Solr 1.4 contains a little advertised feature that I think folks ought to know about, Solritas.  What a silly name, you say!  Yeah, ok, call it the boring ol’ VelocityResponseWriter if you like.  Pronunciation accessibility aside, Solritas makes Solr responses look like a real user interface.  Close enough for government work, as they say.  We’ve effectively used Solritas on a number of our recent engagements.  After bringing clients’ data into Solr for the first time, wow, they have a search box, facets, and even toggle-able inline hit score explanations and request/response dumps (when using &debugQuery=true).  And we like to toss the clients’ logo in the upper left for good measure, because we can and it’s easy.

Follow a few straightforward steps to enable Solritas on the example application which ships with Solr 1.4:

  1. Add the VelocityResponseWriter libraries directory to the example solrconfig.xml
  2. Register it as a response writer plugin, named “velocity”.
  3. Copy a templates directory –
    cp -R contrib/velocity/src/main/solr/conf/velocity example/solr/conf/velocity
  4. And while we’re at it, we’ll create a nicely named (URLs matter too) request handler mapping that defaults to the velocity writer type, and a few other basic settings that make sense for the example data.

A diff of the only configuration file touched here, solrconfig.xml, is provided at the end to give you the details of the above steps.

Now Start up Solr

cd example
java -jar start.jar

Index the example data (from another shell)

cd example/exampledocs
java -jar post.jar *.xml

and visit http://localhost:8983/solr/itas. It should looks something like this screenshot:


Hey, I want my XML back!  Ok, append &wt=xml to the URL; it’s the same results, just presented differently.

Adding debugQuery=true to a request, we get a pleasant inline view of Lucene’s explanation for the score computation of each result, see for yourself at http://localhost:8983/solr/itas?q=ipod&debugQuery=true, looks something like this, with explanations initially hidden and toggled into visibility individually by clicking the “toggle explain” link that now appears:


There’s a lot of interesting power and potential lurking within Solritas.  Here are a few notable existing capabilities:

  • Templates are searched for in the request parameters, a file system base directory, and the classpath – in that order.  This allows the requesting client to override a sub-template such as the header, on a per-request basis, plugins bundled in JAR files may contain their own templates, and the file system can override embedded templates.
  • Several useful “tools” (as Velocity calls them) are available to format numbers and dates, URL encode strings, and other practically necessary utilities to make data presentable to humans and hyperlinks.
  • Straightforward access within a template to the entire Solr request and response objects

The following diffs are the only changes needed to light up Solritas on Solr 1.4’s example application –

>   <!-- Added path to VelocityResponseWriter library and dependencies -->
>   <lib dir="../../contrib/velocity/src/main/solr/lib"/>
>   <!-- /itas mapping for Solritas view with some basic good defaults: like dismax, facet on cat -->
>   <requestHandler name="/itas">
>      <lst name="defaults">
>        <str name="v.template">browse</str>
>        <str name="title">Solritas Demonstration</str>
>        <str name="wt">velocity</str>
>        <str name="defType">dismax</str>
>        <str name="q.alt">*:*</str>
>        <str name="rows">10</str>
>        <str name="fl">*,score</str>
>        <str name="facet">on</str>
>        <str name="facet.field">cat</str>
>        <str name="facet.mincount">1</str>
>        <str name="qf">
>           text^0.5 features^1.0 name^1.2 sku^1.5 id^10.0 manu^1.1 cat^1.4
>        </str>
>      </lst>
>   </requestHandler>
>   <!-- Light up VelocityResponseWriter -->
>   <queryResponseWriter name="velocity" class="org.apache.solr.request.VelocityResponseWriter"/>

About Erik Hatcher

Read more from this author


Contact us today to learn how Lucidworks can help your team create powerful search and discovery applications for your customers and employees.