[Bioperl-l] help needed with __DATA__

[Shawn Hoon]

You need to reset the file handle.
Note sure if this is the best thing to do, but you can use the tell and 
seek function of perl:

In subroutine read of base.pm I did the following:
 >> lines added
<snip>

  >>my $offset;
sub read {
     my $self = shift;

     my $renzs = new Bio::Restriction::EnzymeCollection(-empty => 1);
     >>seek DATA,($offset||=tell DATA), 0;

     while (<DATA>) {
         chomp;
</snip>

seems to work for me on perl 5.6.1

hth,

shawn


On Sunday, July 13, 2003, at 12:03  PM, Heikki Lehvaslaiho wrote:

> Could anyone suggest a fix, please:
>
> In the new Bio::Restriction analysis system (the relevant classes are 
> in
> bioperl-live CVS head) the SeqIO approach is used to to load the 
> default
> set of enzymes into a EnzymeCollection from Bio::Restriction::IO::base
> class. I've put the tab delimited list of default enzymes at the end of
> the file after a '__DATA__' line and read it using 'while (<DATA>)'.
>
> This seems to cause a problem. Consecutive calls to code does not seem
> to reset the DATA file handle.
>
> In the following code, the three blocks should work identically but 
> only
> the block executed first has enzymes in the collection.  One can
> reshuffle the order of code blocks and only the first one will work.
>
> ---------------------------------------------
> use Bio::Restriction::EnzymeCollection;
> use Bio::Restriction::IO;
> use strict;
>
> # 1.
> {
> my $renzs = new Bio::Restriction::EnzymeCollection();
> print scalar $renzs->each_enzyme, "\n";
> }
> #2.
> {
> my $renzs2 = new Bio::Restriction::EnzymeCollection();
> print scalar $renzs2->each_enzyme, "\n";
> }
> #3.
> {
> my $in  = new Bio::Restriction::IO; # defaults to 'base.pm'
> my $renzs3 = $in->read;
> print scalar $renzs3->each_enzyme, "\n";
> }
> ---------------------------
>
> Adding 'close DATA' into base.pm will not help. The file handle is not
> automatically opened.
>
> Is this result reproducible?
>
> I am using 'v5.8.0 built for i386-linux-thread-multi'
>
> I could use here documents instead, I guess,  but DATA should work, 
> too.
> 	
> 	-Heikki
>
>
>
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>