Skip to main content

Distributed Pipeline Framework for Python

Project description

|Build Status| # LineUp - Distributed Pipeline Framework for Python

Lineup is a redis-based
`pipeline <http://en.wikipedia.org/wiki/Pipeline_(software)>`__
framework that turns horizontal scalling seamless.

It's currently providing parallelism through python threads and is
pretty useful for writing systems where the workers make lots of network
I/O.

It scales horizontally, so you can simply run more workers in as many
machines you want.

Installing
----------

.. code:: bash

easy_install curdling
curd install lineup

Philosophy
----------

Lineup focus in:

1. Simplicity: easy to create otherwise complex pipelines
2. Easy-scale: just run more workers and you're good.

Defining steps
--------------

Steps must always implement the method ``consume(self, instructions)``
and always call ``self.produce()`` with it's portion of work.

Example: a pipeline that downloads files
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

This is a copy of the `examples <examples>`__ folder.

.. code:: python

import re
import codecs
import requests

from lineup import Step
from lineup import Pipeline

class Download(Step):
def after_consume(self, instructions):
self.log(
"Done downloading %s",
instructions['url'],
)

def before_consume(self):
self.log("The downloader is ready")

def consume(self, instructions):
url = instructions['url']
method = instructions.get('method', 'get').lower()

http_request = getattr(requests, method)
response = http_request(url)
instructions['download'] = {
'content': response.content,
'headers': dict(response.headers),
'status_code': response.status_code,
}
self.produce(instructions)


class Cache(Step):
def after_consume(self, instructions):
self.log("Done caching %s", instructions.keys())

def before_consume(self):
self.log("The cacher is also ready")

def get_slug(self, url):
url = re.sub(r'^https?://', '', url)
url = re.sub(r'\W+', '-', url)
return url

def consume(self, instructions):
url = instructions['url']
name = self.get_slug(url)
with codecs.open(name, 'wb', 'utf-8') as fd:
fd.write(instructions['download']['content'])
fd.close()
instructions['cached_at'] = {
'filename': fd.name,
}
self.produce(instructions)



# Defining pipelines

class SimpleUrlDownloader(Pipeline):
name = 'downloader'
steps = [Download, Cache]

Command line
------------

When running from the command line, lineup will recursively try to
import all python files given as ``--working-dir`` argument, which
defaults to the relative equivalent to ``os.getcwd()``.

In other words it will find your stuff automatically in dir you run the
pipeline from, or from the ``--working-dir`` arg.

Type ``lineup --help`` for more info.

Listing available pipelines
~~~~~~~~~~~~~~~~~~~~~~~~~~~

LineUp scans the given working directory recursively for pipeline class
declarations, here is how you can list all the pipelines that lineup can
find:

.. code:: bash

lineup list pipelines

Running a pipeline in foreground
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. code:: bash

# Lineup will connect to this redis
export LINEUP_REDIS_URI='redis://0@localhost:6379'

lineup downloader run --output=rpush@example-output

.. figure:: example/run.png
:alt: example/run.png

example/run.png
Feeding a pipeline through command line
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. code:: bash

lineup downloader push {"url": "http://github.com/gabrielfalcao.keys"}

.. figure:: example/push.png
:alt: example/run.png

example/run.png
Feeding a pipeline programatically
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.. code:: python

from lineup import JSONRedisBackend
from example.pipelines import SimpleUrlDownloader

pipeline = SimpleUrlDownloader(JSONRedisBackend)
pipeline.input.put({
"url": "http://github.com/gabrielfalcao.keys"
})

Stopping all running pipelines
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

This broadcasts a "stop" signal in the backend.

.. code:: bash

lineup downloader stop

Contributing
------------

Install Dependencies
~~~~~~~~~~~~~~~~~~~~

.. code:: bash

curd install -r development.txt

Run tests
~~~~~~~~~

.. code:: bash

make test

.. |Build Status| image:: https://travis-ci.org/weedlabs/lineup.png?branch=master
:target: https://travis-ci.org/weedlabs/lineup

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for lineup, version 0.1.8
Filename, size File type Python version Upload date Hashes
Filename, size lineup-0.1.8.tar.gz (13.2 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page