[Bioperl-l] How put tmp file in the tmp directory when using Bio::Tools::Run::StandAloneBlastPlus?

[Chris Fields]

On Jun 6, 2010, at 7:33 PM, Peng Yu wrote:

> Chris,
> 
> Sorry if I ask some question that is documented. I'm trying to read as
> much relevant document as I can, but I can't guarantee read everything
> that is relevant.

I found the appropriate answer to your question within 2 minutes of reading your post, using a simple ack search (or grep if you prefer) on the StandAloneBlastPlus documentation, then used perldoc as previously mentioned.  If you are unwilling to spend 5x that (only 10 minutes), and are unwilling to read the relevant documentation, and are so ready to admit it, why would you expect us to spend our (extremely valuable) time helping you?  

> I don't understand why "registered" files and databases are relevant
> for my use case. (BTW, what are "registered" files and what they are
> for?).  I don't specify any database option when I call
> StandAloneBlastPlus in my example. If I call blast directly, I'll not
> get any tempfiles. Therefore, it is reasonable to expect not seeing
> any tempfiles with bioperl module as I understand StandAloneBlastPlus
> is a wrapper for blast+. What is the purpose of leaving temp files in
> the current directory? Is it better not to leave any tempfiles in the
> current directory by default?

We have provided you an answer.  You possibly have a point (why not remove the tempfiles), but you'll have to ask the module author(s) that for their reasoning.  You can always make that suggestion, or perhaps even submit a patch, and work around the problem in the meantime.  Contribute something constructive.  We value that.  

But your posts at this point come across as rote complaining, not suggestions, and your general lack of appreciation/gratitude for any answers we've already given you does not exactly endear yourself to those whose time is very limited.

chris


> On Sun, Jun 6, 2010 at 6:37 PM, Chris Fields <cjfields at illinois.edu> wrote:
>> Peng,
>> 
>> You really need to read all relevant documentation prior to firing off emails to the list.  This is answered in the StandAloneBlastPlus installed documentation ('perldoc Bio::Tools::Run::StandAloneBlastPlus'):
>> 
>>  Cleaning up temp files
>>       Temporary analysis files produced under a single factory instances can
>>       be unlinked by running
>> 
>>        $fac->cleanup;
>> 
>>       Tempfiles are generally not removed unless this method is explicitly
>>       called. "cleanup()" only unlinks "registered" files and databases. All
>>       temporary files are automatically registered; in particular,
>>       "anonymous" databases (such as
>> 
>>        $fac->Bio::Tools::Run::StandAloneBlastPlus->new(
>>          -db_data => 'myseqs.fas',
>>          -create => 1
>>        );
>> 
>>       without a "-db_name" specification) are registered for cleanup. Any
>>       file or database can be registered with an internal method:
>> 
>>        $fac->_register_temp_for_cleanup('testdb');
>> 
>> 
>> 
>> chris
>> 
>> On Jun 6, 2010, at 5:52 PM, Peng Yu wrote:
>> 
>>> The following perl program generate tmp files at the current
>>> directory, which is annoying. Is there a way to put the tmp files in a
>>> tmp directory say /tmp?
>>> 
>>> #!/usr/bin/env perl
>>> 
>>> use strict;
>>> use warnings;
>>> use Bio::Tools::Run::StandAloneBlastPlus;
>>> use Bio::Perl;
>>> 
>>> my $factory = Bio::Tools::Run::StandAloneBlastPlus->new();
>>> 
>>> my $seq1 = Bio::Perl::read_sequence('first.fa');
>>> my $seq2 = Bio::Perl::read_sequence('second.fa');
>>> print $seq1->seq, "\n";
>>> print $seq2->seq, "\n";
>>> 
>>> my $blast_result=$factory->bl2seq(-method=>'blastn',
>>>  -query=> $seq1,
>>>  -subject=> $seq2
>>> );
>>> 
>>> --
>>> Regards,
>>> Peng
>>> _______________________________________________
>>> Bioperl-l mailing list
>>> Bioperl-l at lists.open-bio.org
>>> http://lists.open-bio.org/mailman/listinfo/bioperl-l
>> 
>> 
> 
> 
> 
> -- 
> Regards,
> Peng
> 
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l