This is very early prototype.

Olapy is a python implementation of an XMLA / MDX server compatible with Excel.

This integration reached this point:

• The server is running in Zope process. Since olapy uses spyne, I could reuse Products.ERP5.Document.SOAPBinding.
• We have a test simulating a very basic case.

To try this:

• Install the business template
• In portal_web_service create a new "Olapy Connection" document.
• Use the URL of this Olapy Connection in Excel ( similar to https://softinstXXX.host.vifib.net/portal_web_services/1/ )

What is still TODO:

• I don't have an Excel instance to test with, so I cannot confirm this really work.
• The cube data is dummy data. Next step will probably to be to serve some prefilled pandas arrays.
• I would like to be able to configure the cube and dimensions directly on the "Olapy Connection" object. I would also like it to work directly using inventory API. It should not be too hard to "convert" MDX queries in Inventory API queries.
• We would need to have some authentication and (configurable) security. For exemple, only sales agents can access sales cube, product managers can only see data related to their products, etc.
• At the moment, Olapy only seem to supports a subset of Excel ( not yet support CUBEMEMBER / CUBEVALUE , no drill down, discovery sometimes hardcoded values, MDX parser is simple regexps ), this will be improved too. I could not connect using python xmla or http://webpivottable.com/demo/

This depends on slapos!204