Skip to content


Subversion checkout URL

You can clone with
Download ZIP
A Xapian backend for Haystack

Added commit argument to update

update has now the same signature as in Django-Haystack.
latest commit a3a3a4e7cf
@viorels viorels authored jorgecarleitao committed
Failed to load latest commit information.
tests Fixed tests for Django 1.7 and 1.8.
.gitignore Updated .gitignore to ignore MANIFEST
.travis.yml Fixed tests in Travis.
AUTHORS Updated narrow_queries in backend search to AND queries rather than O…
LICENSE Updated to include README.rst
README.rst Updated README and requirements with latest updates.
requirements.txt Updated README and requirements with latest updates. Updated narrow_queries in backend search to AND queries rather than O… Added commit argument to update


Xapian backend for Django-Haystack

Xapian-haystack is a backend of Django-Haystack for the Xapian search engine. Thanks for checking it out.

You can find more information about Xapian here.


Xapian-Haystack provides all the standard features of Haystack:

  • Weighting
  • Faceted search (date, query, etc.)
  • Sorting
  • Spelling suggestions
  • EdgeNGram and Ngram (for autocomplete)


  • Python 2.4+ (Python 3.3 not support yet).
  • Django 1.6+
  • Django-Haystack 2.0.X
  • Xapian 1.0.13+

In particular, we build this backend on Travis using:

  • Python 2.7.6
  • Django 1.6, 1.7 and 1.8
  • Django-Haystack (master)
  • Xapian 1.2.8 (libxapian22)


First you need to install Xapian in your machine. We recommend installing it on the virtual environment using this gist: activate the virtual environment and run the script.

You can test if the installation was successful by running:

python -c "import xapian"

Finally, install Xapian-Haystack by running:

pip install git+


Xapian is configured as other backends of Haystack. You have to define the connection to the database, which is done to a path to a directory, e.g:

    'default': {
        'ENGINE': 'xapian_backend.XapianEngine',
        'PATH': os.path.join(os.path.dirname(__file__), 'xapian_index')

The backend has the following optional settings:

  • HAYSTACK_XAPIAN_LANGUAGE: the stemming language; the default is english and the list of available languages can be found here.
  • HAYSTACK_XAPIAN_WEIGHTING_SCHEME: a tuple with parameters to be passed to the weighting scheme BM25. By default, it uses the same parameters as Xapian recommends; this setting allows you to change them.
  • HAYSTACK_XAPIAN_FLAGS: the options used to parse AutoQueries; the default is FLAG_PHRASE | FLAG_BOOLEAN | FLAG_LOVEHATE | FLAG_WILDCARD | FLAG_PURE_NOT See here for more information on what they mean.
  • HAYSTACK_XAPIAN_STEMMING_STRATEGY: This option lets you chose the stemming strategy used by Xapian. Possible values are STEM_NONE, STEM_SOME, STEM_ALL, STEM_ALL_Z, where STEM_SOME is the default. See here for more information about the different strategies.


Xapian-Haystack has a test suite in continuous deployment in Travis. The script .travis.yml contains the steps required to run the test suite.


The source code can be found in github.


Xapian-Haystack is maintained by Jorge C. Leitão; David Sauve was the main contributor of Xapian-Haystack and Xapian-Haystack was originally funded by Trapeze. ANtlord implemented support for EdgeNgram and Ngram.


Xapian-haystack is free software licenced under GNU General Public Licence v2 and Copyright (c) 2009, 2010, 2011, 2012 David Sauve, 2009, 2010 Trapeze, 2014 Jorge C. Leitão. It may be redistributed under the terms specified in the LICENSE file.

Questions, Comments, Concerns:

Feel free to open an issue here or pull request your work.

You can ask questions on the django-haystack mailing list or in the irc #haystack.

Something went wrong with that request. Please try again.