pysolr
is a lightweight Python wrapper for Apache Solr. It provides an
interface that queries the server and returns results based on the query.
- Basic operations such as selecting, updating & deleting.
- Index optimization.
- "More Like This" support (if set up in Solr).
- Spelling correction (if set up in Solr).
- Timeout support.
- SolrCloud awareness
- Python 2.7 - 3.6
- Requests 2.9.1+
- Optional -
simplejson
- Optional -
kazoo
for SolrCloud mode
pysolr is on PyPI:
$ pip install pysolr
Or if you want to install directly from the repository: python setup.py install
, or drop the pysolr.py
file anywhere on your PYTHONPATH
.
Basic usage looks like:
# If on Python 2.X
from __future__ import print_function
import pysolr
# Setup a Solr instance. The timeout is optional.
solr = pysolr.Solr('http://localhost:8983/solr/', timeout=10, auth=<type of authentication>)
# Do a health check.
solr.ping()
# How you'd index data.
solr.add([
{
"id": "doc_1",
"title": "A test document",
},
{
"id": "doc_2",
"title": "The Banana: Tasty or Dangerous?",
"_doc": [
{ "id": "child_doc_1", "title": "peel" },
{ "id": "child_doc_2", "title": "seed" },
]
},
])
# Note that the add method has commit=True by default, so this is
# immediately committed to your index.
# You can index a parent/child document relationship by
# associating a list of child documents with the special key '_doc'. This
# is helpful for queries that join together conditions on children and parent
# documents.
# Later, searching is easy. In the simple case, just a plain Lucene-style
# query is fine.
results = solr.search('bananas')
# The ``Results`` object stores total results found, by default the top
# ten most relevant results and any additional data like
# facets/highlighting/spelling/etc.
print("Saw {0} result(s).".format(len(results)))
# Just loop over it to access the results.
for result in results:
print("The title is '{0}'.".format(result['title']))
# For a more advanced query, say involving highlighting, you can pass
# additional options to Solr.
results = solr.search('bananas', **{
'hl': 'true',
'hl.fragsize': 10,
})
# You can also perform More Like This searches, if your Solr is configured
# correctly.
similar = solr.more_like_this(q='id:doc_2', mltfl='text')
# Finally, you can delete either individual documents,
solr.delete(id='doc_1')
# also in batches...
solr.delete(id=['doc_1', 'doc_2'])
# ...or all documents.
solr.delete(q='*:*')
# For SolrCloud mode, initialize your Solr like this:
zookeeper = pysolr.ZooKeeper("zkhost1:2181,zkhost2:2181,zkhost3:2181")
solr = pysolr.SolrCloud(zookeeper, "collection1", auth=<type of authentication>)
Simply point the URL to the index core:
# Setup a Solr instance. The timeout is optional.
solr = pysolr.Solr('http://localhost:8983/solr/core_0/', timeout=10)
# Setup a Solr instance. The trailing slash is optional.
solr = pysolr.Solr('http://localhost:8983/solr/core_0/', search_handler='/autocomplete', use_qt_param=False)
If use_qt_param
is True
it is essential that the name of the handler is exactly what is configured
in solrconfig.xml
, including the leading slash if any (though with the qt
parameter a leading slash is not
a requirement by SOLR). If use_qt_param
is False
(default), the leading and trailing slashes can be
omitted.
If search_handler
is not specified, pysolr will default to /select
.
The handlers for MoreLikeThis, Update, Terms etc. all default to the values set in the solrconfig.xml
SOLR ships
with: mlt
, update
, terms
etc. The specific methods of pysolr's Solr
class (like more_like_this
,
suggest_terms
etc.) allow for a kwarg handler
to override that value. This includes the search
method.
Setting a handler in search
explicitly overrides the search_handler
setting (if any).
# Setup a Solr instance in a kerborized enviornment
from requests_kerberos import HTTPKerberosAuth, OPTIONAL
kerberos_auth = HTTPKerberosAuth(mutual_authentication=OPTIONAL, sanitize_mutual_error_response=False)
solr = pysolr.Solr('http://localhost:8983/solr/', auth=kerberos_auth)
# Setup a CloudSolr instance in a kerborized environment
from requests_kerberos import HTTPKerberosAuth, OPTIONAL
kerberos_auth = HTTPKerberosAuth(mutual_authentication=OPTIONAL, sanitize_mutual_error_response=False)
zookeeper = pysolr.ZooKeeper("zkhost1:2181/solr, zkhost2:2181,...,zkhostN:2181")
solr = pysolr.SolrCloud(zookeeper, "collection", auth=kerberos_auth)
# Setup a Solr instance in an https environment
solr = pysolr.Solr('http://localhost:8983/solr/', verify=path/to/cert.pem)
# Setup a CloudSolr instance in a kerborized environment
zookeeper = pysolr.ZooKeeper("zkhost1:2181/solr, zkhost2:2181,...,zkhostN:2181")
solr = pysolr.SolrCloud(zookeeper, "collection", verify=path/to/cert.perm)
# Setup a Solr instance. The trailing slash is optional.
# All request to solr will result in a commit
solr = pysolr.Solr('http://localhost:8983/solr/core_0/', search_handler='/autocomplete', always_commit=True)
always_commit
signals to the Solr object to either commit or not commit by default for any solr request.
Be sure to change this to True if you are upgrading from a version where the default policy was alway commit by default.
Functions like add
and delete
also still provide a way to override the default by passing the commit
kwarg.
It is generally good practice to limit the amount of commits to solr. Excessive commits risk opening too many searcher or using too many system resources.
pysolr
is licensed under the New BSD license.
The run-tests.py
script will automatically perform the steps below and is recommended for testing by
default unless you need more control.
Downloading, configuring and running Solr 4 looks like this:
./start-solr-test-server.sh
The test suite requires the unittest2 library:
Python 2:
python -m unittest2 tests
Python 3:
python3 -m unittest tests