Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Web pages capture using Django & CasperJS
Python JavaScript HTML
Latest commit 19a19ba @leplatrem leplatrem Merge pull request #37 from nikolaik/python3
Python3 and Django 1.8 support


django-screamshot is a very naive implementation of Web pages capture with CasperJS (aaAAaah!, phantomjs:))

(See the issues pages for more details about what remains to be done.)

Checkout screamshotter, the simplest Django project powered by django-screamshot.


First make sure you have either the casperjs or phantomjs command in your PATH, using related installation instructions:

Then install the egg :

pip install django-screamshot


  • Add screamshot to your INSTALLED_APPS

As a screenshot Web API

Add it to your project URLs :

urlpatterns = patterns('',
    url(r'^capture/$',  include('screamshot.urls', namespace='screamshot', app_name='screamshot')),

You can then obtain a screenshot using the following GET parameters :

The website URL to capture. This can be a fully qualified URL, or the name of a URL to be reversed in your Django project. Note: do not forget to encode the url.
CSS3 selector. It will restrict the screenshot to the selected element.
HTTP method to be used (default: GET)
Viewport width (default: 1400)
Viewport height (default: 900)
HTTP data to be posted (default: {})
CSS3 selector. The screenshot will be performed only once this selector is satisfied. Typical usage: if your page contains a heavy javascript processing, you can add a CSS class on an element when the processing is finished to make sure the screenshot will get the page properly rendered.
If render=html, it will return an HTML page containing the image and where the print diaplo box will be automatically opened.
Resize image (width x height, e.g: 500x500), need install PIL or Pillow.
If true, then resulting image is cropped to match specified size.

For example : http://server/capture/?url=

As a template tag

You can include screenshots in your pages using a template tag. It will perform the capture and return the base64 version of the resulting image.

This is very useful if you don't want to expose the capture API publicly.

{% base64capture URL SELECTOR %}

For example, in a SVG template :

{% load screamshot %}

   xlink:href="data:{% base64capture "company:map" "#map" %}"
   width="640" />

If you run the capture server on a different instance, you can specify the root url for reversing (default is local) :


As a library - render local Django template

Sometimes, you don't have access to the request object. A typical example would be creating a PDF receipt for a customer and emailing it. Both of these tasks can take a while, so it is natural to push them into some queue (like RabbitMQ). But if your pdf-rendering task get's executed, you don't have access to the request object. Don't worry - you can still use screamshot as a library. Here's how.

from screamshot.utils import render_template

# you can either render the template to a TemporaryFile:

with render_template('my-template.html', {'context': 'variables'}) as output:
    # do anything you want with the output
    # like attach it to email message, etc.

# or you can specify a path instead:
    {'context': 'variables'},

Please note, that in order to load your static files, screamshot will try to replace all STATIC_URL occurence with a local path to your static files (only if they are not hosted via https of course)

Customizing the page rendering

The CasperJS script appends the screamshot CSS class on the body element. You can easily customize the rendering for printing using this CSS marker in your CSS stylesheet:

.screamshot #navigation {
  display: none;
.screamshot #main {
  margin: 2em;

Capture views with authentication

You can use Basic HTTP authentication in your Django project, create a dedicated user for screenshots and capture the full URL with credentials (http://user:password@host/page/).

Alternatively, you can use a specific view decorator.

Define the authorized IP to capture your pages in your settings :


And use the provided decorator :

from screamshot.decorators import login_required_capturable

def your_view(request):

Renderer command and CLI arguments

You can specify which renderer you would like to use, by setting the CAPTURE_METHOD setting. The default value is 'casperjs'. Possible values are 'casperjs' and 'phantomjs'.

    'CAPTURE_METHOD': 'phantomjs',

By default, we look for thr CasperJS/PhantomJS binary in the PATH environment variable (like which), but you can bypass this:

    'CASPERJS_CMD': '/home/you/Downloads/apps/casperjs',
    'PHANTOMJS_CMD': '/home/you/Downloads/apps/phantomjs'

Please note, that the CAPTURE_METHOD setting specifies which location would be evaluated, i.e. if you set CAPTURE_METHOD to 'phantomjs', PHANTOMJS_CMD would be evaluated.

You can also specify PhantomJS/CasperJS extra-args, such as
--disk-cache=true with the CLI_ARGS setting :
    'CLI_ARGS': ['--disk-cache=true', '--max-disk-cache-size=30000']

See related documentation on PhantomJS and CasperJS homepages.

You can also override the capture script. A default implementation uses capture script written for CasperJS. A default capture script for PhantomJS is also provided.

If you have your own script which you would like to use, specify it in CAPTURE_SCRIPT option.

    'CAPTURE_SCRIPT': '/home/you/scripts/capture.js',

Notes about runserver

If you want to test it using runserver, you won't be able to capture pages coming from the same instance.

Run it twice (on two ports) and configure CAPTURE_ROOT_URL.




  • Lesser GNU Public License
Something went wrong with that request. Please try again.