att/rcloud.solr

Search controller

Closed this issue · 1 comments

Big feature.

Add a controller to sit inbetween searches and solr. It will grab a large number of search results and control:

  • create test second solr instance #38
  • store results from all sources #39
  • sort results #40
  • implement paging #43
  • invalidate cache
  • sort dates and chars #46

Things to consider:

This is ultimately all about combining multiple sources, so bare that in mind when designing this. One source is main, there may be many alternative sources.

In the first iteration we're going to go for the following behaviour:

page: 0 indicates a new search. cache is discarded and new searches made against solr.
page: >0 indicates paging through an old search. Cache doesn't get invalidated.

Need to think about circumstances whereby cache should be invalidated (and trigger a new search) when page>0.