Spark整合kafka0.10.0新特性（一）(2)

不为空 * */

def onStart(currentOffsets: ju.Map[TopicPartition, jl.Long]): Consumer[K, V] } /**

* Subscribe to a collection of topics.

* @param topics collection of topics to subscribe * @param kafkaParams Kafka

*

* configuration parameters to be used on driver. The same params will be used on executors,

* with minor automatic modifications applied. * Requires \

* with Kafka broker(s) specified in host1:port1,host2:port2 form.

* @param offsets: offsets to begin at on initial startup. If no offset is given for a * TopicPartition, the committed offset (if applicable) or kafka param * auto.offset.reset will be used. *

* 订阅一组topic消息 *kafkaParams参数：

*配置参数将会在driver上使用，并且相同参数也会发送到executor上使用，在executor上可能略微的配置修改

*通知需要设置bootstrap.servers指定broker *

*offsets参数：

* offsets伴随着startup开始,如果对于TopicPartition没有指定offset， * 提交的offset或auto.offset.reset将会被使用 * * */

private case class Subscribe[K, V](

topics: ju.Collection[jl.String],