def search(self,
search_term,
room=None,
max_results=10,
order_by_recent=False):
# type (str, str, int, bool) -> List[int, int]
"""Search for events in the index.
Returns the score and the column id for the event.
"""
queryparser = tantivy.QueryParser.for_index(
self._index,
[
self.body_field, self.name_field, self.topic_field,
self.room_field
],
)
# This currently supports only a single room since the query parser
# doesn't seem to work with multiple room fields here.
if room:
query_string = "{} AND room:{}".format(search_term,
sanitize_room_id(room))
else:
query_string = search_term
try:
query = queryparser.parse_query(query_string)
except ValueError:
raise InvalidQueryError(f"Invalid search term: {search_term}")
if order_by_recent:
collector = tantivy.TopDocs(
max_results, order_by_field=self.timestamp_field)
else:
collector = tantivy.TopDocs(max_results)
result = self._searcher.search(query, collector)
retrieved_result = []
for score, doc_address in result:
doc = self._searcher.doc(doc_address)
column = doc.get_first(self.column_field)
retrieved_result.append((score, column))
return retrieved_result
def test_simple_search_in_dir(self, dir_index):
_, index = dir_index
query = index.parse_query("sea whale", ["title", "body"])
top_docs = tantivy.TopDocs(10)
result = index.searcher().search(query, top_docs)
assert len(result) == 1
def test_simple_search_after_reuse(self, dir_index):
index_dir, _ = dir_index
index = Index(schema(), str(index_dir))
query = index.parse_query("sea whale", ["title", "body"])
top_docs = tantivy.TopDocs(10)
result = index.searcher().search(query, top_docs)
assert len(result) == 1
def test_simple_search_in_ram(self, ram_index):
index = ram_index
query = index.parse_query("sea whale", ["title", "body"])
top_docs = tantivy.TopDocs(10)
result = index.searcher().search(query, top_docs)
assert len(result) == 1
_, doc_address = result[0]
searched_doc = index.searcher().doc(doc_address)
assert searched_doc["title"] == ["The Old Man and the Sea"]
def test_and_query(self, ram_index):
index = ram_index
query = index.parse_query("title:men AND body:summer", default_field_names=["title", "body"])
# look for an intersection of documents
top_docs = tantivy.TopDocs(10)
searcher = index.searcher()
result = searcher.search(query, top_docs)
# summer isn't present
assert len(result) == 0
query = index.parse_query("title:men AND body:winter", ["title", "body"])
result = searcher.search(query, top_docs)
assert len(result) == 1
def test_delete_update(self, ram_index):
query = ram_index.parse_query("Frankenstein", ["title"])
top_docs = tantivy.TopDocs(10)
result = ram_index.searcher().search(query, top_docs)
assert len(result) == 1
writer = ram_index.writer()
with pytest.raises(ValueError):
writer.delete_documents("fake_field", "frankenstein")
with pytest.raises(ValueError):
writer.delete_documents("title", b"frankenstein")
writer.delete_documents("title", "frankenstein")
writer.commit()
ram_index.reload()
result = ram_index.searcher().search(query, top_docs)
assert len(result) == 0