This package has support for all the search endpoints, as well as a suite of functions for managing a Solr database, including adding and deleting documents.

Important search functions

  • solr_search() - General search, only returns documents

  • solr_all() - General search, including all non-documents in addition to documents: facets, highlights, groups, mlt, stats.

  • solr_facet() - Faceting only (w/o general search)

  • solr_highlight() - Highlighting only (w/o general search)

  • solr_mlt() - More like this (w/o general search)

  • solr_group() - Group search (w/o general search)

  • solr_stats() - Stats search (w/o general search)

Important Solr management functions

Vignettes

See the vignettes for help browseVignettes(package = "solrium")

Performance

v0.2 and above of this package will have wt=csv as the default. This should give significant performance improvement over the previous default of wt=json, which pulled down json, parsed to an R list, then to a data.frame. With wt=csv, we pull down csv, and read that in directly to a data.frame.

The http library we use, crul, sets gzip compression header by default. As long as compression is used server side, you're good to go on compression, which should be a good peformance boost. See https://wiki.apache.org/solr/SolrPerformanceFactors#Query_Response_Compression for notes on how to enable compression.

There are other notes about Solr performance at https://wiki.apache.org/solr/SolrPerformanceFactors that can be used server side/in your Solr config, but aren't things to tune here in this R client.

Let us know if there's any further performance improvements we can make.

Author

Scott Chamberlain [email protected]