A simple framework that makes it easy to work with data streams in Python.
Spout is a small and simple framework that makes it easy to work with data streams in Python. In particular, Spout was designed with the processing and consumption of live data sources in mind.
At the heart of Spout is the concept of a Stream (which is defined in an abstract Stream class). This defines the basic operations that can be performed upon a data stream:
The items in one stream can be “mapped” to another stream. This is done by applying a supplied Function to each item in the input stream, to produce another output stream.
The items in a stream can be “filtered”, so that the resultant stream only contains items that match a given criteria. This is done by using a supplied Predicate to test each item in the input stream, and copies it to the output stream if it passes the test criteria.
The items in a stream are used in some calculations or functionality that provides no further output to the stream. This is done by applying the supplied Operation to each item in the stream.
To use Spout, you first need to create an input data stream. A data stream is simply an instantiation of a Stream or any of its children (which can be found in the streams.py file). The Stream class has been specifically designed so that it is easy to extend and wrap around currently existing data sources that you might have, such as files or databases.
Some existing examples of stream data sources can by found in sources.py.
For example, to create a Stream out of the lines in a plain text file:
from spout.sources import FileInputStream s = FileInputStream(“test.txt”)
Now that you have your data in a stream, you simply have to process it! This can be done by creating and using your own Functions, Predicates or Operations (see above).
For example, to print out all the lines in a text file that start with a digit, but with the digit stripped, we can create our own Predicate and Function and pass these to the .filter() and .map() functions:
from spout.sources import FileInputStream from spout.structs import Function, Predicate from spout.outputs import PrintOperation
- class StartsWithDigit(Predicate):
- def test(self, obj):
- return obj.is_digit()
- class StripFirstChar(Function):
- def apply(self, input):
- return input[1:]
s = FileInputStream(“test.txt”) s .filter(StartsWithDigit()) .map(StripFirstChar()) .for_each(PrintOperation())
Spout is available in the Python Package Index (PyPI), and so the easiest way to install it is through pip:
$ pip install spout
However, it is also possible to install the repository from the source, through the setup.py utility:
$ python setup.py install
The inspiration for Spout’s fluent interface came largely from the OpenIMAJ streaming framework. http://www.openimaj.org/