requirement: splitting on the number of lines/reads about fastp

[sfchen]

This will be implemented:)

[sfchen]

Hi @Toliman06 , this feature is implemented.

You can use -S lines to limit the lines of each split file, instead of using -s to limit the total split file number.

For example, -S 10000 to set every output file will be near 10000 lines.

You can use either -s or -S for splitting output, but you cannot choose both.

Could you please try latest code or binary (not the release) and update this issue?

[Toliman06]

Nice! I will try this ans keep you in touch

[Toliman06]

The program have a strange behavior:
`root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# fastp -i 4d64173b-71dc-4aef-a877-fda22c960a66_1.fastq.gz -S 10000 -o test.fastq
root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# wc -l 0174.test.fastq

11996 0174.test.fastq
root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# wc -l 0165.test.fastq

12000 0165.test.fastq
root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# wc -l 0138.test.fastq

12000 0138.test.fastq
root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# wc -l 0116.test.fastq

12000 0116.test.fastq
root@seqonesandbox:/data/analysis/bbb9ba7-b4e062897804/input_files# wc -l 0073.test.fastq

12000 0073.test.fastq
`
There is 12000 lines per file instead of 10000.

PS: perhaps the number of reads per file could be more interesting that the number of lines (as a read is supposed to be 4 lines...)

[sfchen]

This is reasonable.

Since fastp reads data by blocks, a block contains 1000 reads (4000 lines). And fastp also writes by blocks, so the output may be a little more than 10000 lines.

Since the lines setting are usually much greater than 1000 lines, so this will not cause any problems.

[Toliman06]

Nice! It could be helpfull to mention it in the readme file section.
I will close this subject.