| 1 | package WWW::Search::PubMed; | ||
| 2 | |||
| 3 | =head1 NAME | ||
| 4 | |||
| 5 | WWW::Search::PubMed - Search the NCBI PubMed abstract database. | ||
| 6 | |||
| 7 | =head1 SYNOPSIS | ||
| 8 | |||
| 9 | use WWW::Search; | ||
| 10 | my $s = new WWW::Search ('PubMed'); | ||
| 11 | $s->native_query( 'ACGT' ); | ||
| 12 | while (my $r = $s->next_result) { | ||
| 13 | print $r->title . "\n"; | ||
| 14 | print $r->description . "\n"; | ||
| 15 | } | ||
| 16 | |||
| 17 | =head1 DESCRIPTION | ||
| 18 | |||
| 19 | WWW::Search::PubMed provides a WWW::Search backend for searching the | ||
| 20 | NCBI/PubMed abstracts database. | ||
| 21 | |||
| 22 | =head1 VERSION | ||
| 23 | |||
| 24 | This document describes WWW::Search::PubMed version 1.004, | ||
| 25 | released 31 October 2007. | ||
| 26 | |||
| 27 | =head1 REQUIRES | ||
| 28 | |||
| 29 | L<WWW::Search|WWW::Search> | ||
| 30 | L<XML::DOM|XML::DOM> | ||
| 31 | |||
| 32 | =cut | ||
| 33 | |||
| 34 | our($VERSION) = '1.004'; | ||
| 35 | |||
| 36 | 2 | ▶ | use strict; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 37 | 2 | ▶ | use warnings; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 38 | |||
| 39 | require WWW::Search; | ||
| 40 | require WWW::SearchResult; | ||
| 41 | 2 | ▶ | use WWW::Search::PubMed::Result; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 42 | 2 | ▶ | use base qw(WWW::Search); |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 43 | |||
| 44 | 0 | ▶ | use XML::DOM; |
|
Subroutine
BEGIN: called CC=1 SCAR=2.6
| |||
| 45 | our $debug = 0; | ||
| 46 | |||
| 47 | use constant ARTICLES_PER_REQUEST => 20; | ||
| 48 | use constant QUERY_ARTICLE_LIST_URI => 'http://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&retmax=500'; # term=ACTG | ||
| 49 | use constant QUERY_ARTICLE_INFO_URI => 'http://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed'; # &id=12167276&retmode=xml | ||
| 50 | |||
| 51 | =begin private | ||
| 52 | |||
| 53 | =item C<< native_setup_search ( $query, $options ) >> | ||
| 54 | |||
| 55 | Sets up the NCBI search using the supplied C<$query> string. | ||
| 56 | |||
| 57 | =end private | ||
| 58 | |||
| 59 | =cut | ||
| 60 | |||
| 61 | sub native_setup_search { | ||
| 62 | my $self = shift; | ||
| 63 | my $query = shift; | ||
| 64 | my $options = shift; | ||
| 65 | |||
| 66 | $self->user_agent( "WWW::Search::PubMed/${VERSION} libwww-perl/${LWP::VERSION}; <http://kasei.us/code/pubmed/>" ); | ||
| 67 | |||
| 68 | my $ua = $self->user_agent(); | ||
| 69 | my $url = QUERY_ARTICLE_LIST_URI . '&term=' . WWW::Search::escape_query($query); | ||
| 70 | my $response = $ua->get( $url ); | ||
| 71 | my $success = $response->is_success; | ||
| 72 | if ($success) { | ||
| 73 | my $parser = new XML::DOM::Parser; | ||
| 74 | my $content = $response->content; | ||
| 75 | $self->{'_xml_parser'} = $parser; | ||
| 76 | my $doc = $parser->parse( $content ); | ||
| 77 | |||
| 78 | $self->{'_count'} = eval { ($doc->getElementsByTagName('Count')->item(0)->getChildNodes)[0]->getNodeValue() } || 0; | ||
| 79 | |||
| 80 | my @articles; | ||
| 81 | my $ids = $doc->getElementsByTagName('Id'); | ||
| 82 | my $n = $ids->getLength; | ||
| 83 | foreach my $i (0 .. $n - 1) { | ||
| 84 | my $node = $ids->item( $i ); | ||
| 85 | my @children = $node->getChildNodes(); | ||
| 86 | push(@articles, + $children[0]->getNodeValue() ); | ||
| 87 | } | ||
| 88 | $self->{'_article_ids'} = \@articles; | ||
| 89 | } else { | ||
| 90 | return undef; | ||
| 91 | } | ||
| 92 | } | ||
| 93 | |||
| 94 | =begin private | ||
| 95 | |||
| 96 | =item C<< native_retrieve_some >> | ||
| 97 | |||
| 98 | Requests search results from NCBI, adding the results to the WWW::Search object's cache. | ||
| 99 | |||
| 100 | =end private | ||
| 101 | |||
| 102 | =cut | ||
| 103 | |||
| 104 | sub native_retrieve_some { | ||
| 105 | my $self = shift; | ||
| 106 | |||
| 107 | return undef unless scalar (@{ $self->{'_article_ids'} || [] }); | ||
| 108 | my $ua = $self->user_agent(); | ||
| 109 | my $url = QUERY_ARTICLE_INFO_URI . '&id=' . join(',', splice(@{ $self->{'_article_ids'} },0,ARTICLES_PER_REQUEST)) . '&retmode=xml'; | ||
| 110 | warn 'Fetching URL: ' . $url if ($debug); | ||
| 111 | my $response = $ua->get( $url ); | ||
| 112 | if ($response->is_success) { | ||
| 113 | my $content = $response->content; | ||
| 114 | if ($debug) { | ||
| 115 | open (my $fh, ">/tmp/pubmed.article.info"); | ||
| 116 | print { $fh } $content; | ||
| 117 | close($fh); | ||
| 118 | warn "Saved response in /tmp/pubmed.article.info\n"; | ||
| 119 | } | ||
| 120 | my $doc = $self->{'_xml_parser'}->parse( $content ); | ||
| 121 | my $articles = $doc->getElementsByTagName('PubmedArticle'); | ||
| 122 | my $n = $articles->getLength; | ||
| 123 | warn "$n articles found\n" if ($debug); | ||
| 124 | my $count = 0; | ||
| 125 | foreach my $i (0 .. $n - 1) { | ||
| 126 | my $article = $articles->item( $i ); | ||
| 127 | my $id = ($article->getElementsByTagName('PMID')->item(0)->getChildNodes)[0]->getNodeValue(); | ||
| 128 | warn "$id\n" if ($debug); | ||
| 129 | my $title = ($article->getElementsByTagName('ArticleTitle')->item(0)->getChildNodes)[0]->getNodeValue(); | ||
| 130 | warn "\t$title\n" if ($debug); | ||
| 131 | my $url = 'http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=' . $id . '&dopt=Abstract'; | ||
| 132 | my @authors; | ||
| 133 | my $authornodes = $article->getElementsByTagName('Author'); | ||
| 134 | my $n = $authornodes->getLength; | ||
| 135 | foreach my $i (0 .. $n - 1) { | ||
| 136 | my ($author, $fname, $lname); | ||
| 137 | eval { | ||
| 138 | $author = $authornodes->item($i); | ||
| 139 | $lname = ($author->getElementsByTagName('LastName')->item(0)->getChildNodes)[0]->getNodeValue(); | ||
| 140 | $fname = substr( ($author->getElementsByTagName('ForeName')->item(0)->getChildNodes)[0]->getNodeValue(), 0, 1) . '.'; | ||
| 141 | }; | ||
| 142 | if ($@) { | ||
| 143 | warn $@ if ($debug); | ||
| 144 | next unless ($lname); | ||
| 145 | } else { | ||
| 146 | push(@authors, join(' ', $lname, $fname)); | ||
| 147 | } | ||
| 148 | } | ||
| 149 | my $author = join(', ', @authors); | ||
| 150 | warn "\t$author\n" if ($debug); | ||
| 151 | |||
| 152 | my $journal = $self->get_text_node( $article, 'MedlineTA' ); | ||
| 153 | my $page = $self->get_text_node( $article, 'MedlinePgn' ); | ||
| 154 | my $volume = $self->get_text_node( $article, 'Volume' ); | ||
| 155 | my $issue = $self->get_text_node( $article, 'Issue' ); | ||
| 156 | my $pmid = $self->get_text_node( $article, 'PMID' ); | ||
| 157 | my $abstract = $self->get_text_node( $article, 'AbstractText' ); | ||
| 158 | |||
| 159 | my @date; | ||
| 160 | { | ||
| 161 | my $date = $article->getElementsByTagName('PubDate')->item(0); | ||
| 162 | push(@date, $self->get_text_node( $date, 'Year' )); | ||
| 163 | push(@date, $self->get_text_node( $date, 'Month' )); | ||
| 164 | push(@date, $self->get_text_node( $date, 'Day' )); | ||
| 165 | } | ||
| 166 | |||
| 167 | my $hit = new WWW::Search::PubMed::Result; | ||
| 168 | |||
| 169 | my $source = ''; | ||
| 170 | my $date = join(' ', grep defined, @date); | ||
| 171 | $hit->date( $date ); | ||
| 172 | $hit->year( $date[0] ) if (defined($date[0])); | ||
| 173 | $hit->month( $date[1] ) if (defined($date[1])); | ||
| 174 | $hit->day( $date[2] ) if (defined($date[2])); | ||
| 175 | |||
| 176 | $source = "${journal}. " | ||
| 177 | . ($date ? "${date}; " : '') | ||
| 178 | . ($volume ? "${volume}" : '') | ||
| 179 | . ($issue ? "(${issue})" : '') | ||
| 180 | . ($page ? ":$page" : ''); | ||
| 181 | $source = "(${source})" if ($source); | ||
| 182 | warn "\t$source\n" if ($debug); | ||
| 183 | |||
| 184 | $hit->add_url( $url ); | ||
| 185 | $hit->title( $title ); | ||
| 186 | |||
| 187 | $hit->pmid( $pmid ); | ||
| 188 | $hit->abstract( $abstract ) if ($abstract); | ||
| 189 | |||
| 190 | my $desc = join(' ', grep {$_} ($author, $source)); | ||
| 191 | $hit->description( $desc ); | ||
| 192 | push( @{ $self->{'cache'} }, $hit ); | ||
| 193 | $count++; | ||
| 194 | warn "$count : $title\n" if ($debug); | ||
| 195 | } | ||
| 196 | return $count; | ||
| 197 | } else { | ||
| 198 | warn "Uh-oh." . $response->error_as_HTML(); | ||
| 199 | return undef; | ||
| 200 | } | ||
| 201 | |||
| 202 | } | ||
| 203 | |||
| 204 | =begin private | ||
| 205 | |||
| 206 | =item C<< get_text_node ( $node, $name ) | ||
| 207 | |||
| 208 | Returns the text contained in the named descendent of the XML $node. | ||
| 209 | |||
| 210 | =end private | ||
| 211 | |||
| 212 | =cut | ||
| 213 | |||
| 214 | sub get_text_node { | ||
| 215 | my $self = shift; | ||
| 216 | my $node = shift; | ||
| 217 | my $name = shift; | ||
| 218 | my $text = eval { ($node->getElementsByTagName($name)->item(0)->getChildNodes)[0]->getNodeValue() }; | ||
| 219 | if ($@) { | ||
| 220 | warn "XML[$name]: $@" if ($debug); | ||
| 221 | return undef; | ||
| 222 | } else { | ||
| 223 | warn "XML[$name]: $text\n" if ($debug); | ||
| 224 | return $text; | ||
| 225 | } | ||
| 226 | } | ||
| 227 | |||
| 228 | 1; | ||
| 229 | |||
| 230 | __END__ | ||