error parsing regexp: Invalid repeat count

[user0380]

Hi, I'm using seqkit to extract CDS region from various sequences. This is the command I used:

seqkit locate -G -r -p 'ATG([ATCGN]{3}){2083}T(AG|AA|GA)'

It produced this error info:

[ERRO] error parsing regexp: invalid repeat count: {2083}

This regexp is valid in Perl. Is seqkit's regexp compatible with Perl?

Thanks!

[shenwei356]

I just noticed that the built-in Go regexp package enforces a hard-coded limit of 1000 for explicit repetition quantifiers 😮

A workaround solution:

seqkit locate -G -r -p 'ATG(...){1000}(...){1000(...){83}T(AG|AA|GA)'

([ATCGN]{3}){1000} won't work ~

Another one with a post-filter step is recommended.

seqkit locate -G -r -i -p 'ATG([ATCGN]{3})+?T(AG|AA|GA)' | csvtk filter2 -t -f '($end-$start+1)/3-2 == 2083'