Search code examples
marklogicmarklogic-dhfml-gradle

Is it possible to use token replacements in DHF flow files?

On MarkLogic 10.0-3, using ml-data-hub plugin 5.2.1, I would like to specify variable source and target databases in mapping steps of flows when deploying to various environments (saliman.properties plugin 1.4.6) based on the mlAppName property, e.g.:

[...snip, snip...]
    "2" : {
          "name" : "myStep",
          "description" : "a step in my flow",
          "options" : {
            "additionalCollections" : [ ],
            "sourceQuery" : "cts.collectionQuery([\"targetCollection\"])",
            "mapping" : {
              "name" : "MyFlow-myStep",
              "version" : 0
            },
            "targetEntity" : "MyEntity",
            "sourceDatabase" : "%%mlAppName%%-STAGING",
            "collections" : [ "myStep", "MyEntity" ],
            "permissions" : "data-hub-operator,read,data-hub-operator,update",
            "validateEntity" : false,
            "sourceCollection" : "sourceCollection",
            "outputFormat" : "json",
            "targetDatabase" : "%%mlAppName%%-FINAL"
          },
          "customHook" : {
            "module" : "",
            "parameters" : { },
            "user" : "",
            "runBefore" : false
          },
          "retryLimit" : 0,
          "batchSize" : 100,
          "threadCount" : 4,
          "stepDefinitionName" : "entity-services-mapping",
          "stepDefinitionType" : "MAPPING"
        },
[...snip, snip...]

However, the token substitutions don't happen. Is this impossible, or is there a better best-practice method to implement this? Thanks.

Solution

  • It's unfortunately not yet supported. You'd have to use a custom Gradle task to manipulate the contents of the documents after they've been loaded. Please create a ticket in the DHF github project though to request this feature, it's a worthwhile one to add.