lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Scott Dawson <sc.e.daw...@gmail.com>
Subject Re: Custom JSON
Date Thu, 16 Oct 2014 19:48:49 GMT
Noble,
Thanks. You're right. I had some things incorrectly configured but now I
can put structured JSON into Solr using the out-of-the-box solrconfig.xml.

One additional question: Is there any way to query Solr and receive the
original structured JSON document in response? Or does the flattening
process that happens during indexing obliterate the original structure with
no way to reconstruct it?

Thanks again,
Scott

On Thu, Oct 16, 2014 at 2:10 PM, Noble Paul <noble.paul@gmail.com> wrote:

> The end point  /update/json/docs is enabled implicitly in Solr irrespective
> of the solrconfig.xml
> In schemaless the fields are created automatically by solr.
>
> If you have all the fields created in your schema.xml it will work .
>
> if you  need an id field please use a copy field to create one
>
> --Noble
>
> On Thu, Oct 16, 2014 at 8:42 PM, Scott Dawson <sc.e.dawson@gmail.com>
> wrote:
>
> > Hello,
> > I'm trying to use the new custom JSON feature described in
> > https://issues.apache.org/jira/browse/SOLR-6304. I'm running Solr
> 4.10.1.
> > It seems that the new feature, or more specifically, the
> /update/json/docs
> > endpoint is not enabled out-of-the-box except in the schema-less example.
> > Is there some dependence of the feature on schemaless mode? I've tried
> > pulling the endpoint definition and related pieces of the
> > example-schemaless solrconfig.xml and adding those to the "standard"
> > solrconfig.xml in the main example but I've run into a cascade of issues.
> > Right now I'm getting a "This IndexSchema is not mutable" exception when
> I
> > try to post to the /update/json/docs endpoint.
> >
> > My real question is -- what's the easiest way to get this feature up and
> > running quickly and is this documented somewhere? I'm trying to do a
> quick
> > proof-of-concept to verify that we can move from our current flat JSON
> > ingestion to a more natural use of structured JSON.
> >
> > Thanks,
> > Scott Dawson
> >
>
>
>
> --
> -----------------------------------------------------
> Noble Paul
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message