Full Text Indexers Excitement and Zope

Saturday, 01 April 2006

Both Django and Turbogears Developers are excited about MerQuery Full text indexer. MerQuery plans to be the Full Text Indexer[1] for Python frameworks. The developer of MerQuery plans to use Apache Project's Lucene Indexing and search technology.

The first thoughts were, hmm.. doesn't Zope already have a full text indexer already. Yes, it does. Zope has had Full Text capabilities from early days of Zope2. Read this workshop paper given by Jim Fulton at Pycon'01.

Ah,talk of Zope2 makes people give a 1000 yard stare. "Zope is not teh cool" some say. Let that pass and look at whether Zope3 is doing any "cool" stuff. Lucene?

Hoorah! it apparently does!. If you look carefully the recipe was written in 2004. And, the recipe itself is a great tutorial on how to write new apps for zope3.

It was never claimed that it is easy to write Wikis in 20 minutes using Zope3. But then, how many times do you write your own wikis? But, where Zope2 and Zope3 shine is in the non-obvious stuff like Text indexing, fine grained permissions,object publishing and the list goes on.

I'm excited about new Python frameworks like Django and Pylons, but I'm also at a loss to understand cold shoulder given to Zope3. Zope3 deserves better. I agree[2] that Zope.org is not as sexy as TG or Django websites.Going by Zopistas' claim that Zope3 is aiming to eat J2EE's lunch, Zope.org does very little to support that claim.

Zope3 needs a "Message" and Zope.org needs a Massage(Maybe a lot).

[1] Read David Mertz's excellent article - Developing a Full-Text Indexer in Python to understand what a Full Text Indexer does.

[2] Read Jeff Shell's blog - Griddle Noise he is a passionate Zope user.