Split RabbitMQPlugin into ExtractorRoutingService and MessageService (#155)

* Initial split into services

* single MessageService file

* Add stubbed EventSinkService (using fake MessageService)

* Included an example of how to call EventSinkService from within an HTML template

* Wiring updates

* wiring for ExtractorRoutingService

* more wiring

* Sync to switch branches

* further wiring of new services

* more wiring

* enough updates to compile!

* add Singleton marker

* initialize factory to account for lack of onStart

* Filled out EventSinkService impl

* Adjust direct calls to Amplitude to use EventSinkService

* add noop for MessageService

* fix: cleanup commented out yucky code

* Added missing tracking for file uploads

* Remove direct usage of Amplitude Javascript SDK

* Fix logFileUploadEvent, starting to think about auth question

* add manual submission event

* Update RabbitMQMessageService.scala

* Fix failing test case / wrong package name

* Fix failing test case, revert package name change

* Parameterize exchange_type (default value is "topic")

* Removed old call to Amplitude. It was breaking the upload page. File div would disappear after upload even though the file was successfully added.

* Added rabbitmq info back to status. Created a getInfo() function to let the implementation specify what to use.
Might want to use the same pattern for future plugin -> services conversions.

Co-authored-by: Mike Lambert <[email protected]>
Co-authored-by: Luigi Marini <[email protected]>

Author: 3 people

app/api/Datasets.scala

@@ -52,8 +52,10 @@ class  Datasets @Inject()(
   relations: RelationService,
   userService: UserService,
   thumbnailService : ThumbnailService,
+  routing: ExtractorRoutingService,
   appConfig: AppConfigurationService,
-  esqueue: ElasticsearchQueue) extends ApiController {
+  esqueue: ElasticsearchQueue,
+  sinkService: EventSinkService) extends ApiController {
 
   lazy val chunksize = play.Play.application().configuration().getInt("clowder.chunksize", 1024*1024)
 
@@ -606,9 +608,7 @@ class  Datasets @Inject()(
         val mdMap = metadata.getExtractionSummary
 
         //send RabbitMQ message
-        current.plugin[RabbitmqPlugin].foreach { p =>
-          p.metadataAddedToResource(metadataId, metadata.attachedTo, mdMap, Utils.baseUrl(request), request.apiKey, request.user)
-        }
+        routing.metadataAddedToResource(metadataId, metadata.attachedTo, mdMap, Utils.baseUrl(request), request.apiKey, request.user)
 
         events.addObjectEvent(request.user, id, x.name, EventType.ADD_METADATA_DATASET.toString)
 
@@ -655,9 +655,7 @@ class  Datasets @Inject()(
                 val mdMap = metadata.getExtractionSummary
 
                 //send RabbitMQ message
-                current.plugin[RabbitmqPlugin].foreach { p =>
-                  p.metadataAddedToResource(metadataId, metadata.attachedTo, mdMap, Utils.baseUrl(request), request.apiKey, request.user)
-                }
+                routing.metadataAddedToResource(metadataId, metadata.attachedTo, mdMap, Utils.baseUrl(request), request.apiKey, request.user)
 
                 events.addObjectEvent(request.user, id, x.name, EventType.ADD_METADATA_DATASET.toString)
 
@@ -739,10 +737,8 @@ class  Datasets @Inject()(
         }
 
         // send extractor message after attached to resource
-        current.plugin[RabbitmqPlugin].foreach { p =>
-          metadataIds.foreach { mId =>
-            p.metadataRemovedFromResource(mId, ResourceRef(ResourceRef.dataset, id), Utils.baseUrl(request), request.apiKey, request.user)
-          }
+        metadataIds.foreach { mId =>
+          routing.metadataRemovedFromResource(mId, ResourceRef(ResourceRef.dataset, id), Utils.baseUrl(request), request.apiKey, request.user)
         }
 
         Ok(toJson(Map("status" -> "success", "count" -> metadataIds.size.toString)))
@@ -2500,8 +2496,10 @@ class  Datasets @Inject()(
         val baseURL = controllers.routes.Datasets.dataset(id).absoluteURL(https(request))
 
         // Increment download count if tracking is enabled
-        if (tracking)
+        if (tracking) {
           datasets.incrementDownloads(id, user)
+          sinkService.logDatasetDownloadEvent(dataset, user)
+        }
 
         // Use custom enumerator to create the zip file on the fly
         // Use a 1MB in memory byte array