	Data Structures	Source Location
		`lang/python/`

Caution: the Architecture Guide is not updated in lockstep with the code base and is not necessarily correct or complete for any specific release.

WiredTiger includes a Python API, which is useful for scripting and experimentation.

The following simple example show some highlights of the API:

import os
from wiredtiger import wiredtiger_open
 
# Connect to the database and open a session
os.system('rm -rf WT_HOME')
os.makedirs('WT_HOME')
 
conn = wiredtiger_open('WT_HOME', 'create')
session = conn.open_session()
 
# Create a simple table
session.create('table:T', 'key_format=S,value_format=S')
 
# Open a cursor and insert a record
cursor = session.open_cursor('table:T', None)
 
cursor.set_key('key1')
cursor.set_value('value1')
cursor.insert()
 
# Iterate through the records
cursor.reset()
for key, value in cursor:
    print('Got record: %s : %s' % (key, value))
 
conn.close()

The API is implemented using SWIG. SWIG imports the <wiredtiger.h> file, and from it creates the classes and C support files needed to allow Python to import the WiredTiger library.

Not every facility that is present in the C API is available in Python. In particular, setting up custom collators, extractors, compressors, encryptors and file system implementations must all be done within the C framework.