Solr: best documented, easy to use, stable Python

问题:

I want to use Lucene Solr in Python. There seems to be multiple APIs for this purpose. They seem to suffer dependency hell and stability issues, and Solr doesnt ship with python bindings anymore. And I cant find any documentation for the user who is not familiar with Solr.

I am leaning on Sunburnt over pysolr and solrpy for being the most mature. Is my evaluation correct?

Please recommend stable python bindings for Solr with good standalone (does not presuppose Solr knowledge) documentation. Or should I learn how to interact with the "servlet" and make use of the existing documentation?

回答1:

Always know Solr independently of any client library like Sunburnt, pysolr, solrpy, etc. Just as you have to know relational databases before using any ORM. Moreover, no ORM documentation will teach you relational databases, and no one would expect it to do so.

回答2:

Haystack has an awesome API and supports the backends: Solr, Whoosh and Xapian.

I have successfully used Haystack several places, in development with Whoosh backend and in production using the Solr backend. it has been pretty smooth.

回答3:

I normally use mysolr in my projects: http://mysolr.redtuna.org/

回答4:

I ended up using sunburnt, which seems very feature-rich and reading a lot of their and Solr's documentation.