gistfile1.py

import os
import base64
from subprocess import call

from django import http
from django.conf import settings
from django_seo_js.backends import SEOBackendBase


SITE_CONF = getattr(settings, 'SITE_CONF')
BASE_URL = "%s%s" % (SITE_CONF['protocol'], SITE_CONF['domain'])
SNAPSHOTS_DIR = getattr(settings, 'SNAPSHOTS_DIR')
PHANTOM_RUNNER = os.path.join(os.path.dirname(__file__), 'phantom-runner.js')


## Backends
class PhantomJSBackend(SEOBackendBase):
    """
    A custom backend for django-seo-js that will use phantomjs to
    render the page, and store it in a predefined location.
    """
    def get_response_for_url(self, url):
        """
        Accepts a fully-qualified url.
        Returns an HttpResponse, passing through all headers and the status code.
        """
        if not url or "//" not in url:
            raise ValueError("Missing or invalid url: %s" % url)
        fragment = url.split('_escaped_fragment_=')[1]
        return self.render_phantomjs_response(fragment)

    def update_url(self, url):
        """
        Force an update of the cache for a particular URL.
        Returns True on success, False on fail.
        """
        self._get_or_create_snapshot(url, force_update=True)
        return True

    def _get_fragment_hash(self, fragment):
        return base64.urlsafe_b64encode(fragment)

    def _get_or_create_snapshot(self, fragment, force_update=False):
        """
        Generate and cache the html for URL using phantomjs.
        """
        outfile = os.path.join(
            SNAPSHOTS_DIR, "%s.html" % self._get_fragment_hash(fragment))
        url = "%s/#\!%s" % (BASE_URL, fragment)  # note we are escaping the '!'

        if not os.path.exists(outfile) or force_update:
            phantom_cmd = 'phantomjs %s %s > %s' % (PHANTOM_RUNNER, url, outfile)
            call(phantom_cmd, shell=True)

        return outfile

    def render_phantomjs_response(self, fragment):
        """
        Tries to return the requested snapshot. If it does not exist,
        generate it first, save it, then return it.
        """
        snapshot_file = self._get_or_create_snapshot(fragment)
        snapshot_content = open(snapshot_file, 'rb').read()
        return http.HttpResponse(snapshot_content, content_type="text/html")