3 # parse_ncbi_results retreives files of search results from ncbi,
4 # and is released under the terms of the GPL version 2, or any later
5 # version, at your option. See the file README and COPYING for more
8 # Copyright 2004 by Don Armstrong <don@donarmstrong.com>.
10 # $Id: ss,v 1.1 2004/06/29 05:26:35 don Exp $
22 parse_ncbi_results [options]
28 --dir, -D directory to stick results into [default .]
29 --name, -n file naming scheme [default ${search}_results.$format]
30 --terms, -t file of search terms [default -]
31 --debug, -d debugging level [default 0]
32 --help, -h display this help
33 --man, -m display manual
41 Debug verbosity. (Default 0)
45 Display brief useage information.
55 parse_ncbi_results -D ./ncbi_results/ -n '${search}_name.html' < search_parameters
57 Will pretty much do what you want
63 use vars qw($DEBUG $REVISION);
66 ($REVISION) = q$LastChangedRevision: 1$ =~ /LastChangedRevision:\s+([^\s]+)/;
67 $DEBUG = 0 unless defined $DEBUG;
70 use XML::Parser::Expat;
73 # XXX parse config file
75 my %options = (debug => 0,
82 GetOptions(\%options,'keyword|k=s','debug|d+','help|h|?','man|m');
85 pod2usage() if $options{help};
86 pod2usage({verbose=>2}) if $options{man};
88 $DEBUG = $options{debug};
91 use constant {NAME => 0,
102 my @csv_fields = qw(name hits rzscore refseq location alias database terms description function);
106 for my $file_name (@ARGV) {
107 my $file = new IO::File $file_name, 'r' or die "Unable to open file $file_name $!";
110 my @gene = map {s/^\"//; s/\"$//; $_;} split /(?<=\")\,(?=\")/, $_;
111 $genes{$gene[NAME]}{name} = $gene[NAME];
112 $genes{$gene[NAME]}{database}{$gene[DBNAME]}++;
113 $genes{$gene[NAME]}{hits}++;
114 $genes{$gene[NAME]}{terms}{$gene[KEYWORD]}++;
115 add_unique_parts($genes{$gene[NAME]},'refseq',$gene[REFSEQ]);
116 add_if_better($genes{$gene[NAME]},'description',$gene[DESCRIPTION]);
117 add_if_better($genes{$gene[NAME]},'location',$gene[LOCATION]);
118 add_unique_parts($genes{$gene[NAME]},'function',split(/; /, $gene[FUNCTION]));
119 add_unique_parts($genes{$gene[NAME]},'alias', split(/; /, $gene[ALIAS]));
123 print join(',',map {qq("$_")} @csv_fields),qq(\n);
124 for my $gene (keys %genes) {
125 $genes{$gene}{rzscore} = scalar keys %{$genes{$gene}{terms}};
126 next if $genes{$gene}{rzscore} == 1 and exists $genes{$gene}{terms}{antigen};
127 $genes{$gene}{rzscore} -= 1 if exists $genes{$gene}{terms}{antigen};
128 print STDOUT join (',',
129 map {s/"//g; qq("$_")}
132 if (ref $value eq 'HASH') {
133 join('; ',map {qq($_:$$value{$_})} keys %$value);
135 elsif (ref $value eq 'ARRAY') {
141 } @{$genes{$gene}}{@csv_fields}
146 sub add_unique_parts{
147 my ($hr,$key,@values) = @_;
148 if (not defined $$hr{key}) {
149 $$hr{$key} = [@values];
153 @temp_hash{@{$$hr{$key}}} = (1) x scalar @{$$hr{$key}};
154 $temp_hash{@values} = (1) x scalar @values;
155 $$hr{$key} = [keys %temp_hash];
160 my ($hash, $key, $value) = @_;
161 if (not defined $$hash{$key}) {
162 $$hash{$key} = $value;
164 elsif (length $$hash{$key} < length $value and $value !~ /^NO\s+(LOCATION|DESCRIPTION)+$/) {
165 $$hash{$key} = $value;