def transient(self, data):
data['$uri'] = url_for('entities_api.view', id=data.get('id'))
data['$ui'] = entity_url(data.get('id'))
if data.get('$bulk'):
data['$writeable'] = False
else:
collection_id = data.get('collection_id')
data['$writeable'] = request.authz.can_write(collection_id)
return data
def entity_links(self, data, pk, schemata):
return {
'self': url_for('entities_api.view', id=pk),
# 'similar': url_for('entities_api.similar', id=pk),
# 'documents': url_for('entities_api.documents', id=pk),
'references': url_for('entities_api.references', id=pk),
'tags': url_for('entities_api.tags', id=pk),
'ui': entity_url(pk)
}
def reconcile_index():
domain = app_ui_url.strip('/')
api_key = request.authz.role.api_key if request.authz.logged_in else None
meta = {
'name': settings.APP_TITLE,
'identifierSpace': 'http://rdf.freebase.com/ns/type.object.id',
'schemaSpace': 'http://rdf.freebase.com/ns/type.object.id',
'view': {
'url': entity_url('{{id}}')
},
'preview': {
'url': entity_url('{{id}}') + '?api_key=%s' % api_key,
'width': 800,
'height': 400
},
'suggest': {
'entity': {
'service_url':
domain,
'service_path':
url_for('reconcile_api.suggest_entity', api_key=api_key)
},
'type': {
'service_url': domain,
'service_path': url_for('reconcile_api.suggest_type')
},
'property': {
'service_url': domain,
'service_path': url_for('reconcile_api.suggest_property')
}
},
'defaultTypes': [{
'id': 'Entity',
'name': 'Persons and Companies'
}]
}
return jsonify(meta)
def reconcile_op(query):
"""Reconcile operation for a single query."""
parser = SearchQueryParser({
'limit': query.get('limit', '5'),
'strict': 'false'
}, request.authz)
name = query.get('query', '')
schema = query.get('type') or 'Thing'
entity = {
'id': 'fake',
'names': [name],
'fingerprints': [fingerprints.generate(name)],
'schemata': ensure_list(schema),
'schema': schema
}
for p in query.get('properties', []):
entity[p.get('pid')] = ensure_list(p.get('v'))
query = SimilarEntitiesQuery(parser, entity=entity)
matches = []
for doc in query.search().get('hits').get('hits'):
source = doc.get('_source')
match = {
'id': doc.get('_id'),
'name': source.get('name'),
'score': min(100, doc.get('_score') * 10),
'uri': entity_url(doc.get('_id')),
'match': source.get('name') == name
}
for type_ in get_freebase_types():
if source['schema'] == type_['id']:
match['type'] = [type_]
matches.append(match)
log.info("Reconciled: %r -> %d matches", name, len(matches))
return {
'result': matches,
'num': len(matches)
}
def generate_matches_sheet(workbook,
sheet,
collection,
match_collection,
authz,
links=True,
one_sheet=False,
offset=0,
limit=1000):
from aleph.views.serializers import MatchSchema
if one_sheet:
sheet_label = "All matches (top %s per collection)" % limit
else:
sheet_label = "%s (top %s)" % (match_collection.label, limit)
sheet.set_zoom(125)
parser = QueryParser({}, authz, limit=limit)
q_match = Match.find_by_collection(collection.id, match_collection.id)
matches = MatchQueryResult({}, q_match, parser=parser, schema=MatchSchema)
if offset < 3:
sheet.write(0, 0, '', workbook.header_format)
sheet.write(1, 0, 'Score', workbook.header_format)
sheet.merge_range(0, 1, 0, 4, collection.label, workbook.header_format)
sheet.write(1, 1, 'Name', workbook.header_format)
sheet.write(1, 2, 'Type', workbook.header_format)
sheet.write(1, 3, 'Country', workbook.header_format)
sheet.write(1, 4, 'Source URL', workbook.header_format)
sheet.merge_range(0, 5, 0, 8, sheet_label, workbook.header_format)
sheet.write(1, 5, 'Name', workbook.header_format)
sheet.write(1, 6, 'Type', workbook.header_format)
sheet.write(1, 7, 'Country', workbook.header_format)
if one_sheet:
sheet.write(1, 8, 'Collection', workbook.header_format)
sheet.freeze_panes(2, 0)
sheet.autofilter(1, 1, 2 + len(matches.results), 8)
widths = {}
for row, result in enumerate(matches.results, offset):
sheet.write_number(row, 0, int(result.score))
name = result.entity.get('name')
widths[1] = max(widths.get(1, 0), len(name))
if links:
url = entity_url(result.entity_id)
sheet.write_url(row, 1, url, workbook.link_format, name)
else:
sheet.write_string(row, 1, name)
schema = model.get(result.entity['schema'])
sheet.write_string(row, 2, schema.label)
countries = ', '.join(sorted(result.entity.get('countries', [])))
sheet.write_string(row, 3, countries.upper())
ent_props = result.entity.get('properties', {})
if (ent_props.get('sourceUrl') is not None):
source_url = ', '.join(ent_props.get('sourceUrl'))
else:
source_url = ''
sheet.write_string(row, 4, source_url)
name = result.match.get('name')
widths[5] = max(widths.get(5, 0), len(name))
if links:
url = entity_url(result.match_id)
sheet.write_url(row, 5, url, workbook.link_format, name)
else:
sheet.write_string(row, 5, name)
schema = model.get(result.match['schema'])
sheet.write_string(row, 6, schema.label)
countries = ', '.join(sorted(result.match.get('countries', [])))
sheet.write_string(row, 7, countries.upper())
if one_sheet:
sheet.write_string(row, 8, match_collection.label)
for idx, max_len in widths.items():
max_len = min(70, max(7, max_len + 1))
sheet.set_column(idx, idx, float(max_len))
return sheet
