[SOLVED] Is there a way to reverse generate the ml-gradle project configuration setting based on existing ML application server?

Is there a way to reverse generate the ml-gradle project configuration setting based on existing ML application server?

mlDeploy and mlUndeploy enable to rebuild the ML application from CI/CD. Using ml-gradle is the default way to manage new ML app server. That is great.

How to migrate existing ML app to use ml-gradle to manage? How to automatically rebuild the ML gradle project from an existing ML app servers, particaully different indexes? (We have over 50 different types of indexes on one existing ML app server.) It is really slow and easy to make human mistake by recreating the ml-gradle project manually.

Any suggestions to automate that reverse engineering?

Solution

This is the best approach - https://github.com/marklogic-community/ml-gradle/wiki/Exporting-resources .

The challenge is that MarkLogic doesn't have the concept of an "application" - i.e. once you deploy your resources, they're effectively the same as all the OOTB resources and thus it's difficult to know what you deployed vs what's OOTB.