a crawler for ajax applications using hashbang urls.

creating a html snapshot of your site for googlebot by walking through hashbang links.

sample usage:

./crawlajax.py 'http://2011.jsconf.us/#!/schedule'

this command creates following folder structure and files:

|- www
 \|- schedule
  |\|- index.html
  |- news
  |\|- index.html
  |- speakers
  |\|- index.html
  ...

you can use any web server you preferred now to serve your static html snapshots by forwarding requests with escaped_fragment parameter to this folder.

http://2011.jsconf.us/?_escaped_fragment_=/schedule => http://2011.jsconf.us/googlebot/schedule/index.html

using phantomjs to take html snapshot of rendered dom. you can change snapshot command via construction parameter snapshot_cmd.

for making ajax applications crawlable, check instructions at:

• http://code.google.com/web/ajaxcrawling/docs/getting-started.html

we tried using htmlunit as google suggests at first, but we had performance problems, so we chose this way.

project satisfies only our applications requirements for now, you're free to fork and enrich it according to your needs.

dependencies:

• python-2.6.6
• phantomjs-1.3.0

see instructions to build and install phantomjs:

• http://code.google.com/p/phantomjs/wiki/BuildInstructions
• http://code.google.com/p/phantomjs/wiki/Installation

you can run unit tests via cmd: