fix(sdk): uses offset for ByteBuffer array offset (#209)

ByteBuffer could have an array offset. We consider the length of the
ByteBuffer, but we do not consider the offset and default it to 0. This
PR fixes this problem by adding in the array offset when encrypting.

sdk/src/main/java/io/opentdf/platform/sdk/NanoTDF.java

@@ -140,7 +140,7 @@ public int createNanoTDF(ByteBuffer data, OutputStream outputStream,
             System.arraycopy(iv, 0, actualIV, kIvPadding, iv.length);
         } while (Arrays.equals(actualIV, kEmptyIV));	// if match, we need to retry to prevent key + iv reuse with the policy
 
-        byte[] cipherData = gcm.encrypt(actualIV, authTagSize, data.array(), 0, dataSize);
+        byte[] cipherData = gcm.encrypt(actualIV, authTagSize, data.array(), data.arrayOffset(), dataSize);
 
         // Write the length of the payload as int24
         int cipherDataLengthWithoutPadding = cipherData.length - kIvPadding;