Running samplestack on ml-gradle

[rjrudin]

The main changes here are:

• Most of the custom tasks are gone in favor of existing tasks in ml-gradle
• I re-organized the database config files and modules under database/ml-config and database/ml-modules

I kept MarkLogicSlurpTask largely as-is, as it contains a good amount of logic that is specific to samplestack. And I think it's a nice demo then of how if ml-gradle doesn't do something that you want, it's easy to make a custom task that does exactly what you want.

For testing, I did a new install of ML and ran "gradle appserver", and everything ran fine. This was done on ML 8.0-3.

[grechaw]

I've had a chance to look at these changes -- I think it's a really good step for samplestack, not sure whether the PR is being analyzed by ss team...

[laurelnaiad]

Rob, could you help me understand the purpose behind rearranging the files in the repo? Was this necessary, did it make things easier?

[rjrudin]

Hey Daphne - the separation of ML config and modules is what ml-app-deployer expects by default, and the organization of ml-config is based on the management API.

I think it's a bit easier to understand for someone new to ML - they can go to the management API docs and seem the same breakout in the docs as they do in the ml-config directory.

The ml-modules directory structure is based on how samplestack is already organized.

[grechaw]

I've verified this PR from the java side. I think it's ready to go. I also think its a great step forward for samplestack as it gets rid of many lines of code and takes the build stuff out of customizations.

[rjrudin]

Charles - I'm going to add support for mimetypes and then call ml-gradle a "2.0", and then I'll update the PR for this to include the latest version and the simplified buildscript (since only jcenter() is needed).