ARROW-393: [JAVA] JSON file reader fails to set the buffer size on String data vector

java/tools/src/main/java/org/apache/arrow/tools/Integration.java

@@ -80,7 +80,7 @@ public void execute(File arrowFile, File jsonFile) throws IOException {
           Schema schema = footer.getSchema();
           LOGGER.debug("Input file size: " + arrowFile.length());
           LOGGER.debug("Found schema: " + schema);
-          try (JsonFileWriter writer = new JsonFileWriter(jsonFile);) {
+          try (JsonFileWriter writer = new JsonFileWriter(jsonFile, JsonFileWriter.config().pretty(true));) {
             writer.start(schema);
             List<ArrowBlock> recordBatches = footer.getRecordBatches();
             for (ArrowBlock rbBlock : recordBatches) {

java/tools/src/test/java/org/apache/arrow/tools/TestIntegration.java

@@ -24,9 +24,12 @@
 import static org.apache.arrow.tools.ArrowFileTestFixtures.writeInput;
 import static org.junit.Assert.fail;
 
+import java.io.BufferedReader;
 import java.io.File;
 import java.io.FileNotFoundException;
 import java.io.IOException;
+import java.io.StringReader;
+import java.util.Map;
 
 import org.apache.arrow.memory.BufferAllocator;
 import org.apache.arrow.memory.RootAllocator;
@@ -44,6 +47,11 @@
 import org.junit.Test;
 import org.junit.rules.TemporaryFolder;
 
+import com.fasterxml.jackson.core.util.DefaultPrettyPrinter;
+import com.fasterxml.jackson.core.util.DefaultPrettyPrinter.NopIndenter;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.fasterxml.jackson.databind.SerializationFeature;
+
 public class TestIntegration {
 
   @Rule
@@ -69,7 +77,7 @@ public void testValid() throws Exception {
     File testOutFile = testFolder.newFile("testOut.arrow");
     testOutFile.delete();
 
-    // generate an arow file
+    // generate an arrow file
     writeInput(testInFile, allocator);
 
     Integration integration = new Integration();
@@ -90,6 +98,50 @@ public void testValid() throws Exception {
     integration.run(args3);
   }
 
+  @Test
+  public void testJSONRoundTripWithVariableWidth() throws Exception {
+    File testJSONFile = new File("../../integration/data/simple.json");
+    File testOutFile = testFolder.newFile("testOut.arrow");
+    File testRoundTripJSONFile = testFolder.newFile("testOut.json");
+    testOutFile.delete();
+    testRoundTripJSONFile.delete();
+
+    Integration integration = new Integration();
+
+    // convert to arrow
+    String[] args1 = { "-arrow", testOutFile.getAbsolutePath(), "-json",  testJSONFile.getAbsolutePath(), "-command", Command.JSON_TO_ARROW.name()};
+    integration.run(args1);
+
+    // convert back to json
+    String[] args2 = { "-arrow", testOutFile.getAbsolutePath(), "-json",  testRoundTripJSONFile.getAbsolutePath(), "-command", Command.ARROW_TO_JSON.name()};
+    integration.run(args2);
+
+    BufferedReader orig = readNormalized(testJSONFile);
+    BufferedReader rt = readNormalized(testRoundTripJSONFile);
+    String i, o;
+    int j = 0;
+    while ((i = orig.readLine()) != null && (o = rt.readLine()) != null) {
+      Assert.assertEquals("line: " + j, i, o);
+      ++j;
+    }
+  }
+
+  private ObjectMapper om = new ObjectMapper();
+  {
+    DefaultPrettyPrinter prettyPrinter = new DefaultPrettyPrinter();
+    prettyPrinter.indentArraysWith(NopIndenter.instance);
+    om.setDefaultPrettyPrinter(prettyPrinter);
+    om.enable(SerializationFeature.INDENT_OUTPUT);
+    om.enable(SerializationFeature.ORDER_MAP_ENTRIES_BY_KEYS);
+  }
+
+  private BufferedReader readNormalized(File f) throws IOException {
+    Map<?,?> tree = om.readValue(f, Map.class);
+    String normalized = om.writeValueAsString(tree);
+    return new BufferedReader(new StringReader(normalized));
+  }
+
+
   @Test
   public void testInvalid() throws Exception {
     File testValidInFile = testFolder.newFile("testValidIn.arrow");

java/vector/src/main/java/org/apache/arrow/vector/file/json/JsonFileReader.java

@@ -22,6 +22,7 @@
 import static com.fasterxml.jackson.core.JsonToken.START_ARRAY;
 import static com.fasterxml.jackson.core.JsonToken.START_OBJECT;
 import static java.nio.charset.StandardCharsets.UTF_8;
+import static org.apache.arrow.vector.schema.ArrowVectorType.OFFSET;
 
 import java.io.File;
 import java.io.IOException;
@@ -128,15 +129,12 @@ private void readVector(Field field, FieldVector vector) throws JsonParseExcepti
         valueVector.allocateNew();
         Mutator mutator = valueVector.getMutator();
 
-        int innerVectorCount = count;
-        if (vectorType.getName() == "OFFSET") {
-          innerVectorCount++;
-        }
-        mutator.setValueCount(innerVectorCount);
+        int innerVectorCount = vectorType.equals(OFFSET) ? count + 1 : count;
         for (int i = 0; i < innerVectorCount; i++) {
           parser.nextToken();
           setValueFromParser(valueVector, i);
         }
+        mutator.setValueCount(innerVectorCount);
         readToken(END_ARRAY);
       }
       // if children

java/vector/src/main/java/org/apache/arrow/vector/schema/ArrowVectorType.java

@@ -81,4 +81,19 @@ public String getName() {
   public String toString() {
     return getName();
   }
+
+  @Override
+  public int hashCode() {
+    return type;
+  }
+
+  @Override
+  public boolean equals(Object obj) {
+    if (obj instanceof ArrowVectorType) {
+      ArrowVectorType other = (ArrowVectorType) obj;
+      return type == other.type;
+    }
+    return false;
+  }
+
 }