extract regions

Murcia has asked for the wisdom of the Perl Monks concerning the following question:

hi conferes,

I want to extract regions from a file.

>lmo0024        16802072 upstream sequence, from -300 to +3, size 304
ctttcggacaaagcgtggttgattttattcttaacgaaattccagaatggctaatgggtg
caaaaatccagtagctgcaggagcaaatggtgcagcgctagttggggaggaga
atga
>lmo0025        16802073 upstream sequence, from -27 to +3, size 31
aatataaaaattggaggaatagacaaaatgg
.
.
.

The regions are between the >lmo numbers (inclusive)
I know that

while(<>){
  if(/begin/ ... /end/){
     do
  }
}
[download]

extracts regions
but how to do in this case? Thanks for tips! Murcia

Comment on extract regions Download Code

Replies are listed 'Best First'.
Re: extract regions by Roger (Parson) on Feb 09, 2004 at 10:43 UTC
If all you want to do is to extract data between each >lmo, you could do this instead: `local $/ = '>lmo'; while (<DATA>) { s/\n*?>lmo//; # add s/\n//g; if you want to combine lines next if !$_; print "--- REGION ---\n", $_, "\n"; } __DATA__ >lmo0024 16802072 upstream sequence, from -300 to +3, size 304 ctttcggacaaagcgtggttgattttattcttaacgaaattccagaatggctaatgggtg caaaaatccagtagctgcaggagcaaatggtgcagcgctagttggggaggaga atga >lmo0025 16802073 upstream sequence, from -27 to +3, size 31 aatataaaaattggaggaatagacaaaatgg` [download] And the output - `--- REGION --- 0024 16802072 upstream sequence, from -300 to +3, size 304 ctttcggacaaagcgtggttgattttattcttaacgaaattccagaatggctaatgggtg caaaaatccagtagctgcaggagcaaatggtgcagcgctagttggggaggaga atga --- REGION --- 0025 16802073 upstream sequence, from -27 to +3, size 31 aatataaaaattggaggaatagacaaaatgg` [download]	[reply] [d/l] [select]
Re: extract regions by ysth (Canon) on Feb 09, 2004 at 10:46 UTC
We had basically this same question at parsing question. See some of the answers there (particularly the reference to bioperl).	[reply]
Re: extract regions by Abigail-II (Bishop) on Feb 09, 2004 at 10:39 UTC
It's not clear what you want. A construct like: `while (<>) { if (/begin/ ... /end/) { ... } }` [download] still reads the entire file, line-by-line. If your file consists of 'records' that start with `^lmo\d+`, and end with the next occurence, playing with the range operator doesn't make much sense; you might as well do a straigth `while (<>) { ... }` loop. What is it what you really want? Abigail	[reply] [d/l] [select]
Re: extract regions by flounder99 (Friar) on Feb 09, 2004 at 16:01 UTC
Let's say you want from >lmo0025 to >lmo0026 inclusive. If you know that the next record will be >lmo0027 you can do something like this: `use strict; while (<>) { if (/^>lmo0025/ .. /^>lmo0027/) { last if /^>lmo0027/; print; } }` [download] If you don't know that the next record is >lmo0027 you can do something like this: `use strict; my ($firstfoundflag, $lastfoundflag); while (<>) { if ($firstfoundflag \|\| /^>lmo0025/) { $firstfoundflag++; if ($lastfoundflag \|\| /^>lmo0026/ ) { $lastfoundflag++; if (/^>lmo(?!0026)/) { last; } } print; } }` [download] -- flounder	[reply] [d/l] [select]