skip to navigation
skip to content

crate 0.3.0

Crate client

Latest Version: 0.17.0


This is the database adapter for the crate database. Its main feature is a implementation of the Python DB API 2.0 specification.


Installing via pip

To install the crate client via pip use the following command:

$ pip install crate

To update use:

$ pip install -U crate

Installing via easy_install

If you prefer easy_install which is provided by setuptools use the following command:

$ easy_install crate

To update use:

$ easy_install -U crate


Copyright 2013 Crate-Technology GmbH

Licensed under the Apache License, Version 2.0 (the ‘License’); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an ‘AS IS’ BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Crate Client Usage

Connect to a Database

Before we can start we have to import the crate client:

>>> from crate import client

The client provides a connect() function which is used to establish a connection, the first argument is the url of the server to connect to:

>>> connection = client.connect(crate_host)

Crate is a clustered database providing high availability through replication. In order for clients to make use of this property it is recommended to specify all hosts of the cluster. This way if a server does not respond, the request is automatically routed to the next server:

>>> invalid_host = 'http://not_responding_host:4200'
>>> connection = client.connect([invalid_host, crate_host])

If no servers are given, the default one is used:

>>> connection = client.connect()
>>> connection.client._active_servers

It’s possible to define a default timeout value in seconds for all servers using the optional parameter timeout:

>>> connection = client.connect([crate_host, invalid_host], timeout=5)

Selecting Some Data

To execute a command a cursor has to be opened to perform database operations:

>>> cursor = connection.cursor()
>>> cursor.execute("SELECT name FROM locations where name = ?", ('Algol',))

To retrieve a row we can use one of the cursor’s fetch functions (described below).


fetchone() with each call returns the next row from the results:

>>> result = cursor.fetchone()
>>> pprint(result)

If no more data is available, an empty result is returned:

>>> while cursor.fetchone():
...     pass
>>> cursor.fetchone()


fetch_many() returns a list of all remaining rows, containing no more than the specified size of rows:

>>> cursor.execute("SELECT name FROM locations order by name")
>>> result = cursor.fetchmany(2)
>>> pprint(result)
[[u'Aldebaran'], [u'Algol']]

If a size is not given, the cursor’s arraysize, which defaults to ‘1’, determines the number of rows to be fetched:

>>> cursor.fetchmany()
[[u'Allosimanius Syneca']]

It’s also possible to change the cursors arraysize to an other value:

>>> cursor.arraysize = 3
>>> cursor.fetchmany()
[[u'Alpha Centauri'], [u'Altair'], [u'Argabuthon']]


fetchall() returns a list of all remaining rows:

>>> cursor.execute("SELECT name FROM locations order by name")
>>> result = cursor.fetchall()
>>> pprint(result)
 [u'Allosimanius Syneca'],
 [u'Alpha Centauri'],
 [u'Arkintoofle Minor'],
 [u'Galactic Sector QQ7 Active J Gamma'],
 [u'North West Ripple'],
 [u'Outer Eastern Rim'],

Cursor Description

The description property of the cursor returns a sequence of 7-item sequences containing the column name as first parameter. Just the name field is supported, all other fields are ‘None’:

>>> cursor.execute("SELECT * FROM locations order by name")
>>> result = cursor.fetchone()
>>> pprint(result)
 u'Max Quordlepleen claims that the only thing left ...',
 u'Star System',

>>> result = cursor.description
>>> pprint(result)
((u'date', None, None, None, None, None, None),
 (u'datetime', None, None, None, None, None, None),
 (u'description', None, None, None, None, None, None),
 (u'kind', None, None, None, None, None, None),
 (u'name', None, None, None, None, None, None),
 (u'position', None, None, None, None, None, None))

Closing the Cursor

The following command closes the cursor:

>>> cursor.close()

If a cursor is closed, it will be unusable from this point forward. If any operation is attempted to a closed cursor an ProgrammingError will be raised.

>>> cursor.execute("SELECT * FROM locations")
Traceback (most recent call last):
ProgrammingError: Cursor closed

Closing the Connection

The following command closes the connection:

>>> connection.close()

If a connection is closed, it will be unusable from this point forward. If any operation using the connection is attempted to a closed connection an ProgrammingError will be raised:

>>> cursor.execute("SELECT * FROM locations")
Traceback (most recent call last):
ProgrammingError: Connection closed

>>> cursor = connection.cursor()
Traceback (most recent call last):
ProgrammingError: Connection closed


The Crate client library provides an API to access the powerful Blob storage capabilities of the Crate server.

First, a connection object is required. This can be retrieved by importing the client module and then connecting to one or more crate server:

>>> from crate import client
>>> connection = client.connect(crate_host)

Every table which has Blob support enabled, may act as a container for Blobs. The BlobContainer object for a specific table can be retrieved like this:

>>> blob_container = connection.get_blob_container('myfiles')
>>> blob_container
<BlobContainer 'myfiles'>

The returned container object can now be used to manage the contained Blobs.

Uploading Blobs

To upload a Blob the put method can be used. This method takes a file like object and an optional SHA-1 digest as argument.

In this example we upload a file without specifying the SHA-1 digest:

>>> from tempfile import TemporaryFile
>>> f = TemporaryFile()
>>> _ = f.write(b"this is the content of the file")
>>> f.flush()

The actual put - it returns the computed SHA-1 digest upon completion:

>>> print(blob_container.put(f))


Omitting the SHA-1 digest results in one extra read of the file contents to compute the digest before the actual upload starts. Therefore, if the application already has a SHA-1 digest for the content, or is able to compute the digest on another read upfront, providing the digest will lead to better performance.

Here is another example, which provides the digest in the call:

>>> _ =
>>> blob_container.put(f, digest='6d46af79aa5113bd7e6a67fae9ab5228648d3f81')


The above call returned False because the object already existed. Since the digest is already known by the caller and it makes no sense to return it again, a boolean gets returned which indicates if the Blob was newly created or not.

Retrieving Blobs

Retrieving a blob can be done by using the get method like this:

>>> res = blob_container.get('6d46af79aa5113bd7e6a67fae9ab5228648d3f81')

The result is a generator object which returns one chunk per iteration:

>>> print(next(res))
this is the content of the file

It is also possible to check if a blob exists like this:

>>> blob_container.exists('6d46af79aa5113bd7e6a67fae9ab5228648d3f81')

Deleting Blobs

To delete a blob just call the delete method, the resulting boolean states whether a blob existed under the given digest or not:

>>> blob_container.delete('6d46af79aa5113bd7e6a67fae9ab5228648d3f81')
File Type Py Version Uploaded on Size
crate-0.3.0-py27-none-any.whl (md5) Python Wheel py27 2014-03-21 40KB
crate-0.3.0.tar.gz (md5) Source 2014-01-27 36KB