I was always under the impression that a search component couldn't
modify the output of a previous search component. If it can, then the
highlight component could add its results to the output of the query
component, and we're done.

Upayavira (who sees the confusion on people's faces often when he shows
them Solr highlighting)

On Fri, 27 May 2016, at 06:27 PM, Erick Erickson wrote:
> Maybe you'd be better off using a custom search component.
> instead of a doc transformer. The intent of a doc transformer
> is, as you've discovered, working on single docs at a time. You
> want to manipulate the whole response which seems to fit more
> naturally into a search component. Make sure to put it after
> the highlight component (i.e. last-components).
> 
> Best,
> Erick
> 
> On Fri, May 27, 2016 at 6:55 AM, Upayavira <u...@odoko.co.uk> wrote:
> > In a JSON response, we get this:
> >
> > {
> >   "responseHeader": {...},
> >   "response": { "docs": [...] },
> >   "highlighting": {...}
> >   ...
> > }
> >
> > I'm assuming that the getProcessedDocuments call would give me the docs:
> > {} element, whereas I'm after the whole response so I can retrieve the
> > "highlighting" element.
> >
> > Make sense?
> >
> > On Fri, 27 May 2016, at 02:45 PM, Mikhail Khludnev wrote:
> >> Upayavira,
> >>
> >> It's not clear what do you mean in "results themselves", perhaps you mean
> >> SolrDocuments ?
> >>
> >> public abstract class ResultContext {
> >>  ..
> >>   public Iterator<SolrDocument> getProcessedDocuments() {
> >>     return new DocsStreamer(this);
> >>   }
> >>
> >> On Fri, May 27, 2016 at 4:15 PM, Upayavira <u...@odoko.co.uk> wrote:
> >>
> >> > Yes, I've seen that. I can see the getDocList() method will presumably
> >> > give me the results themselves, but I need the full response so I can
> >> > get the highlighting details, but I can't see them anywhere.
> >> >
> >> > On Thu, 26 May 2016, at 09:39 PM, Mikhail Khludnev wrote:
> >> > > public abstract class ResultContext {
> >> > >
> >> > >  /// here are all results
> >> > >   public abstract DocList getDocList();
> >> > >
> >> > >   public abstract ReturnFields getReturnFields();
> >> > >
> >> > >   public abstract SolrIndexSearcher getSearcher();
> >> > >
> >> > >   public abstract Query getQuery();
> >> > >
> >> > >   public abstract SolrQueryRequest getRequest();
> >> > >
> >> > > On Thu, May 26, 2016 at 11:25 PM, Upayavira <u...@odoko.co.uk> wrote:
> >> > >
> >> > > > Hi Mikhail,
> >> > > >
> >> > > > Is there really? If I look at ResultContext, I see it is an abstract
> >> > > > class, completed by BasicResultContext. I don't see any context 
> >> > > > method
> >> > > > there. I can see a getContext() on SolrQueryRequest which just 
> >> > > > returns
> >> > a
> >> > > > hashmap. Will I find the response in there? Is that what you are
> >> > > > suggesting?
> >> > > >
> >> > > > Upayavira
> >> > > >
> >> > > > On Thu, 26 May 2016, at 06:28 PM, Mikhail Khludnev wrote:
> >> > > > > Hello,
> >> > > > >
> >> > > > > There is a protected ResultContext field named context.
> >> > > > >
> >> > > > > On Thu, May 26, 2016 at 5:31 PM, Upayavira <u...@odoko.co.uk> 
> >> > > > > wrote:
> >> > > > >
> >> > > > > > Looking at the code for a sample DocTransformer, it seems that a
> >> > > > > > DocTransformer only has access to the document itself, not to the
> >> > whole
> >> > > > > > results. Because of this, it isn't possible to use a
> >> > DocTransformer to
> >> > > > > > merge, for example, the highlighting results into the main
> >> > document.
> >> > > > > >
> >> > > > > > Am I missing something?
> >> > > > > >
> >> > > > > > Upayavira
> >> > > > > >
> >> > > > >
> >> > > > >
> >> > > > >
> >> > > > > --
> >> > > > > Sincerely yours
> >> > > > > Mikhail Khludnev
> >> > > > > Principal Engineer,
> >> > > > > Grid Dynamics
> >> > > > >
> >> > > > > <http://www.griddynamics.com>
> >> > > > > <mkhlud...@griddynamics.com>
> >> > > >
> >> > >
> >> > >
> >> > >
> >> > > --
> >> > > Sincerely yours
> >> > > Mikhail Khludnev
> >> > > Principal Engineer,
> >> > > Grid Dynamics
> >> > >
> >> > > <http://www.griddynamics.com>
> >> > > <mkhlud...@griddynamics.com>
> >> >
> >>
> >>
> >>
> >> --
> >> Sincerely yours
> >> Mikhail Khludnev
> >> Principal Engineer,
> >> Grid Dynamics
> >>
> >> <http://www.griddynamics.com>
> >> <mkhlud...@griddynamics.com>

Reply via email to