ARROW-633/634: [Java] Add FixedSizeBinary support in Java and integration tests (Updated)

integration/integration_test.py

@@ -463,6 +463,46 @@ def generate_column(self, size, name=None):
         return self.column_class(name, size, is_valid, values)
 
 
+class FixedSizeBinaryType(PrimitiveType):
+
+    def __init__(self, name, byte_width, nullable=True):
+        super(FixedSizeBinaryType, self).__init__(name, nullable=nullable)
+        self.byte_width = byte_width
+
+    @property
+    def numpy_type(self):
+        return object
+
+    @property
+    def column_class(self):
+        return FixedSizeBinaryColumn
+
+    def _get_type(self):
+        return OrderedDict([('name', 'fixedsizebinary'), ('byteWidth', self.byte_width)])
+
+    def _get_type_layout(self):
+        return OrderedDict([
+            ('vectors',
+             [OrderedDict([('type', 'VALIDITY'),
+                           ('typeBitWidth', 1)]),
+              OrderedDict([('type', 'DATA'),
+                           ('typeBitWidth', self.byte_width)])])])
+
+    def generate_column(self, size, name=None):
+        is_valid = self._make_is_valid(size)
+        values = []
+
+        for i in range(size):
+            draw = (np.random.randint(0, 255, size=self.byte_width)
+                    .astype(np.uint8)
+                    .tostring())
+            values.append(draw)
+
+        if name is None:
+            name = self.name
+        return self.column_class(name, size, is_valid, values)
+
+
 class StringType(BinaryType):
 
     @property
@@ -525,6 +565,22 @@ def _get_buffers(self):
         ]
 
 
+class FixedSizeBinaryColumn(PrimitiveColumn):
+
+    def _encode_value(self, x):
+        return ''.join('{:02x}'.format(c).upper() for c in x)
+
+    def _get_buffers(self):
+        data = []
+        for i, v in enumerate(self.values):
+            data.append(self._encode_value(v))
+
+        return [
+            ('VALIDITY', [int(x) for x in self.is_valid]),
+            ('DATA', data)
+        ]
+
+
 class StringColumn(BinaryColumn):
 
     def _encode_value(self, x):
@@ -719,6 +775,9 @@ def get_field(name, type_, nullable=True):
         return BinaryType(name, nullable=nullable)
     elif type_ == 'utf8':
         return StringType(name, nullable=nullable)
+    elif type_.startswith('fixedsizebinary_'):
+        byte_width = int(type_.split('_')[1])
+        return FixedSizeBinaryType(name, byte_width=byte_width, nullable=nullable)
 
     dtype = np.dtype(type_)
 
@@ -751,7 +810,8 @@ def _generate_file(name, fields, batch_sizes, dictionaries=None):
 def generate_primitive_case(batch_sizes, name='primitive'):
     types = ['bool', 'int8', 'int16', 'int32', 'int64',
              'uint8', 'uint16', 'uint32', 'uint64',
-             'float32', 'float64', 'binary', 'utf8']
+             'float32', 'float64', 'binary', 'utf8',
+             'fixedsizebinary_19', 'fixedsizebinary_120']
 
     fields = []
 

java/vector/src/main/codegen/data/ArrowTypes.tdd

@@ -57,6 +57,11 @@
       fields: [],
       complex: false
     },
+    {
+      name: "FixedSizeBinary",
+      fields: [{name: "byteWidth", type: int}],
+      complex: false
+    }
     {
       name: "Bool",
       fields: [],

java/vector/src/main/codegen/data/ValueVectorTypes.tdd

@@ -122,6 +122,21 @@
         }
       ]
     },
+    {
+      major: "Fixed",
+      width: -1,
+      javaType: "byte[]",
+      boxedType: "ArrowBuf",
+      minor: [
+        {
+          class: "FixedSizeBinary",
+          typeParams: [ {name: "byteWidth", type: "int"} ],
+          arrowType: "org.apache.arrow.vector.types.pojo.ArrowType.FixedSizeBinary",
+          friendlyType: "byte[]",
+          fields: [{name: "buffer", type: "ArrowBuf"}],
+        }
+      ]
+    },
     {
       major: "VarLen",
       width: 4,

java/vector/src/main/codegen/templates/HolderReaderImpl.java

@@ -128,6 +128,10 @@ public void read(Nullable${name}Holder h) {
     holder.buffer.getBytes(holder.start, bytes, 0, ${type.width});
     ${friendlyType} value = new BigDecimal(new BigInteger(bytes), holder.scale);
     return value;
+  <#elseif minor.class == "FixedSizeBinary">
+    byte[] value = new byte [holder.byteWidth];
+    holder.buffer.getBytes(0, value, 0, holder.byteWidth);
+    return value;
   <#else>
     ${friendlyType} value = new ${friendlyType}(this.holder.value);
     return value;

java/vector/src/main/java/org/apache/arrow/vector/BaseFixedWidthVector.java

@@ -42,7 +42,7 @@
  */
 public abstract class BaseFixedWidthVector extends BaseValueVector
         implements FixedWidthVector, FieldVector, VectorDefinitionSetter {
-  private final byte typeWidth;
+  private final int typeWidth;
 
   protected int valueAllocationSizeInBytes;
   protected int validityAllocationSizeInBytes;
@@ -54,7 +54,7 @@ public abstract class BaseFixedWidthVector extends BaseValueVector
   protected int valueCount;
 
   public BaseFixedWidthVector(final String name, final BufferAllocator allocator,
-                                      FieldType fieldType, final byte typeWidth) {
+                                      FieldType fieldType, final int typeWidth) {
     super(name, allocator);
     this.typeWidth = typeWidth;
     field = new Field(name, fieldType, null);

java/vector/src/main/java/org/apache/arrow/vector/BitVector.java

@@ -58,7 +58,7 @@ public BitVector(String name, BufferAllocator allocator) {
    * @param allocator allocator for memory management.
    */
   public BitVector(String name, FieldType fieldType, BufferAllocator allocator) {
-    super(name, allocator, fieldType, (byte) 0);
+    super(name, allocator, fieldType, 0);
     reader = new BitReaderImpl(BitVector.this);
   }
 

java/vector/src/main/java/org/apache/arrow/vector/BufferLayout.java

@@ -70,7 +70,7 @@ public static BufferLayout dataBuffer(int typeBitWidth) {
       case 128:
         return VALUES_128;
       default:
-        throw new IllegalArgumentException("only 8, 16, 32, or 64 bits supported");
+        throw new IllegalArgumentException("only 8, 16, 32, 64, or 128 bits supported");
     }
   }
 
@@ -90,7 +90,7 @@ public static BufferLayout byteVector() {
 
   private final BufferType type;
 
-  private BufferLayout(BufferType type, int typeBitWidth) {
+  BufferLayout(BufferType type, int typeBitWidth) {
     super();
     this.type = Preconditions.checkNotNull(type);
     this.typeBitWidth = (short) typeBitWidth;