Merge pull request #54 from DataSketches/quantiles-sketch-to-string

to string UDFs

src/main/java/com/yahoo/sketches/pig/quantiles/DoublesSketchToString.java

@@ -0,0 +1,41 @@
+/*
+ * Copyright 2019, Verizon Media.
+ * Licensed under the terms of the Apache License 2.0. See LICENSE file at the project root for terms.
+ */
+
+package com.yahoo.sketches.pig.quantiles;
+
+import java.io.IOException;
+
+import org.apache.pig.EvalFunc;
+import org.apache.pig.data.DataByteArray;
+import org.apache.pig.data.Tuple;
+
+import com.yahoo.memory.Memory;
+import com.yahoo.sketches.quantiles.DoublesSketch;
+
+/**
+ * This UDF is to get a human-readable summary of a given sketch.
+ */
+public class DoublesSketchToString extends EvalFunc<String> {
+
+  @Override
+  public String exec(final Tuple input) throws IOException {
+    if (input == null) {
+      return null;
+    }
+    if (input.size() != 1) {
+      throw new IllegalArgumentException("expected one input");
+    }
+
+    if (!(input.get(0) instanceof DataByteArray)) {
+      throw new IllegalArgumentException("expected a DataByteArray as a sketch, got "
+          + input.get(0).getClass().getSimpleName());
+    }
+    final DataByteArray dba = (DataByteArray) input.get(0);
+    final DoublesSketch sketch = DoublesSketch.wrap(Memory.wrap(dba.get()));
+
+    return sketch.toString();
+  }
+
+}

src/main/java/com/yahoo/sketches/pig/quantiles/StringsSketchToString.java

@@ -0,0 +1,44 @@
+/*
+ * Copyright 2019, Verizon Media.
+ * Licensed under the terms of the Apache License 2.0. See LICENSE file at the project root for terms.
+ */
+
+package com.yahoo.sketches.pig.quantiles;
+
+import java.io.IOException;
+import java.util.Comparator;
+
+import org.apache.pig.EvalFunc;
+import org.apache.pig.data.DataByteArray;
+import org.apache.pig.data.Tuple;
+
+import com.yahoo.memory.Memory;
+import com.yahoo.sketches.ArrayOfStringsSerDe;
+import com.yahoo.sketches.quantiles.ItemsSketch;
+
+/**
+ * This UDF is to get a human-readable summary of a given sketch.
+ */
+public class StringsSketchToString extends EvalFunc<String> {
+
+  @Override
+  public String exec(final Tuple input) throws IOException {
+    if (input == null) {
+      return null;
+    }
+    if (input.size() != 1) {
+      throw new IllegalArgumentException("expected one input");
+    }
+
+    if (!(input.get(0) instanceof DataByteArray)) {
+      throw new IllegalArgumentException("expected a DataByteArray as a sketch, got "
+          + input.get(0).getClass().getSimpleName());
+    }
+    final DataByteArray dba = (DataByteArray) input.get(0);
+    final ItemsSketch<String> sketch =
+        ItemsSketch.getInstance(Memory.wrap(dba.get()), Comparator.naturalOrder(),
+            new ArrayOfStringsSerDe());
+    return sketch.toString();
+  }
+
+}

src/test/java/com/yahoo/sketches/pig/quantiles/DoublesSketchToStringTest.java

@@ -0,0 +1,57 @@
+/*
+ * Copyright 2019, Verizon Media.
+ * Licensed under the terms of the Apache License 2.0. See LICENSE file at the project root for terms.
+ */
+
+package com.yahoo.sketches.pig.quantiles;
+
+import java.util.Arrays;
+
+import org.apache.pig.EvalFunc;
+import org.apache.pig.data.DataByteArray;
+import org.apache.pig.data.TupleFactory;
+
+import org.testng.annotations.Test;
+
+import com.yahoo.sketches.quantiles.DoublesSketch;
+
+import org.testng.Assert;
+
+public class DoublesSketchToStringTest {
+  private static final TupleFactory TUPLE_FACTORY = TupleFactory.getInstance();
+
+  @Test
+  public void nullInputTuple() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    final String result = func.exec(null);
+    Assert.assertNull(result);
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void emptyInputTuple() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    final String result = func.exec(TUPLE_FACTORY.newTuple());
+    Assert.assertNull(result);
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void tooManyInputs() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    func.exec(TUPLE_FACTORY.newTuple(2));
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void wrongTypeForSketch() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    func.exec(TUPLE_FACTORY.newTuple(Arrays.asList(1.0)));
+  }
+
+  @Test
+  public void normalCase() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    final DoublesSketch sketch = DoublesSketch.builder().build();
+    final String result = func.exec(TUPLE_FACTORY.newTuple(Arrays.asList(new DataByteArray(sketch.toByteArray()))));
+    Assert.assertNotNull(result);
+  }
+
+}

src/test/java/com/yahoo/sketches/pig/quantiles/StringsSketchToStringTest.java

@@ -0,0 +1,63 @@
+/*
+ * Copyright 2019, Verizon Media.
+ * Licensed under the terms of the Apache License 2.0. See LICENSE file at the project root for terms.
+ */
+
+package com.yahoo.sketches.pig.quantiles;
+
+import java.util.Arrays;
+import java.util.Comparator;
+
+import org.apache.pig.EvalFunc;
+import org.apache.pig.data.DataByteArray;
+import org.apache.pig.data.TupleFactory;
+
+import org.testng.annotations.Test;
+
+import com.yahoo.sketches.ArrayOfItemsSerDe;
+import com.yahoo.sketches.ArrayOfStringsSerDe;
+import com.yahoo.sketches.quantiles.ItemsSketch;
+
+import org.testng.Assert;
+
+public class StringsSketchToStringTest {
+  private static final TupleFactory TUPLE_FACTORY = TupleFactory.getInstance();
+
+  private static final Comparator<String> COMPARATOR = Comparator.naturalOrder();
+  private static final ArrayOfItemsSerDe<String> SER_DE = new ArrayOfStringsSerDe();
+
+  @Test
+  public void nullInputTuple() throws Exception {
+    final EvalFunc<String> func = new StringsSketchToString();
+    final String result = func.exec(null);
+    Assert.assertNull(result);
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void emptyInputTuple() throws Exception {
+    final EvalFunc<String> func = new StringsSketchToString();
+    final String result = func.exec(TUPLE_FACTORY.newTuple());
+    Assert.assertNull(result);
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void tooManyInputs() throws Exception {
+    final EvalFunc<String> func = new StringsSketchToString();
+    func.exec(TUPLE_FACTORY.newTuple(2));
+  }
+
+  @Test(expectedExceptions = IllegalArgumentException.class)
+  public void wrongTypeForSketch() throws Exception {
+    final EvalFunc<String> func = new StringsSketchToString();
+    func.exec(TUPLE_FACTORY.newTuple(Arrays.asList(1.0)));
+  }
+
+  @Test
+  public void normalCase() throws Exception {
+    final EvalFunc<String> func = new DoublesSketchToString();
+    ItemsSketch<String> sketch = ItemsSketch.getInstance(COMPARATOR);
+    final String result = func.exec(TUPLE_FACTORY.newTuple(new DataByteArray(sketch.toByteArray(SER_DE))));
+    Assert.assertNotNull(result);
+  }
+
+}