X-Git-Url: https://git.donarmstrong.com/?p=function2gene.git;a=blobdiff_plain;f=bin%2Fget_ncbi_xml_results;fp=bin%2Fget_ncbi_xml_results;h=3ee477861abb0ca8032c60eaf93e0df36977916e;hp=0000000000000000000000000000000000000000;hb=d09b67e0af77d6f2818e41d6b4d648cff651c79d;hpb=5d4602c246e5d2c22435bda4c07116251f1fa546 diff --git a/bin/get_ncbi_xml_results b/bin/get_ncbi_xml_results new file mode 100755 index 0000000..3ee4778 --- /dev/null +++ b/bin/get_ncbi_xml_results @@ -0,0 +1,172 @@ +#! /usr/bin/perl + +# get_ncbi_results retreives files of search results from ncbi, and is +# released under the terms of the GPL version 2, or any later version, +# at your option. See the file README and COPYING for more +# information. + +# Copyright 2004 by Don Armstrong . + +# $Id: ss,v 1.1 2004/06/29 05:26:35 don Exp $ + + +use warnings; +use strict; + + +use Getopt::Long; +use Pod::Usage; + +=head1 NAME + + get_ncbi_results [options] + +get_ncbi_results - Retrieve search results from NCBI using parameters +passed on stdin. + +=head1 SYNOPSIS + + + Options: + --format, -f format of search results to return [default xml] + --database, -b database to search for results [default gene] + --dir, -D directory to stick results into [default .] + --name, -n file naming scheme [default ${search}_results.$format] + --terms, -t file of search terms [default -] + --debug, -d debugging level [default 0] + --help, -h display this help + --man, -m display manual + +=head1 OPTIONS + +=over + +=item B<--debug, -d> + +Debug verbosity. (Default 0) + +=item B<--help, -h> + +Display brief useage information. + +=item B<--man, -m> + +Display this manual. + +=back + +=head1 EXAMPLES + + get_ncbi_results -f xml -b gene -D ./results/ -n '${search}_name.xml' < search_parameters + +Will pretty much do what you want + +=cut + + + +use vars qw($DEBUG $REVISION); + +BEGIN{ + ($REVISION) = q$LastChangedRevision: 1$ =~ /LastChangedRevision:\s+([^\s]+)/; + $DEBUG = 0 unless defined $DEBUG; +} + +use IO::File; +use URI::ParamMunge; +use LWP::UserAgent; + +# XXX parse config file + +my %options = (debug => 0, + help => 0, + man => 0, + format => 'xml', + database => 'gene', + dir => '.', + name => '${search}_results.$format', + terms => '-', + pubmed_site => 'http://www.ncbi.nlm.nih.gov', + pubmed_search_url => '/entrez/query.fcgi?db=gene&cmd=search&term=12q24*+AND+homo[Orgn]&doptcmdl=Brief&dispmax=1000', + pubmed_get_url => '/entrez/query.fcgi?db=gene&cmd=Text&dopt=XML', + ); + +GetOptions(\%options,'format|f=s','database|b=s','name|n=s', + 'terms|t=s','dir|D=s','debug|d+','help|h|?','man|m'); + +pod2usage() if $options{help}; +pod2usage({verbose=>2}) if $options{man}; + +$DEBUG = $options{debug}; + +if (not -d $options{dir}) { + die "$options{dir} does not exist or is not a directory"; +} + +#open search terms file +my $terms; +if ($options{terms} eq '-') { + $terms = \*STDIN; +} +else { + $terms = new IO::File $options{terms}, 'r' or die "Unable to open file $options{terms}: $!"; +} + +my $ua = new LWP::UserAgent(agent=>"DA_get_ncbi_results/$REVISION"); + +#For every term +while (<$terms>) { + # Get uids to retrieve + chomp; + my $search = $_; + my $format = $options{format}; + my $url = uri_param_munge($options{pubmed_site}.$options{pubmed_search_url}, + {term => $search, + db => $options{database}, + }, + ); + my $request = HTTP::Request->new('GET', $url); + my $response = $ua->request($request); + $response = $response->content; + my @gene_ids = $response =~ m/\[GeneID\:\s+(\d+)\]/g; + + my $file_name = eval qq("$options{name}") or die $@; + my $xml_file = new IO::File "$options{dir}/$file_name", 'w' or die "Unable to open $options{dir}/$file_name: $!"; + + # Get XML file + my @current_ids; + while (@current_ids = splice(@gene_ids,0,20)) { + $url = uri_param_munge($options{pubmed_site}.$options{pubmed_get_url}, + {dopt => uc($options{format}), + db => $options{database}, + }, + ) .'&' . join('&',map {qq(uid=$_)} @current_ids); + $request = HTTP::Request->new('GET', $url); + $response = $ua->request($request); + $response = $response->content; + # For some dumb reason, they send us xml with html + # entities. Ditch them. + #$response = decode_entities($response); + $response =~ s/\>/>/gso; + $response =~ s/\</ and suffix a ditch them. + $response =~ s/^\s*
//gso;
+	  $response =~ s#
\s*$##gso; + + $response =~ s#<\?xml[^>]+>##gso; + $response =~ s#]+>##gso; + + print {$xml_file} $response; + sleep 10; + } + undef $xml_file; +} + + + + + + +__END__