Add support for levenshtein string function about kuzu HOT 3 CLOSED

At a minimum, a comparison on the fly should work fine without the need for indexing and maintenance. We should have an operator in the expression evaluator that returns the distance between two strings.

from kuzu.

I'm not sure that's the most efficient way to do it. Does Neo4j or Neptune anybody else do something like this? I know Neo4j implements a full-text index based on Lucene and BM25, which considers subword features, allowing for typos. Surely using an existing library to create an index upfront on specific properties that would be searched via text input queries would be more efficient?

from kuzu.

I found that duckdb has an implementation of levenshtein(s1, s2): https://duckdb.org/docs/sql/functions/char

from kuzu.