Hi,
We have a full text index on one of our node label properties. We are getting incorrect results when querying the full text index when searching for terms with special characters.
For eg
graph.query test "CREATE
(book1:Book {name: 'catch-22'}),
(book2:Book {name: 'slaughterhouse-five'}),
(book3:Book {name: 'the book thief'}),
(book4:Book {name: 'dummy|name'})"
index creation
GRAPH.QUERY test "CALL db.idx.fulltext.createNodeIndex('Book', 'name')"
Now when I query for catch-22 or slaughterhouse-five it doesn’t return any results
graph.query test "CALL db.idx.fulltext.queryNodes('Book', '(@name:slaughterhouse\\-five)' ) YIELD node, score
RETURN node.name order by score desc"
No data to visualize. Raw information is presented below.
1) 1) "node.name"
2) (empty list or set)
It returns row when only slaughterhouse or five is specified. Could someone please help on this?
From redisearch docs, special characters are supposed to be escaped with \\ but it doesn’t seem to be working.
Is the term slaughterhouse-five tokenised and the individual words are stored separately and not the whole term? if so, can this behaviour be changed so that the whole term is indexed as well in addition to the separate words?
Thanks in advance.