Why using catchall fields at all with eDisMax?

Question

I see no point in using catch all fields in a Solr schema if using eDisMax as query parser, since eDismax has the ability to search across multiple fields in a user friendly way like:

?q=hello wolrd&qf=title abstract

Am I missing something?

Answer 1

it has it use cases, for example if you tokenize in a different way. You can tokenize catchall field in the way to match the most (imagine you have a brand field that you are tokenizing as Keyword, and you tokenize catchall with the StandardTokenizer).

So, even if you added all your fields to the qf param, it would still find no hit in some cases, but with something like this

?q=hello wolrd&qf=title^10 abstract^8 brand^6 catchall

it would