Thanks for the excellent clarification. I'll ask the sunspot guys about the
localparams issue. I have a patch that would fix it
Thanks
Brendan
On May 16, 2013, at 1:42 PM, Chris Hostetter <hossman_luc...@fucit.org> wrote:
>
> : I would then like to refer to these 'pseudo' field later in the request
> : string. I thought this would be how I'd do it:
> :
> : f.my_facet_key.facet.prefix=a_given_prefix
> ...
>
>
> that syntax was proposed in SOLR-1351 and a patch was made available, but
> it was never commited (it only supported a subset of faceting, needed more
> tests, and had unclear behavior about how the defaults where picked if
> you combined f.key.facet.foo + f.field.facet.foo + facet.foo)
>
> : I thought this would work, however it doesn't appear to. What does work is
> : if I define the prefix and mincount in the local params:
> :
> : facet.field={!ex=dt key=my_facet_key
> facet.prefix=a_given_prefix}the_facet_field
>
> Correct, SOLR-4717 added support to Solr 4.3 for specifying all of the
> facet options as local params such that that syntax would work. Given th
> way the use of Solr and localparams have evolved over the years it was
> considered a more natural and logical way to specify facet option on a per
> field or per key basis.
>
> : Is this expected? I'm also using sunspot and they construct the queries
> : with keys as in my first example, i.e. facet.field={!ex=dt
> : key=my_facet_key}the_facet_field&f.my_facet_key.facet.prefix=a_given_prefix
>
> I can't comment on that ... i'm not sure why sunspot would assume that
> behavior would work (unless someone looked at SOLR-1351 once upon a time
> and assumed that would definitely be official at some point)
>
> -Hoss
