[Java] Should ListVector.reset() reset the position(index)?

```java
RootAllocator rootAllocator = new RootAllocator();
ListVector identity = ListVector.empty("identity", rootAllocator);
VectorSchemaRoot root = new VectorSchemaRoot(List.of(identity));
root.allocateNew();

UnionListWriter writer = identity.getWriter();

byte[] bytes1 = "hello1".getBytes(StandardCharsets.UTF_8);
ArrowBuf buffer = rootAllocator.buffer(bytes1.length);
buffer.setBytes(0,bytes1);
writer.startList();
writer.varChar().writeVarChar(0,bytes1.length,buffer);
writer.endList();
root.setRowCount(1);
System.out.println(root.contentToTSVString());

root.getFieldVectors().forEach(ValueVector::reset); //reset fieldVector

byte[] bytes2 = "hello2".getBytes(StandardCharsets.UTF_8);
buffer.setBytes(0,bytes2);
writer.startList();
writer.varChar().writeVarChar(0,bytes2.length,buffer);
writer.endList();

root.setRowCount(1);
System.out.println(root.contentToTSVString());
```

the result is
```
identity
["hello1"]

identity
null
```

The above results are confusing。

The correct way is using the `writer.setPosition(index)`,

I don't know if it's a bug or other something.

### Component(s)

Java

先使用 ListVector、ValueVector.reset() 和 UnionListWriter.setPosition(index) 執行 Java 重現，以觀察位置行為。閱讀相關的 reset 和 writer 位置實作，接著確立重設後的預期結果，並為示範的序列新增 coverage，讓行為不再模稜兩可。

由索引模型根據 Issue 內容生成。