def entity_documents(entity, state):
"""Try and find all documents mentioning a particular entity."""
shoulds = [{"term": {"entities.id": entity.get('id')}}]
text_queries = []
for name in entity.get('names', []):
text_queries.append(phrase_match(name, 'text'))
text_queries.append(phrase_match(name, 'text_latin'))
shoulds.append(phrase_match(name, 'title'))
shoulds.append(phrase_match(name, 'summary'))
state.highlight.append(name)
for fp in entity.get('fingerprints', []):
text_queries.append(phrase_match(fp, 'text'))
state.highlight.append(fp)
# for ident in entity.get('identifiers', []):
# text_queries.append(multi_match(ident, ['text']))
shoulds.append(child_record({"bool": {"should": text_queries}}))
# TODO: add in other entity info like phone numbers, addresses, etc. for
# ranking.
state.raw_query = {"bool": {"should": shoulds, "minimum_should_match": 1}}
# pprint(state.raw_query)
return documents_query(state)
def text_query(text):
"""Part of a query which finds a piece of text."""
if text is None or not len(text.strip()):
return match_all()
return {
"bool": {
"minimum_should_match":
1,
"should": [
meta_query_string(text),
child_record({"bool": {
"should": [text_query_string(text)]
}})
]
}
}
def text_query(text):
"""Part of a query which finds a piece of text."""
if text is None or not len(text.strip()):
return match_all()
return {
"bool": {
"minimum_should_match": 1,
"should": [
meta_query_string(text),
child_record({
"bool": {
"should": [text_query_string(text)]
}
})
]
}
}
def text_query(text):
""" Construct the part of a query which is responsible for finding a
piece of thext in the selected documents. """
if text is None or not len(text.strip()):
return match_all()
return {
"bool": {
"minimum_should_match": 1,
"should": [
meta_query_string(text),
child_record({
"bool": {
"should": [text_query_string(text)]
}
})
]
}
}
def analyze_terms(terms, seen=None):
if seen is None:
seen = set()
for term in terms:
query = {
"bool": {
"minimum_should_match": 1,
"should": [
meta_query_string(term, literal=True),
child_record({
"bool": {
"should": [text_query_string(term, literal=True)]
}
})
]
}
}
for doc_id in query_doc_ids(query):
if doc_id not in seen:
analyze_document.delay(doc_id)
seen.add(doc_id)
def analyze_terms(terms, seen=None):
if seen is None:
seen = set()
for term in terms:
term = normalize_strong(term)
query = {
"bool": {
"minimum_should_match":
1,
"should": [
meta_query_string(term),
child_record(
{"bool": {
"should": [text_query_string(term)]
}})
]
}
}
for doc_id in query_doc_ids(query):
if doc_id not in seen:
analyze_document.delay(doc_id)
seen.add(doc_id)