[Feature][Ossfile] 期望确实能提高单cvs文件的并行度

### Search before asking

- [x] I had searched in the [feature](https://github.com/apache/seatunnel/issues?q=is%3Aissue+label%3A%22Feature%22) and found no similar feature requirement.

### Description

作为源端时，目前csv文件记录数为两百万+的时候，跑的速度很慢，要十几分钟是最乐观的，能不能先把这个csv文件切成多个小csv文件，然后并行读写

### Usage Scenario

_No response_

### Related issues

_No response_

### Are you willing to submit a PR?

- [ ] Yes I am willing to submit a PR!

### Code of Conduct

- [x] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct)