def peek_query(args):
if not isinstance(args, MultiDict):
args = MultiDict(args)
text = args.get('q', '').strip()
q = text_query(text)
filters = parse_filters(args)
for entity in args.getlist('entity'):
filters.append(('entities.id', entity))
q = filter_query(q, filters, [])
q = add_filter(q, {
'not': {
'terms': {
'collection_id': authz.collections(authz.READ)
}
}
})
q = {
'query': q,
'size': 0,
'aggregations': {
'collections': {
'terms': {'field': 'collection_id', 'size': 30}
}
},
'_source': False
}
# import json
# print json.dumps(q, indent=2)
result = get_es().search(index=get_es_index(), body=q,
doc_type=TYPE_DOCUMENT)
aggs = result.get('aggregations', {}).get('collections', {})
buckets = aggs.get('buckets', [])
q = Collection.all_by_ids([b['key'] for b in buckets])
q = q.filter(Collection.creator_id != None) # noqa
objs = {o.id: o for o in q.all()}
roles = {}
for bucket in buckets:
collection = objs.get(bucket.get('key'))
if collection is None or collection.private:
continue
if collection.creator_id in roles:
roles[collection.creator_id]['total'] += bucket.get('doc_count')
else:
roles[collection.creator_id] = {
'name': collection.creator.name,
'email': collection.creator.email,
'total': bucket.get('doc_count')
}
roles = sorted(roles.values(), key=lambda r: r['total'], reverse=True)
roles = [format_total(r) for r in roles]
total = result.get('hits', {}).get('total')
return format_total({
'roles': roles,
'active': total > 0,
'total': total
})
def peek_query(args):
if not isinstance(args, MultiDict):
args = MultiDict(args)
text = args.get('q', '').strip()
q = text_query(text)
filters = parse_filters(args)
for entity in args.getlist('entity'):
filters.append(('entities.id', entity))
q = filter_query(q, filters, [])
q = add_filter(
q,
{'not': {
'terms': {
'collection_id': authz.collections(authz.READ)
}
}})
q = {
'query': q,
'size': 0,
'aggregations': {
'collections': {
'terms': {
'field': 'collection_id',
'size': 30
}
}
},
'_source': False
}
# import json
# print json.dumps(q, indent=2)
result = get_es().search(index=get_es_index(),
body=q,
doc_type=TYPE_DOCUMENT)
aggs = result.get('aggregations', {}).get('collections', {})
buckets = aggs.get('buckets', [])
q = Collection.all_by_ids([b['key'] for b in buckets])
q = q.filter(Collection.creator_id != None) # noqa
objs = {o.id: o for o in q.all()}
roles = {}
for bucket in buckets:
collection = objs.get(bucket.get('key'))
if collection is None or collection.private:
continue
if collection.creator_id in roles:
roles[collection.creator_id]['total'] += bucket.get('doc_count')
else:
roles[collection.creator_id] = {
'name': collection.creator.name,
'email': collection.creator.email,
'total': bucket.get('doc_count')
}
roles = sorted(roles.values(), key=lambda r: r['total'], reverse=True)
roles = [format_total(r) for r in roles]
total = result.get('hits', {}).get('total')
return format_total({'roles': roles, 'active': total > 0, 'total': total})
def analyze_terms(terms, seen=None):
if seen is None:
seen = set()
for term in terms:
for doc_id in query_doc_ids(text_query(term)):
if doc_id not in seen:
analyze_document.delay(doc_id)
seen.add(doc_id)
