alexmavr / django-spyglass

Django app that provides real-time content notifications through distributed web crawling

django-spyglass

Spyglass is a django application that provides real-time content monitoring of remote websites through distributed web crawling.

Spyglass is configured by specifying a django Model to be populated, a series of websites to be crawled, and an XPath for every combination of model fields and websites.

When a query is submitted to spyglass, a polling task is assigned to the crawler network for completion. The crawlers regularly poll the desired websites until a match is found. Then, a new model instance is created and the user is notified.

Spyglass is licenced under the BSD license.

Available at the PyPI as 'django-spyglass'.

Features

Basic Features

• A polling interval can be specified for each website to prevent continuous polling.
• User-specified search terms are satisfied with fuzzy string matching.
• Queries can be one-time only, or persistent (notifying the user on future changes).
• Notifications are performed with a generic hook which defaults to email notification.

Configuration in a nutshell

• Specify a django model to be populated
• Add a collection of sites to be crawled
• For each site, specify an XPath for every field of the model
• Deploy the application and run any number of crawlers

Requirements

Installation

• Python 2.7+
• Django 1.4.3+ (1.5.1 not tested)
• django-tastypie 0.9.11+

Deployment

• spyglass-crawlie is the default crawler that can be used with spyglass

Configuration

For detailed instructions, please view the documentation

1. Add spyglass to INSTALLED_APPS in your settings.py.

2. Add METAMODEL='<path-to-your-model>' in settings.py for the model to be populated.

   For example, METAMODEL='core.models.NewsStory'

3. Add

   import spyglass.urls
   
   url(r'^', include(spyglass.urls))
   

   to your urls.py

4. Add some sites on spyglass's Site model through fixtures or direct access

5. For each Site entry, create a DataField entry for each field of the metamodel with an XPath where the crawlers can find the data to populate the field. This is explained in detail at the documentation

6. Set up as the required amount of spyglass-crawlie instances by editing their serverconf and userconf files to target the spyglass server

Extras

Class diagram: https://raw2.github.com/afein/SoftEng/master/SDD/files/spyglass.png

Crawler state diagram: https://github.com/afein/SoftEng/blob/master/SDD/files/state.jpg