Ways to send binary/structured documents to SOLR?

Question

I am using SOLR's ExtractingRequestHandler to ingest the text of documents.

The examples in the documentation all use curl to stream documents, like so:

curl 'http://.../extract?literal.id=doc1&commit=true' -F "[email protected]"

That works just fine, but there is this note:

using "curl" or other command line tools to post documents to Solr is nice for testing, but not the recommended update method for best performance.

The section of the documentation on sending documents to SOLR is a TODO. What are some of the alternative methods, and how are they better?

Answer 1

Typically you use a library in your client language that does all communication with Solr (e.g. for Java use SolrJ, for PHP SolPHP (there are others), etc).

Libraries provide things such as ease of use, error handling, proper encapsulation, etc.