Navigation

Managing Your Code

You can customize Smart Mastering by implementing custom match functions, actions, and merge functions. These functions will be part of your normal application code.

Writing

When writing custom functions, you’ll need to decide where your functions should be stored. The Smart Mastering libraries do not require your functions to be in a particular place; you can store your functions in any module that fits into your overall code structure. That said, a common practice is to segment your Smart Mastering custom functions from other code. If you are using ml-gradle to deploy your project, including projects using the Data Hub Framework, the recommendation is to put your modules under the src/main/ml-modules/ directory. For instance, you might have the following paths:

  • src/main/ml-modules/smart-mastering/match/my-match-algorithm-1.xqy
  • src/main/ml-modules/smart-mastering/match/my-match-algorithm-2.sjs
  • src/main/ml-modules/smart-mastering/action/my-action.sjs
  • src/main/ml-modules/smart-mastering/merge/my-merge-algorithm.sjs

Note that this is a recommended structure, not a required one. This layout uses one file for each algorithm function, but you have the option to group them either all together or by type. For instance, you can create a match-algorithms.sjs, as long as you give each function a distinct name.

XQuery

When writing XQuery functions, you will need to choose a namespace for your library modules. Smart Mastering has no special requirements for your namespaces. If you have established a pattern for your applications, you should follow that. If you haven’t, you could use a pattern such as:

http://{company URL}/{project}/smart-mastering/{function}/{algorithm}

For example:

http://example.com/big-hub/smart-mastering/match/favorite-color

Using

To make use of your custom functions, you need to deploy them (assuming you are using a modules database) and then refer to them in your match or merge options.

Deploying

Deploy your Smart Mastering custom functions the same way you deploy the rest of your application code. If you are using ml-gradle, running gradle mlLoadModules or gradle mlReloadModules will deploy all of your application’s source code, including your Smart Mastering customizations. (See ml-gradle’s documentation on Common Tasks.)

Configuration

To make use of your custom functions, add them to your match or merge options.

how to point the configuration to modules