[Bio] / FigKernelScripts / validate_fasta.pl Repository:
ViewVC logotype

View of /FigKernelScripts/validate_fasta.pl

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.1 - (download) (as text) (annotate)
Wed Oct 10 23:39:40 2007 UTC (12 years, 4 months ago) by redwards
Branch: MAIN
CVS Tags: mgrast_dev_08112011, rast_rel_2009_05_18, mgrast_dev_08022011, rast_rel_2014_0912, rast_rel_2008_06_18, myrast_rel40, rast_rel_2008_06_16, mgrast_dev_05262011, rast_rel_2008_12_18, mgrast_dev_04082011, rast_rel_2008_07_21, rast_rel_2010_0928, rast_2008_0924, mgrast_version_3_2, mgrast_dev_12152011, rast_rel_2008_04_23, mgrast_dev_06072011, rast_rel_2008_09_30, rast_rel_2009_0925, rast_rel_2010_0526, rast_rel_2014_0729, mgrast_dev_02212011, rast_rel_2010_1206, mgrast_release_3_0, mgrast_dev_03252011, rast_rel_2010_0118, mgrast_rel_2008_0924, mgrast_rel_2008_1110_v2, rast_rel_2009_02_05, rast_rel_2011_0119, mgrast_rel_2008_0625, mgrast_release_3_0_4, mgrast_release_3_0_2, mgrast_release_3_0_3, mgrast_release_3_0_1, mgrast_dev_03312011, mgrast_release_3_1_2, mgrast_release_3_1_1, mgrast_release_3_1_0, mgrast_dev_04132011, rast_rel_2008_10_09, mgrast_dev_04012011, rast_release_2008_09_29, mgrast_rel_2008_0806, mgrast_rel_2008_0923, mgrast_rel_2008_0919, rast_rel_2009_07_09, rast_rel_2010_0827, mgrast_rel_2008_1110, myrast_33, rast_rel_2011_0928, rast_rel_2008_09_29, mgrast_rel_2008_0917, rast_rel_2008_10_29, mgrast_dev_04052011, mgrast_dev_02222011, rast_rel_2009_03_26, mgrast_dev_10262011, rast_rel_2008_11_24, rast_rel_2008_08_07, HEAD
Initial release of programs to check and fix valid fasta format

#__perl__


=pod 

=head1 validate_fasta.pl

Check a file and see whether it is valid fasta. Returns 0 if valid, or an error code if it is not valid.
I have tried to catch most of the common mistakes.

=cut

use strict;
my $filename=shift || die "$0 <fasta filename>";



# define some error codes. I just made these up. The rough idea is that based on the error code here, you can parse the file and correct it automatically.

my %error=(
	"File not found" => 1,
	"File contains mac newlines" => 2,
	"File contains spaces before identifier" => 3,
	"File contains numbers where there should be sequence" => 4,
	"Identifier line is empty" => 5,
	"Sequence is empty" => 6,
	"Duplicate identifiers" => 7,
	);
	

unless (-e $filename) {caught_error("File not found")}

my $parsedfirstline=0;
my $seq; my %identifier;
if ($filename =~ /gz$/) {open(IN, "gunzip -c $filename|") || die "Can't open a pipe to the file $filename"}
else {open(IN, $filename) || die "can't find file even though it exists"}
while (<IN>)
{
	if (/\r/) {caught_error("File contains mac newlines")}
	if (/^\s+\>/) {caught_error("File contains spaces before identifier")}
	if ($_ !~ /^>/ && /\d/) {caught_error("File contains numbers where there should be sequence")}	
	if (/^>$/) {caught_error("Identifier line is empty")}
	if (/^>/)
	{
		if ($parsedfirstline)
		{	
			# there won't be any sequence when we read the first line
			$seq =~ s/[\s\d]//g;
			unless ($seq) {caught_error("Sequence is empty")}
		}
		undef $seq;
		$parsedfirstline=1;
	}
	if ($_ !~ /^>/) {$seq .= $_}
	if (/^>(\S+)/)
	{
		if ($identifier{$1}) {caught_error("Duplicate identifiers")}
		$identifier{$1}=1;
	}
}
# make sure there is some sequence in the last line
$seq =~ s/[\s\d]+//g;
unless ($seq) {caught_error("Sequence is empty")}

exit(0);



sub caught_error {
	my $err=shift;
	if (defined $error{$err}) 
	{
		print STDERR "$error{$err}: $err\n";
		exit($error{$err});
	}
	else
	{
		print STDERR "-100: Caught an unusual error: $err\n";
		exit(-100);
	}
}


MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3