Org-mode to feed a database

[Vikas Rawal <vikaslists@agrarianresearch.org>]

This is, in all likelihood, a crazy idea. Please excuse me for that.

I often need to get a number of people enter complex survey data into a database using some kind of forms. The forms I need are fairly complex (with several nested grids, for example). I need radio-buttons and drop-down menus-equivalents.

I was thinking if it would be worth it to use Org-mode for data entry.

The idea would be something like this. We have a template org file that is copied for each survey respondent. The template contains “blank” org tables and some source blocks. Responses are filled into the org tables, and then the source blocks are evaluated to feed the data into the database. You can version-control the whole thing so that different people can create these records and push it to a centralised git repository.

Has anyone done anything like this? What has been the experience?

There are two aspects that immediately come to my mind.

1. Is org adequately feature rich for such a scenario? There are two concerns I can think of right now. 
   1.1 Would feeding information from one Org table to another become too complicated?
   1.2 Is there a way to build in tab-completion for columns in Org-table using pre-specified categories? 

2. In my case, people who would enter the data into org files would not have used emacs before. So, there is a learning curve. Is it worth it? Or should one just build a web-based graphic application for such people to use?

Vikas