This is an unofficial Python library for Datasift.
It currently supports compiling CSDLs via the REST API and consuming an HTTP Stream via the Streaming API.
from datasift import (
DataSiftUser,
DataSiftDefinition,
DataSiftStream,
DataSiftStreamListener
)
# Fill in account credentials
DATASIFT_USERNAME = "my_username"
DATASIFT_API_KEY = "my_api_key"
# Create user
user = DataSiftUser(DATASIFT_USERNAME, DATASIFT_API_KEY)
# Create, compile CSDL
csdl = 'interaction.type == "twitter" and '\
'(interaction.content contains "skrillex")'
definition = DataSiftDefinition(csdl, user)
definition.compile()
# Define listener
class MyListener(DataSiftStreamListener):
def on_data(self, data):
print data
# Create stream, listen
listener = MyListener()
stream = DataSiftStream(listener, definition, user)
stream.listen()
All code is released under the MIT license. Please read the LICENSE.txt file for more details.
Thanks to joshthecoder and his tweepy library for demonstrating HTTP stream consumption via httplib.