add pair end reads functions

[yinshiyi]

• compare read header from r1 to r2, if 99% match pass
• r1 also should not be has identical record.seq() for matching record.head()

record.head(): Returns the ID/identifier string (&str).
record.seq(): Returns the sequence string (&str).

how about the fastq that is not ordered by read header.
record i for r1 matches to record j for r2.
check that first, if not, try to sort r1 and r2 by read header.
search record 1 from r1 in 1-1e6 in r2, if not match, then discard and keep searching until match or end of file

[yinshiyi]

https://github.com/linsalrob/fastq-pair
has done a great amount of amazing work
but does not support compressed fastq.gz file

[yinshiyi]

1e6 is a reasonable depth for most use case, after sorting by qname, if none of the r1 is found in the first million reads of r2, report such and request user to investigate their files