Skip to main content

Distributed component model for Python.

Project description

PyCOM is simple and easy-to-use distributed component model written in Python. PyCOM makes different parts of your network application isolated and independent, while allowing easy and straightforward interaction between them. PyCOM may be seen as an easy and lightweight web services replacement.

Ideology highlights:

  • Non-intrusive design without black magic and lots of auto-generated code

  • Effective, easy-to-implement and portable protocol

  • Support for binary attachments of any size

  • Support for stateful services via HTTP-alike sessions

  • Low level enough to build your own frameworks

  • … and still simple enough to be used as is

  • Python 2 and Python 3 support out-of-box

Technical highlights:

  • Pluggable protocol support (default is JSON over ZeroMQ aka zerojson)

  • Easily extensible core library

  • Does not required special “container” software

  • Introspection support for services

  • Comprehensive test suite and documentation

  • Free software (new BSD license)

There is ongoing effort to create a C++ client library for PyCOM: https://bitbucket.org/divius/libpycom

Main concepts

With PyCOM you build your application as a number of services, each running in it’s own process (or even on it’s own computer). You maintain a PyCOM nameserver for finding services by their names (by the way, nameserver itself is a service).

Services provide interfaces, i.e. a named way of interacting with service. They are somewhat similar to interfaces in e.g. Java, but note that PyCOM does not perform any checks on interfaces. Interface usually has some amount of methods.

Services are identified by path with parts separated by slashes, e.g. /com/foo/group/service.

Interfaces are identified by name with parts separated by dots, e.g. com.foo.my-interface.

Examples

Service example (module service1.module1):

import pycom

@pycom.interface("com.foo.example")
class Query(pycom.Service):

    @pycom.method("create")
    def method_create(self, request):
        return {"field1" : request.args}

pycom.main()

Example command line for running this service (provided nameserver is running on 192.168.10.1:2012):

python -m pycom -a tcp://192.168.10.2:2013 -n tcp://192.168.10.1:2012 service1.module1

Example client code for this service:

import pycom

pycom.configure(nameserver="192.168.10.1:2012")

query = pycom.locate("com.foo.example")
print query.invoke("create", 42)
# Prints {"field1" : 42}
print query.introspect()
# Prints a lot of introspection information

Quick start

To test (logs will be saved in test.log):

$ python test.py

To build HTML documentation (requires Sphinx):

$ python setup.py build_sphinx
$ <your-browser> build/sphinx/html/index.html

To install:

$ python setup.py install

or via pip:

$ pip install pycom

Do not forget to read about known issues in the current version: http://packages.python.org/pycom/status.html#known-issues

Support

PyCOM repository and issue tracker are hosted on BitBucket.

Download releases: http://pypi.python.org/pypi/pycom#downloads

Latest source code: https://bitbucket.org/divius/pycom/overview

Report bugs: https://bitbucket.org/divius/pycom/issues

Read documentation: http://www.pycom.org

Enjoy =)

Project details


Download files

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

Source Distribution

pycom-0.4.1.tar.gz (36.8 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page