Skip to main content

Dependency-based Python module reloader

Project description

https://secure.travis-ci.org/jparise/python-reloader.png?branch=master

This library implements a dependency-based module reloader for Python. Unlike the builtin reload() function, this reloader will reload the requested module and all other modules that are dependent on that module.

A detailed discussion of the reloader’s implementation is available here:

http://www.indelible.org/ink/python-reloading/

Usage

The reloader works by tracking dependencies between imported modules. It must first be enabled in order to track those dependencies. The reloader has no dependency information for modules that were imported before it was enabled or after it is disabled, so you’ll probably want to enable the reloader early in your application’s startup process.

import reloader
reloader.enable()

# Import additional modules
import module1
import module2

To manually reload an imported module, pass it to the reloader’s reload() method:

import example
reloader.reload(example)

Note that you must pass the module object itself and not a string containing the module’s name. If you only have the module’s name, you can fetch the module object from the global sys.modules dictionary:

reloader.reload(sys.modules['example'])

You can also query a module’s dependencies for informational or debugging purposes:

reloader.get_dependencies(example)

You can disable the reloader’s dependency tracking at any time:

reloader.disable()

Blacklisting Modules

There may be times when you don’t want a module and its dependency hierarchy to be reloaded. The module might rarely change and be expensive to import, for example. To support these cases, you can explicit “blacklist” modules from the reloading process using the blacklist argument to enable().

reloader.enable(blacklist=['os', 'ConfigParser'])

The blacklist can be any iterable listing the fully-qualified names of modules that should be ignored. Note that blacklisted modules will still appear in the dependency graph for completeness; they will just not be reloaded.

An Interactive Example

This example demonstrates how easily the reloader can be used from the interactive Python interpretter. Imagine you have the module example.py open in a text editor, and it contains the following:

print "I am example.py"

Our interactive session starts like this:

>>> import reloader
>>> reloader.enable()
>>> import example
I am example.py

Now modify example.py in your text editor. You can then reload the example in your interactive session:

>>> reloader.reload(example)
I am the modified example.py

This is a simplistic example that doesn’t fully demonstrate the power of the reloader’s dependency-based module tracking, but it hopefully illustrates the basic usage and utility of the system.

The __reload__() Callback

If a module has a __reload__() function, it will be called with a copy of the original module’s dictionary after it has been reloaded. This provides a convenient mechanism for preserving state between reloads.

Consider a module named counts that contains the following code:

COUNTER = 0

The module’s COUNTER variable will be reset to 0 when the module is reloaded:

>>> import counts
>>> counts.COUNTER += 1
>>> counts.COUNTER
1
>>> reloader.reload(counts)
>>> counts.COUNTER += 1
1

We can preserve the value of COUNTER across reloads by adding a __reload__() function to the counts module:

def __reload__(state):
    global COUNTER
    COUNTER = state['COUNTER']

Now when we reload counts:

>>> import counts
>>> counts.COUNTER += 1
>>> counts.COUNTER
1
>>> reloader.reload(counts)
>>> counts.COUNTER += 1
>>> counts.COUNTER
2

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

reloader-0.4.tar.gz (5.1 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