Bug #53623 for Bio-Trace-ABIF: open_abif => Use of uninitialized value in numeric ge (>=) at ... /Bio/Trace/ABIF.pm line 217

Tue Jan 12 14:23:30 2010 MEC [...] stowers.org - Ticket created

Subject:	open_abif => Use of uninitialized value in numeric ge (>=) at ... /Bio/Trace/ABIF.pm line 217
Date:	Tue, 12 Jan 2010 13:22:14 -0600
To:	"'bug-bio-trace-abif [...] rt.cpan.org'" <bug-bio-trace-abif [...] rt.cpan.org>
From:	"Cook, Malcolm" <MEC [...] stowers.org>

when using version 1.02 Use of uninitialized value in numeric ge (>=) at /n/site/inst/shared/sys/lib/perl5/site_perl/5.8.8/Bio/Trace/ABIF.pm line 217 Bio::Trace::ABIF::_scan_tags('Bio::Trace::ABIF=HASH(0xd8b86f0)') called at /n/site/inst/shared/sys/lib/perl5/site_perl/5.8.8/Bio/Trace/ABIF.pm line 199 Bio::Trace::ABIF::open_abif('Bio::Trace::ABIF=HASH(0xd8b86f0)', 'AB1/yMoBY-P20-A5-YBL007C-MoBY-seq-R_C15.ab1') called at /home/mec/bin/ABIFqtrim line 43 while opening the attached ab1 file ?? Malcolm Cook Stowers Institute for Medical Research - Bioinformatics Kansas City, Missouri USA

Download yMoBY-P20-A5-YBL007C-MoBY-seq-R_C15.ab1
application/octet-stream 319.9k

Message body not shown because it is not plain text.

Tue Jan 12 15:32:08 2010 vita [...] cpan.org - Given to VITA

Tue Jan 12 15:32:09 2010 vita [...] cpan.org - Requestor MEC@stowers.org deleted

Tue Jan 12 15:33:21 2010 vita [...] cpan.org - Requestor MEC@stowers.org added

Tue Jan 12 16:01:34 2010 vita [...] cpan.org - Correspondence added

On Tue Jan 12 14:23:30 2010, MEC@stowers.org wrote: Show quoted text

> when using version 1.02 > > Use of uninitialized value in numeric ge (>=) at > /n/site/inst/shared/sys/lib/perl5/site_perl/5.8.8/Bio/Trace/ABIF.pm > line 217 [...] > while opening the attached ab1 file > > ??

It seems like the ab1 file has the offset to the directory index corrupted (and it seems that only those bytes are corrupted): the offset points to a location that it is beyond the end of the file. I have manually fixed your file. Attached you find a version that should be fine (but I cannot guarantee 100%). That's a strange error: is it the only problematic file or there are more giving the same error? In the latter case, I would look for a bug in the software that creates the ab1 files. Please check if the attached file is ok for you and let me know. Nicola

Subject:

FIXED-yMoBY-P20-A5-YBL007C-MoBY-seq-R_C15.ab1

Download FIXED-yMoBY-P20-A5-YBL007C-MoBY-seq-R_C15.ab1
application/octet-stream 319.9k

Message body not shown because it is not plain text.

Tue Jan 12 16:01:36 2010 The RT System itself - Status changed from 'new' to 'open'

Wed Jan 13 13:32:30 2010 vita [...] cpan.org - Correspondence added

On Tue Jan 12 16:01:34 2010, VITA wrote: Show quoted text

> On Tue Jan 12 14:23:30 2010, MEC@stowers.org wrote:

> > when using version 1.02 > > > > Use of uninitialized value in numeric ge (>=) at > > /n/site/inst/shared/sys/lib/perl5/site_perl/5.8.8/Bio/Trace/ABIF.pm > > line 217 [...] > > while opening the attached ab1 file

I've fixed this in the upcoming 1.03 release. Now, open_abif() returns 0 (=failed to open file) instead of printing those error messages when it encounters a problem on opening the ab1 file. It also outputs an "IO Error" message on stderr in case of error, so you know there's a problem with the content of the ab1 file. Nicola

Wed Jan 13 13:32:32 2010 vita [...] cpan.org - Status changed from 'open' to 'resolved'

Wed Jan 13 15:20:42 2010 MEC [...] stowers.org - Correspondence added

Subject:	RE: [rt.cpan.org #53623] open_abif => Use of uninitialized value in numeric ge (>=) at ... /Bio/Trace/ABIF.pm line 217
Date:	Wed, 13 Jan 2010 14:19:25 -0600
To:	"'bug-Bio-Trace-ABIF [...] rt.cpan.org'" <bug-Bio-Trace-ABIF [...] rt.cpan.org>
From:	"Cook, Malcolm" <MEC [...] stowers.org>

Hi, The ab1 file was created with '3730 Data Collection' v. 3.0. It was not otherwise touched by any processes than opening it with Bio::Trace::ABIF (so far as I know - other than moving it around the network). Your fix fixed it - I can now extract meta data from your "fixed" version of my buggy file. You're the man! Still no idea what created the problem. Can you shed any light on possibilities? Or, can you suggest how I might pose the same question to ABI? Perhaps it is a known issue. Also, can you tell me: is there an easy 'fix' for such corrupt files that does not involve manual intervention? I have a few such files. I'm making reference to ABI's document "ABIF File format specification" and understand the issue a little better. Thanks, Malcolm Cook Stowers Institute for Medical Research - Bioinformatics Kansas City, Missouri USA Show quoted text

> -----Original Message----- > From: Nicola Vitacolonna via RT > [mailto:bug-Bio-Trace-ABIF@rt.cpan.org] > Sent: Tuesday, January 12, 2010 3:02 PM > To: Cook, Malcolm > Subject: [rt.cpan.org #53623] open_abif => Use of > uninitialized value in numeric ge (>=) at ... > /Bio/Trace/ABIF.pm line 217 > > <URL: https://rt.cpan.org/Ticket/Display.html?id=53623 > > > On Tue Jan 12 14:23:30 2010, MEC@stowers.org wrote:

> > when using version 1.02 > > > > Use of uninitialized value in numeric ge (>=) at > > /n/site/inst/shared/sys/lib/perl5/site_perl/5.8.8/Bio/Trace/ABIF.pm > > line 217 [...] > > while opening the attached ab1 file > > > > ??

> > It seems like the ab1 file has the offset to the directory > index corrupted (and it seems that only those bytes are > corrupted): the offset points to a location that it is beyond > the end of the file. I have manually fixed your file. > Attached you find a version that should be fine (but I cannot > guarantee 100%). > > That's a strange error: is it the only problematic file or > there are more giving the same error? > In the latter case, I would look for a bug in the software > that creates the ab1 files. > > Please check if the attached file is ok for you and let me know. > > Nicola > > >

Wed Jan 13 15:20:44 2010 The RT System itself - Status changed from 'resolved' to 'open'

Wed Jan 13 17:10:18 2010 vita [...] cpan.org - Correspondence added

On Wed Jan 13 15:20:42 2010, MEC@stowers.org wrote: Show quoted text

> > The ab1 file was created with '3730 Data Collection' v. 3.0. > > It was not otherwise touched by any processes than opening it with > Bio::Trace::ABIF (so far as I know - other than moving it around the > network). > […] > Still no idea what created the problem. Can you shed any light on > possibilities?

Not an easy guess. From what you say and from the attachment you've sent, the files might be corrupted at the source (the application that writes them does it wrong). A disk or network failure could also be the cause. A wierd bug in Bio::Trace::ABIF is more unlikely, especially if you've opened the files only for reading. Show quoted text

> Also, can you tell me: is there an easy 'fix' for such corrupt files > that does not involve manual intervention? I have a few such files.

Not easier than writing a script to do the job—provided that the problem is the same for all the files. Show quoted text

> I'm making reference to ABI's document "ABIF File format > specification" and understand the issue a little better.

So, the problem of your file was the content of bytes 26-30 (data offset), which specify the location of the directory. I have searched where the directory might possibly be and changed those bytes accordingly. I was lucky enough that it worked. I do not consider this issue as a bug of my module, as long as I cannot reproduce it. So, I'm setting this to resolved. Feel free to re-open the issue if you can prove otherwise. Nicola

Wed Jan 13 17:10:32 2010 vita [...] cpan.org - Status changed from 'open' to 'resolved'