deepsearch is a Haystack extension used to index deep and nested model relationships. Its key features are:

• deepsearch.indexes.DeepSearchIndex class able to index deeply nested related objects and their fields.
• deepsearch.models.FieldBoost model that stores weights of each index field for query-time field boosting.
• resume_index command able to reindex a slice of all objects or only a subset of index fields.
• Celery support for real-time indexing.

Requirements:

• Python >= 2.6
• Django >= 1.4
• django-haystack >= 2.1.0
• South.>= 0.8.1
• celery>=3.1 (optional)

Recommended:

• haystackbrowser (to inspect indexed data)

Setup:

1. Include 'deepsearch' in your INSTALLED_APPS.

2. Configure HAYSTACK_CONNECTIONS.

3. To enable real-time indexing add the following line to settings.py:

   HAYSTACK_SIGNAL_PROCESSOR = 'deepsearch.signals.DeepSearchSignalProcessor'

4. Create search_indexes.py in your app directory.

5. python manage.py init_boosts to update index field weight values. You should run this whenever you modify index schema.

6. python manage.py rebuild_index will update index and deepsearch.models.IndexRelation table.