Comments on: Sunburnt: a python-solr interface

By: Royce

Royce — Tue, 18 Oct 2011 18:51:16 +0000

Is it possible to add python dictionaries that contain sub-dictionaries? For example,

dict(author=dict(book=dict(name=’1984′)))

If possible, how do you represent this type of document in a Solr schema?

-Royce

By: Casper

Casper — Wed, 26 Jan 2011 20:50:44 +0000

The tweaking involves getting rid of the C dependencies, lxml is replaced by xml.etree.ElementTree, urllib is replaced by urllib2, So to be more specific, i works on appengine, but since my usage is limited to specific functionality, mainly “select”, it works for me just fine. I haven’t dealt with the mxdatetime yet, but i guess it’s no problem finding or coding a replacement. Sorry if my comment was a bit misleading.

By: Gora Khargosh

Gora Khargosh — Fri, 21 Jan 2011 00:11:31 +0000

I believe none of the dependencies of sunburnt
work with Google App Engine:
1. lxml — uses Python/C API, so is immediately ruled out.
2. mxDatetime — uses Python/C API, so is immediately ruled out.
3. pytz — is available as gaepytz, but I’m not sure whether it would work as well as pytz itself.

I’m also wondering how Casper got sunburnt to work on App Engine. Do let people know if it is at all possible.

Thanks.
Gora Khargosh.

By: Toby

Toby — Tue, 30 Nov 2010 17:14:09 +0000

Glad to hear it was useful – can you share what you needed to do to make it work on AppEngine?

I don’t use wt=json, because I have to deal with the overhead of XML when sending messages to Solr anyway, so it’s easier for me to not have to worry about two serialization formats!

By: Casper

Casper — Tue, 30 Nov 2010 10:45:06 +0000

Thanks for sharing, with at little twaeking it seems to work on google app engine as well. By the way, may i recommend using the query param wt=json when selecting from solr just to avoid the overhead of xml.

By: claudio

claudio — Tue, 29 Jun 2010 10:32:40 +0000

great work. that’s exactly what i needed.
it would indeed make more sense to take the schema directly from the solr http.

By: Toby White

Toby White — Thu, 18 Feb 2010 10:22:30 +0000

I wrote a separate post comparing the existing Python interfaces, at http://eaddrinu.se/blog/2010/sunburnt.html – in short, there’s nothing wrong with the existing clients, but they didn’t expose the interface I wanted.

By: John

John — Thu, 18 Feb 2010 07:37:06 +0000

Interesting! How does this differ from the Python and other clients at http://www.lucidimagination.com/search/document/CDRG_ch11_11.5 ? Maybe you should contrib or post on the apache wiki?

By: Eric Pugh

Eric Pugh — Mon, 08 Feb 2010 21:41:13 +0000

Love to see a new Python Solr library. I’ve been thinking that have easier access to what is encoded in the schema.xml would be nice for many different types of projects.