taxonomy support?

Taxonomy: Yes, I've been thinking about doing that.

Batch processing: I probably won't do this, but the module is written in such a way that batch processing would be an easy add-on module.

Views: I am interested in doing a general QueryPath views module. If that were done, this would inherit views support. This module currently does not store the data in relational tables, so views building would require a specialized back-end (or decomposition of the data into database tables). I'd welcome patches to do that.

OK, been staring at this a bit and swapped some emails with @mbutcher. I would definitely like to look into adding taxonomy support based on the information coming back. This is what I'm envisioning:

Setting Vocabularies

Setting the target vocabularies for each of the major OpenAmplify semantic components:

• Domain
• Subdomain
• Topics
• Locations
• Education
• Slang
• Flamboyance

If no target vocabulary is chosen for a component, the terms from that component are not added to the node's taxonomy.

Free-Tagging or Not?

If the vocabulary has free-tagging enabled, offer this as an option for the component terms. If the term already exists, use it - if not, add it.

If the vocabulary does not allow free-tagging, then only assign terms to the node that already exist in the vocabulary.

If only one term is allowed, use the highest-ranking term from that OpenAmplify component.

Views Support

If we leverage the existing node-taxonomy framework, Views will fall into place nicely as it already supports taxonomy-based arguments, relationships, sorting and filters when a node-based View is created.

Batch Processing

As for batch operations - I wonder if we could implement a hook_cron callback that would iterate over the node table on each cron run and attempt to execute OpenAmplify against some (sane) user-configured number of nodes?

Or maybe a Drush hook to allow for command-line batch processing?

Or both...?