % thing._label
thing.formed_by = b
if data.get('dissolution'):
d = model.Dissolution()
ts = model.TimeSpan(ident='')
if 'dissolution_clean' in data and data['dissolution_clean']:
if data['dissolution_clean'][0]:
ts.begin_of_the_begin = data['dissolution_clean'][
0].strftime("%Y-%m-%dT%H:%M:%SZ")
if data['dissolution_clean'][1]:
ts.end_of_the_end = data['dissolution_clean'][1].strftime(
"%Y-%m-%dT%H:%M:%SZ")
verbatim = data['dissolution']
ts._label = verbatim
ts.identified_by = model.Name(ident='', content=verbatim)
d.timespan = ts
d._label = "Dissolution of %s" % thing._label
thing.dissolved_by = d
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 timespan_after(before):
ts = model.TimeSpan(ident='')
try:
ts.begin_of_the_begin = before.end_of_the_end
with suppress(AttributeError):
l = f'After {before._label}'
l.identified_by = model.Name(ident='', content=l)
ts._label = l
return ts
except AttributeError:
return None
def timespan_before(after):
ts = model.TimeSpan(ident='')
try:
ts.end_of_the_end = after.begin_of_the_begin
with suppress(AttributeError):
l = f'Before {after._label}'
l.identified_by = model.Name(ident='', content=l)
ts._label = l
return ts
except AttributeError:
return None
def populate_destruction_events(self,
data: dict,
note,
*,
type_map,
location=None):
destruction_types_map = type_map
hmo = get_crom_object(data)
title = data.get('title')
short_title = truncate_with_ellipsis(title, 100) or title
r = re.compile(
r'[Dd]estroyed(?: (?:by|during) (\w+))?(?: in (\d{4})[.]?)?')
m = r.search(note)
if m:
method = m.group(1)
year = m.group(2)
dest_id = hmo.id + '-Destr'
d = model.Destruction(ident=dest_id,
label=f'Destruction of “{short_title}”')
d.referred_to_by = vocab.Note(ident='', content=note)
if year is not None:
begin, end = date_cleaner(year)
ts = timespan_from_outer_bounds(begin, end)
ts.identified_by = model.Name(ident='', content=year)
d.timespan = ts
if method:
with suppress(KeyError, AttributeError):
type_name = destruction_types_map[method.lower()]
otype = vocab.instances[type_name]
event = model.Event(
label=
f'{method.capitalize()} event causing the destruction of “{short_title}”'
)
event.classified_as = otype
d.caused_by = event
data['_events'].append(add_crom_data(data={}, what=event))
if location:
current = parse_location_name(
location, uri_base=self.helper.uid_tag_prefix)
base_uri = hmo.id + '-Place,'
place_data = self.helper.make_place(current, base_uri=base_uri)
place = get_crom_object(place_data)
if place:
data['_locations'].append(place_data)
d.took_place_at = place
hmo.destroyed_by = d
def __call__(self, data: dict):
'''Add modeling for an auction event based on properties of the supplied `data` dict.'''
cno = data['catalog_number']
sale_type = data.get('non_auction_flag', 'Auction')
auction, uid, uri = self.helper.sale_event_for_catalog_number(
cno, sale_type)
auction.identified_by = model.Name(ident='', content=auction._label)
data['uid'] = uid
data['uri'] = uri
add_crom_data(data=data, what=auction)
catalog = get_crom_object(data['_catalog'])
data['_record'] = data['_catalog']
return data
def test_boundary_setter(self):
vocab.add_linked_art_boundary_check()
p = model.Person()
p2 = model.Person()
n = model.Name()
n.content = "Test"
p2.identified_by = n
p.exact_match = p2
# Now, Test should not appear in the resulting JSON of p
factory.linked_art_boundaries = True
js = factory.toJSON(p)
self.assertTrue(not 'identified_by' in js['exact_match'][0])
factory.linked_art_boundaries = False
js = factory.toJSON(p)
self.assertTrue('identified_by' in js['exact_match'][0])
def test_production_mode(self):
# model.factory.production_mode()
# Can't unset the cached hierarchy
# and it causes the test for the hierarchy to fail
model.factory.validate_profile = False
model.factory.validate_properties = False
model.factory.validate_range = False
model.factory.validate_multiplicity = False
p = model.Person()
p.identified_by = model.Name(value="abc")
p.part = model.HumanMadeObject()
js = model.factory.toJSON(p)
model.factory.production_mode(state=False)
def model_sojourn(self, data, loc):
base_uri = self.helper.make_proj_uri('PLACE', '')
cb = data.get('corporate_body', False)
sojourn_type = vocab.Establishment if cb else vocab.Residing
sdata = {
'type': sojourn_type,
'referred_to_by': [],
}
verbatim_date = loc.get('address_date')
if verbatim_date:
date_range = date_cleaner(verbatim_date)
if date_range:
begin, end = date_range
ts = timespan_from_outer_bounds(*date_range)
ts.identified_by = model.Name(ident='', content=verbatim_date)
sdata['timespan'] = add_crom_data(
{
'address_date': verbatim_date,
'begin': begin,
'end': end
}, ts)
current = None
l = loc.get('location')
if l:
current = parse_location_name(l, uri_base=self.helper.proj_prefix)
address = loc.get('address')
if address:
current = {
'name': address,
'part_of': current,
'type': 'address',
}
for k in ('address_note', 'location_note'):
note = loc.get(k)
if note:
sdata['referred_to_by'].append(
vocab.Note(ident='', content=note))
if current:
place_data = self.helper.make_place(current, base_uri=base_uri)
data['_places'].append(place_data)
sdata['place'] = place_data
return sdata
def extract_physical_dimensions(dimstr, **kwargs):
dimensions = dimensions_cleaner(dimstr, **kwargs)
if dimensions:
for orig_d in dimensions:
dimdata = normalized_dimension_object(orig_d, source=dimstr)
if dimdata:
dimension, label = dimdata
if dimension.which == 'height':
dim = vocab.Height(ident='')
elif dimension.which == 'width':
dim = vocab.Width(ident='')
else:
dim = vocab.PhysicalDimension(ident='')
dim.value = dimension.value
dim.identified_by = model.Name(ident='', content=label)
unit = vocab.instances.get(dimension.unit)
if unit:
dim.unit = unit
yield dim
def test_not_multiple_instance(self):
who = model.Person()
n = model.Name(content="Test")
who.identified_by = n
model.factory.multiple_instances_per_property = "error"
self.assertRaises(model.DataError, who.__setattr__, 'identified_by', n)
self.assertEqual(who.identified_by, [n])
model.factory.multiple_instances_per_property = "drop"
who.identified_by = n
self.assertEqual(who.identified_by, [n, n])
# and check that only serialized once
js = model.factory.toJSON(who)
self.assertEqual(len(js['identified_by']), 1)
model.factory.multiple_instances_per_property = "allow"
js = model.factory.toJSON(who)
self.assertEqual(len(js['identified_by']), 2)
def __call__(self, data: dict, location_codes, unique_catalogs):
'''Add information about the ownership of a physical copy of an auction catalog'''
# Add the URI of this physical catalog to `unique_catalogs`. This data will be used
# later to figure out which catalogs can be uniquely identified by a catalog number
# and owner code (e.g. for owners who do not have multiple copies of a catalog).
cno = data['catalog_number']
owner_code = data['owner_code']
copy_number = data.get('copy_number', '')
owner_name = None
entry_record = get_crom_object(data.get('_catalog'))
with suppress(KeyError):
owner_name = location_codes[owner_code]
owner_uri = self.helper.make_proj_uri('ORGANIZATION',
'LOCATION-CODE', owner_code)
data['_owner'] = {
'label':
owner_name,
'uri':
owner_uri,
'referred_to_by': [entry_record],
'identifiers': [
model.Name(ident='', content=owner_name),
model.Identifier(ident='', content=str(owner_code))
],
}
owner = model.Group(ident=owner_uri)
owner.referred_to_by = entry_record
add_crom_data(data['_owner'], owner)
if not owner_code:
warnings.warn(f'Setting empty identifier on {owner.id}')
add_crom_data(data=data['_owner'], what=owner)
catalog = get_crom_object(data)
catalog.current_owner = owner
owner_uri = self.helper.physical_catalog_uri(
cno, owner_code, None
) # None here because we want a key that will stand in for all the copies belonging to a single owner
copy_uri = self.helper.physical_catalog_uri(cno, owner_code,
copy_number)
unique_catalogs[owner_uri].add(copy_uri)
return data
def set_lot_objects(self, lot, cno, lno, auction_of_lot_uri, data,
sale_type):
'''Associate the set of objects with the auction lot.'''
shared_lot_number = self.helper.shared_lot_number_from_lno(lno)
set_type = vocab.AuctionLotSet if sale_type == 'Auction' else vocab.CollectionSet
coll_label = f'Object Set for Lot {cno} {shared_lot_number}'
coll = set_type(ident=f'{auction_of_lot_uri}-Set', label=coll_label)
coll.identified_by = model.Name(ident='', content=coll_label)
est_price = data.get('estimated_price')
if est_price:
self.set_possible_attribute(coll, 'dimension', est_price)
start_price = data.get('start_price')
if start_price:
self.set_possible_attribute(coll, 'dimension', start_price)
ask_price = data.get('ask_price')
if ask_price:
self.set_possible_attribute(coll, 'dimension', ask_price)
lot.used_specific_object = coll
data['_lot_object_set'] = add_crom_data(data={}, what=coll)
def set_la_name(thing, value, title_type=None, set_label=False):
if value is None:
return None
if isinstance(value, tuple):
label, language = value
else:
label = value
language = None
if set_label:
if not label:
warnings.warn(f'Setting empty label on {thing.id}')
thing._label = label
name = model.Name(ident='', content=label)
if title_type is not None:
if isinstance(title_type, model.Type):
name.classified_as = title_type
else:
vocab.add_classification(name, title_type)
thing.identified_by = name
if language is not None:
name.language = language
return name
def set_lot_date(lot, auction_data, event_dates):
'''Associate a timespan with the auction lot.'''
date = implode_date(auction_data, 'lot_sale_')
if date:
begin = implode_date(auction_data, 'lot_sale_', clamp='begin')
end = implode_date(auction_data, 'lot_sale_', clamp='eoe')
bounds = [begin, end]
else:
bounds = []
if bounds:
if auction_data.get('lot_sale_mod'):
# if the lot sale date is marked as uncertain:
# - use the event end date as the lot sale's end_of_the_end
# - if the event doesn't have a known end date, assert no end_of_the_end for the lot sale
if event_dates and event_dates[1]:
bounds[1] = event_dates[1]
else:
bounds[1] = None
ts = timespan_from_outer_bounds(*bounds)
ts.identified_by = model.Name(ident='', content=date)
lot.timespan = ts
def _populate_object_catalog_record(self, data: dict, parent, lot, cno,
rec_num):
hmo = get_crom_object(data)
catalog_uri = self.helper.make_proj_uri('CATALOG', cno)
catalog = vocab.AuctionCatalogText(ident=catalog_uri,
label=f'Sale Catalog {cno}')
record_uri = self.helper.make_proj_uri('CATALOG', cno, 'RECORD',
rec_num)
lot_object_id = parent['lot_object_id']
puid = parent.get('persistent_puid')
puid_id = self.helper.gri_number_id(puid)
record = vocab.ParagraphText(
ident=record_uri,
label=
f'Sale recorded in catalog: {lot_object_id} (record number {rec_num})'
)
record_data = {'uri': record_uri}
record_data['identifiers'] = [
model.Name(ident='', content=f'Record of sale {lot_object_id}'),
puid_id
]
record.part_of = catalog
if parent.get('transaction'):
record.referred_to_by = vocab.PropertyStatusStatement(
ident='',
label='Transaction type for sales record',
content=parent['transaction'])
record.about = hmo
data['_record'] = add_crom_data(data=record_data, what=record)
return record
def make_ymd_timespan(data: dict, start_prefix="", end_prefix="", label=""):
y = f'{start_prefix}year'
m = f'{start_prefix}month'
d = f'{start_prefix}day'
y2 = f'{end_prefix}year'
m2 = f'{end_prefix}month'
d2 = f'{end_prefix}day'
t = model.TimeSpan(ident='')
if not label:
label = ymd_to_label(data[y], data[m], data[d])
if y != y2:
lbl2 = ymd_to_label(data[y2], data[m2], data[d2])
label = f'{label} to {lbl2}'
t._label = label
if not label:
warnings.warn(f'Setting empty name on {t.id}')
t.identified_by = model.Name(ident='', content=label)
t.begin_of_the_begin = ymd_to_datetime(data[y], data[m], data[d])
t.end_of_the_end = ymd_to_datetime(data[y2],
data[m2],
data[d2],
which="end")
return t
def extract_monetary_amount(data,
add_citations=False,
currency_mapping=CURRENCY_MAPPING,
source_mapping=None,
truncate_label_digits=2):
'''
Returns a `MonetaryAmount`, `StartingPrice`, or `EstimatedPrice` object
based on properties of the supplied `data` dict. If no amount or currency
data is found in found, returns `None`.
For `EstimatedPrice`, values will be accessed from these keys:
- amount: `est_price_amount` or `est_price`
- currency: `est_price_currency` or `est_price_curr`
- note: `est_price_note` or `est_price_desc`
- bibliographic statement: `est_price_citation`
For `StartingPrice`, values will be accessed from these keys:
- amount: `start_price_amount` or `start_price`
- currency: `start_price_currency` or `start_price_curr`
- note: `start_price_note` or `start_price_desc`
- bibliographic statement: `start_price_citation`
For `MonetaryAmount` prices, values will be accessed from these keys:
- amount: `price_amount` or `price`
- currency: `price_currency` or `price_curr`
- note: `price_note` or `price_desc`
- bibliographic statement: `price_citation`
'''
amount_type = 'Price'
if 'price' in data or 'price_amount' in data or 'amount' in data:
amnt = model.MonetaryAmount(ident='')
price_amount = data.get('price_amount',
data.get('price', data.get('amount')))
price_currency = data.get(
'currency', data.get('price_currency', data.get('price_curr')))
note = data.get('price_note', data.get('price_desc', data.get('note')))
cite = data.get('price_citation', data.get('citation'))
source = data.get('price_source', '')
elif 'est_price' in data or 'est_price_amount' in data:
amnt = vocab.EstimatedPrice(ident='')
price_amount = data.get('est_price_amount', data.get('est_price'))
price_currency = data.get(
'est_price_currency',
data.get('est_price_curr', data.get('currency')))
amount_type = 'Estimated Price'
note = data.get('est_price_note',
data.get('est_price_desc', data.get('note')))
cite = data.get('est_price_citation', data.get('citation'))
source = data.get('est_price_source', data.get('est_price_so', ''))
elif 'start_price' in data or 'start_price_amount' in data:
amnt = vocab.StartingPrice(ident='')
price_amount = data.get('start_price_amount', data.get('start_price'))
price_currency = data.get(
'start_price_currency',
data.get('start_price_curr', data.get('currency')))
amount_type = 'Starting Price'
note = data.get('start_price_note',
data.get('start_price_desc', data.get('note')))
cite = data.get('start_price_citation', data.get('citation'))
source = data.get('start_price_source', data.get('start_price_so', ''))
elif 'ask_price' in data or 'ask_price_amount' in data:
amnt = vocab.AskingPrice(ident='')
price_amount = data.get('ask_price_amount', data.get('ask_price'))
price_currency = data.get(
'ask_price_currency',
data.get('ask_price_curr', data.get('currency')))
amount_type = 'Asking Price'
note = data.get('ask_price_note',
data.get('ask_price_desc', data.get('note')))
cite = data.get('ask_price_citation', data.get('citation'))
source = data.get('ask_price_source', data.get('ask_price_so', ''))
else:
return None
price_amount_label = price_amount
if price_amount or price_currency:
if cite and add_citations:
amnt.referred_to_by = vocab.BibliographyStatement(ident='',
content=cite)
if note:
amnt.referred_to_by = vocab.Note(ident='', content=note)
if price_amount:
try:
value = price_amount
value = value.replace('[?]', '')
value = value.replace('?', '')
value = value.strip()
if re.search(re.compile(r',\d\d\d'), value):
value = value.replace(',', '')
value = float(value)
label_fmt = '{:,.%df}' % truncate_label_digits
price_amount_label = label_fmt.format(value)
amnt.value = value
except ValueError:
amnt._label = price_amount
amnt.identified_by = model.Name(ident='', content=price_amount)
# warnings.warn(f'*** Not a numeric price amount: {value}')
if price_currency:
price_currency_key = price_currency
try:
price_currency_key = currency_mapping[
price_currency_key.lower()]
except KeyError:
pass
if isinstance(price_currency_key, model.BaseResource):
amnt.currency = price_currency_key
elif price_currency_key in vocab.instances:
amnt.currency = vocab.instances[price_currency_key]
else:
warnings.warn('*** No currency instance defined for %s' %
(price_currency_key, ))
if price_amount_label and price_currency:
amnt._label = '%s %s' % (price_amount_label, price_currency)
elif price_amount:
amnt._label = '%s' % (price_amount, )
return amnt
return None
def _populate_object_present_location(self, data: dict, now_key,
destruction_types_map):
hmo = get_crom_object(data)
location = data.get('present_location')
if location:
loc = location.get('geog')
note = location.get('note')
if loc:
if 'destroyed ' in loc.lower():
self.populate_destruction_events(
data, loc, type_map=destruction_types_map)
elif isinstance(note, str) and 'destroyed ' in note.lower():
# the object was destroyed, so any "present location" data is actually
# an indication of the location of destruction.
self.populate_destruction_events(
data,
note,
type_map=destruction_types_map,
location=loc)
else:
# TODO: if `parse_location_name` fails, still preserve the location string somehow
current = parse_location_name(
loc, uri_base=self.helper.uid_tag_prefix)
inst = location.get('inst')
if inst:
owner_data = {
'label': f'{inst} ({loc})',
'identifiers':
[model.Name(ident='', content=inst)]
}
ulan = None
with suppress(ValueError, TypeError):
ulan = int(location.get('insi'))
if ulan:
owner_data['ulan'] = ulan
owner_data['uri'] = self.helper.make_proj_uri(
'ORG', 'ULAN', ulan)
else:
owner_data['uri'] = self.helper.make_proj_uri(
'ORG', 'NAME', inst, 'PLACE', loc)
else:
owner_data = {
'label':
'(Anonymous organization)',
'uri':
self.helper.make_proj_uri('ORG', 'CURR-OWN',
*now_key),
}
if note:
owner_data['note'] = note
base_uri = hmo.id + '-Place,'
place_data = self.helper.make_place(current,
base_uri=base_uri)
place = get_crom_object(place_data)
make_la_org = pipeline.linkedart.MakeLinkedArtOrganization(
)
owner_data = make_la_org(owner_data)
owner = get_crom_object(owner_data)
acc = location.get('acc')
if acc:
acc_number = vocab.AccessionNumber(ident='',
content=acc)
hmo.identified_by = acc_number
assignment = model.AttributeAssignment(ident='')
assignment.carried_out_by = owner
acc_number.assigned_by = assignment
owner.residence = place
data['_locations'].append(place_data)
data['_final_org'] = owner_data
else:
pass # there is no present location place string
def related_procurement(self,
hmo,
tx_label_args,
current_tx=None,
current_ts=None,
buyer=None,
seller=None,
previous=False,
ident=None,
make_label=None):
'''
Returns a new `vocab.ProvenanceEntry` object (and related acquisition) that is temporally
related to the supplied procurement and associated data. The new procurement is for
the given object, and has the given buyer and seller (both optional).
If the `previous` flag is `True`, the new procurement is occurs before `current_tx`,
and if the timespan `current_ts` is given, has temporal data to that effect. If
`previous` is `False`, this relationship is reversed.
The `make_label` argument, if supplied, is used as a function to generate the
label for the provenance entry. Its arguments (generated in `handle_prev_post_owner`)
are:
* helper: the helper object for the current pipeline
* sale_type: the sale type passed to `handle_prev_post_owner` (e.g. "Auction")
* transaction: the transaction type being handled (e.g. "Sold")
* rel: a string describing the relationship between this provenance entry and the object (e.g. "leading to the previous ownership of")
* N trailing arguments used that are the contents of the `lot_object_key` tuple passed to `handle_prev_post_owner`
'''
def _make_label_default(helper, sale_type, transaction, rel, *args):
strs = [str(x) for x in args]
return ', '.join(strs)
if make_label is None:
make_label = _make_label_default
tx = vocab.ProvenanceEntry(ident=ident)
tx_label = make_label(*tx_label_args)
tx._label = tx_label
tx.identified_by = model.Name(ident='', content=tx_label)
if current_tx:
if previous:
tx.ends_before_the_start_of = current_tx
else:
tx.starts_after_the_end_of = current_tx
modifier_label = 'Previous' if previous else 'Subsequent'
try:
pacq = model.Acquisition(
ident='',
label=f'{modifier_label} Acquisition of: “{hmo._label}”')
pxfer = model.TransferOfCustody(
ident='',
label=f'{modifier_label} Transfer of Custody of: “{hmo._label}”'
)
except AttributeError:
pacq = model.Acquisition(ident='',
label=f'{modifier_label} Acquisition')
pxfer = model.TransferOfCustody(
ident='', label=f'{modifier_label} Transfer of Custody')
pacq.transferred_title_of = hmo
pxfer.transferred_custody_of = hmo
if buyer:
pacq.transferred_title_to = buyer
pxfer.transferred_custody_to = buyer
if seller:
pacq.transferred_title_from = seller
pxfer.transferred_custody_from = seller
tx.part = pacq
tx.part = pxfer
if current_ts:
if previous:
pacq.timespan = timespan_before(current_ts)
else:
pacq.timespan = timespan_after(current_ts)
return tx, pacq
def set_properties(self, data, thing):
super().set_properties(data, thing)
# TODO: this whole title_type thing isn't right. most of the identifiers below aren't titles
title_type = model.Type(ident='http://vocab.getty.edu/aat/300417193',
label='Title')
name = None
if 'label' in data:
name = set_la_name(thing,
data['label'],
title_type,
set_label=True)
for author in data.get('created_by', []):
thing.created_by = author
for a in data.get('used_for', []):
thing.used_for = a
for a in data.get('about', []):
thing.about = a
for c in data.get('classified_as', []):
thing.classified_as = c
for t in data.get('translations', []):
n = set_la_name(thing, t, title_type)
if name is not None:
n.translation_of = name
for content, itype, notes in data.get('qualified_identifiers', []):
ident = itype(content=content)
if not content:
warnings.warn(f'Setting empty identifier on {thing.id}')
thing.identified_by = ident
for n in notes:
ident.referred_to_by = n
code_type = None # TODO: is there a model.Type value for this sort of code?
for c in data.get('classifications', []):
if isinstance(c, model.Type):
classification = c
else:
cid, label = c
name = model.Name()
name.classified_as = title_type
name.content = label
classification = model.Type(label=label)
if not label:
warnings.warn(f'Setting empty name on {classification.id}')
classification.identified_by = name
code = model.Identifier()
code.classified_as = code_type
if not cid:
warnings.warn(f'Setting empty identifier on {code.id}')
code.content = cid
classification.identified_by = code
thing.about = classification
for c in data.get('indexing', []):
if isinstance(c, tuple):
cid, label = c
name = model.Name()
name.classified_as = title_type
name.content = label
indexing = model.Type(label=label)
if not label:
warnings.warn(f'Setting empty name on {indexing.id}')
indexing.identified_by = name
code = model.Identifier()
code.classified_as = code_type
code.content = cid
if not cid:
warnings.warn(f'Setting empty identifier on {code.id}')
indexing.identified_by = code
else:
indexing = c
thing.about = indexing
parents = data.get('part_of', [])
for parent_data in parents:
parent = get_crom_object(parent_data)
thing.part_of = parent
children = data.get('part', [])
for child_data in children:
child = get_crom_object(child_data)
thing.part = child
for carrier in data.get('carried_by', []):
hmo = get_crom_object(carrier)
thing.carried_by = hmo
for dimension in data.get('dimensions', []):
thing.dimension = dimension
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)