Another path to follow could be to design a specific collection(index) for
the auto-suggestion.
In there you can define the analysis chain as you like ( for example using
edge-ngram filtering on top of tokenisation) to provide infix
autocompletion.
Then you can play with your queries as you like an
Yes, but it is better than nothing. Don’t let the unavailable perfect solution
keep you from implementing the available good solution.
If you want to easily use fuzzy search with edismax, check out the patch
submitted with SOLR-629.
wunder
Walter Underwood
wun...@wunderwood.org
http://observer.
I thought about that - but that does not solve the issue. If the user types
in a misspelled word, and that word is somewhere in the middle of the
field, that result will not be returned.
2017-06-28 19:00 GMT+02:00 Walter Underwood :
> I set up two suggesters, one fuzzy and one analyzing infix. Th
I set up two suggesters, one fuzzy and one analyzing infix. That gives two sets
of suggestions, so the client code has to merge them into one list and toss
duplicates.
They use the same weights, so I can keep the top weighted suggestions.
wunder
Walter Underwood
wun...@wunderwood.org
http://obs
