Add support for unique R1 and R2 prefixes in `split2`

I'd like to be able to do the following:

```
seqkit split2 -1 my_r1.fq.gz -2 my_r2.fq.gz --by-parts 10 --by-parts-prefix "my_r{read}_"
```

And get out:

```
my_r1_001.fq.gz
my_r2_001.fq.gz
my_r1_002.fq.gz
my_r2_002.fq.gz
...
```

Thanks!