MLeap Flavor/Deployment mlflow#3: Java implementation of a scoring se…

…rver for SageMaker (mlflow#327)

* Add SparkJava plugin

* Change default port

* Support multiple instances of SageMakerServer

* Add SageMaker tests

* Rename SageMakerServer >> ScoringServer and add new tests

* New tests

* Format code

* Remove unused import

* Format again

* New test and artifact name update

* Using jetty scoring server

* Select random port

* Port fixes

* Add port number to start method

* Use apache http for tests

* pom versioning

* More tests

* Code formatting

* Line removal

* Format again

* Format code, address comments

* Format code, docs

* Add stream parsing test, fix other test asserts

* Import fix

* Add CSV to predictor error test

* Fix test

* Fix test, format code

* Remove abstractions for testing sys env reading

* Fixes

* Address comments

mlflow/java/pom.xml

@@ -2,10 +2,10 @@
 <project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
   <modelVersion>4.0.0</modelVersion>
   <groupId>org.mlflow</groupId>
-  <artifactId>mlflow-java</artifactId>
+  <artifactId>mlflow</artifactId>
   <version>0.4.2</version>
   <packaging>jar</packaging>
-  <name>mlflow-java</name>
+  <name>mlflow</name>
   <url>http://maven.apache.org</url>
   <properties>
     <maven.compiler.source>1.8</maven.compiler.source>
@@ -36,19 +36,54 @@
     <dependency>
       <groupId>ml.combust.mleap</groupId>
       <artifactId>mleap-spark_2.11</artifactId>
-      <version>0.4.0</version>
+      <version>0.10.3</version>
     </dependency>
     <dependency>
       <groupId>ml.combust.mleap</groupId>
       <artifactId>mleap-runtime_2.11</artifactId>
-      <version>0.10.0</version>
+      <version>0.10.3</version>
     </dependency>
     <dependency>
       <groupId>org.apache.maven.plugins</groupId>
       <artifactId>maven-surefire-plugin</artifactId>
       <version>2.22.0</version>
       <scope>test</scope>
     </dependency>
+    <dependency>
+      <groupId>org.eclipse.jetty</groupId>
+      <artifactId>jetty-server</artifactId>
+      <version>9.4.11.v20180605</version>
+    </dependency>
+    <dependency>
+      <groupId>org.eclipse.jetty</groupId>
+      <artifactId>jetty-servlet</artifactId>
+      <version>9.4.11.v20180605</version>
+    </dependency>
+    <dependency>
+      <groupId>org.eclipse.jetty</groupId>
+      <artifactId>jetty-util</artifactId>
+      <version>9.4.11.v20180605</version>
+    </dependency>
+    <dependency>
+      <groupId>org.slf4j</groupId>
+      <artifactId>slf4j-nop</artifactId>
+      <version>1.6.6</version>
+    </dependency>
+    <dependency>
+      <groupId>org.slf4j</groupId>
+      <artifactId>slf4j-jdk14</artifactId>
+      <version>1.7.25</version>
+    </dependency>
+    <dependency>
+      <groupId>org.apache.httpcomponents</groupId>
+      <artifactId>httpclient</artifactId>
+      <version>4.5.6</version>
+    </dependency>
+    <dependency>
+      <groupId>commons-io</groupId>
+      <artifactId>commons-io</artifactId>
+      <version>2.5</version>
+    </dependency>
   </dependencies>
   <build>
     <plugins>

mlflow/java/src/main/java/org/mlflow/sagemaker/MLeapPredictor.java

@@ -7,7 +7,7 @@ public class MLeapPredictor extends Predictor {
   public MLeapPredictor(String modelDataPath, String inputSchemaPath) {}
 
   @Override
-  protected DataFrame predict(DataFrame input) {
+  protected PredictorDataWrapper predict(PredictorDataWrapper input) {
     throw new UnsupportedOperationException("Not yet implemented!");
   }
 

mlflow/java/src/main/java/org/mlflow/sagemaker/Predictor.java

@@ -1,12 +1,12 @@
 package org.mlflow.sagemaker;
 
-
 /**
  * A generic predictor object that provides a uniform interface for model inference. By extending
  * {@link org.mlflow.sagemaker.Predictor}, models of a specific flavor can expose their inference
  * routines for use by generic tools, such as model containers
  */
 public abstract class Predictor {
   /** Performs inference on the specified input and produces a result */
-  protected abstract DataFrame predict(DataFrame input) throws PredictorEvaluationException;
+  protected abstract PredictorDataWrapper predict(PredictorDataWrapper input)
+      throws PredictorEvaluationException;
 }

mlflow/java/src/main/java/org/mlflow/sagemaker/PredictorDataWrapper.java

@@ -0,0 +1,47 @@
+package org.mlflow.sagemaker;
+
+/** Input/output data representation for use by {@link org.mlflow.sagemaker.Predictor} objects */
+public class PredictorDataWrapper {
+  public enum ContentType {
+    Json,
+    Csv
+  }
+
+  private final String content;
+  private final ContentType contentType;
+
+  /** Constructs a PredictorDataWrapper */
+  public PredictorDataWrapper(String content, ContentType contentType) {
+    this.content = content;
+    this.contentType = contentType;
+  }
+
+  /** @return The type of content contained in the wrapper (JSON, CSV, etc.) */
+  ContentType getContentType() {
+    return this.contentType;
+  }
+
+  /**
+   * Produces a JSON string representation of the PredictorDataWrapper
+   *
+   * @return A string in JSON format
+   */
+  String toJson() {
+    if (this.contentType == ContentType.Json) {
+      return this.content;
+    } else {
+      throw new UnsupportedOperationException(
+          "Converting a data wrapper of a non-JSON content type to JSON is not yet supported.");
+    }
+  }
+
+  /**
+   * Produces a CSV string representation of the PredictorDataWrapper
+   *
+   * @return A string in CSV format
+   */
+  String toCsv() {
+    throw new UnsupportedOperationException(
+        "Converting a data wrapper to CSV is not yet supported.");
+  }
+}