2 # get_pdf tries to get pdfs, and is released
3 # under the terms of the GPL version 2, or any later version, at your
4 # option. See the file README and COPYING for more information.
5 # Copyright 2008 by Don Armstrong <don@donarmstrong.com>.
6 # $Id: perl_script 1352 2009-01-25 02:04:38Z don $
17 get_pdf - try to get a pdf
21 get_pdf [options] reference [references]
24 --debug, -d debugging level (Default 0)
25 --help, -h display this help
26 --man, -m display manual
34 The reference is a pmid
36 =item B<--cgi-proxy, -C>
38 Use this cgi proxy style proxy
42 Debug verbosity. (Default 0)
46 Display brief usage information.
66 my %options = (debug => 0,
72 my %REFERENCE_TYPES = (pmid => 'pmid|p');
75 values %REFERENCE_TYPES,
76 'use_links|use-links!',
77 'cgi_proxy|cgi-proxy|C=s',
78 'http_proxy|http-proxy|H=s',
79 'debug|d+','help|h|?','man|m');
81 pod2usage() if $options{help};
82 pod2usage({verbose=>2}) if $options{man};
84 $DEBUG = $options{debug};
88 if (not grep {exists $options{$_} and
89 defined $options{$_} and
90 $options{$_}} keys %REFERENCE_TYPES) {
94 if (grep {exists $options{$_}
95 and defined $options{$_}
96 and $options{$_}} keys %REFERENCE_TYPES > 1) {
97 push @USAGE_ERRORS,"You can only specify exactly one of the ".(map { "--$_"} keys %REFERENCE_TYPES)." options";
101 push @USAGE_ERRORS,"You must specify at least one reference";
104 pod2usage(join("\n",@USAGE_ERRORS)) if @USAGE_ERRORS;
107 if (exists $options{http_proxy}) {
108 $ENV{http_proxy} = $options{http_proxy};
109 $ENV{HTTP_PROXY} = $options{http_proxy};
110 $ENV{CGI_HTTP_PROXY} = $options{http_proxy};
113 if ($options{pmid}) {
114 my $m = WWW::Mechanize->new(agent => 'Mozilla',cookie_jar => {});
115 for my $pmid (@ARGV) {
117 next unless length $pmid;
118 my $url = "www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=${pmid}&dopt=Abstract";
119 if (exists $options{cgi_proxy}) {
120 $url = $options{cgi_proxy}.$url;
122 $url = "http://${url}";
124 $m->get($url) or die "Unable to get $url";
125 my $orig_mech = $m->clone();
127 my @possible_links = $m->find_all_links(text_regex => qr/to\s*read/i);
128 # try to find the other links
129 push @possible_links,
130 grep {my $attr = $_->attrs(); exists $attr->{title} and $attr->{title} =~ qr/(?:Full\s*Text|PMC)/i}
132 print STDERR map {"article link: ".$_->url_abs()."\n"} @possible_links if $DEBUG;
133 die "No links" unless @possible_links;
137 print "trying ".$possible_links[0]->url()."\n" if $DEBUG;
138 $m->get($possible_links[0]->url()) or
139 die "Unable to follow link";
140 # try to find pdf link
141 my $pdf_m = find_pdf_link($m) or
142 die "Unable to find pdf";
143 my $fh = IO::File->new($pmid.'.pdf','w') or
144 die "Unable to open ${pmid}.pdf for writing: $!";
145 print {$fh} $pdf_m->content or
146 die "Unable to write to ${pmid}.pdf: $!";
148 die "Unable to close ${pmid}.pdf filehandle: $!";
150 shift @possible_links;
151 } while ($@ and @possible_links);
157 print STDERR "$@\n" if $DEBUG;
158 if ($options{use_links}) {
160 # links2 doesn't like the leading http:// of proxies for some reason
161 exists $options{http_proxy}?('-http-proxy',(map {s{http://}{}; $_} $options{http_proxy})):(),
164 rename('temp.pdf',"${pmid}.pdf") if -e 'temp.pdf';
171 sub check_subframes {
173 my @sub_frames = $m->find_all_links(tag_regex=>qr/^i?frame$/);
174 print STDERR "subframes: \n" if $DEBUG;
175 p @sub_frames if $DEBUG;
176 for my $frame (@sub_frames) {
177 my $r = $m->get($frame->url_abs());
178 print STDERR "trying: ".$frame->url_abs()."\n" if $DEBUG;
179 if ($r->header('Content-Type') =~ /pdf/) {
182 print STDERR "failed: ".$r->header('Content-Type')."\n" if $DEBUG;
184 for my $frame (@sub_frames) {
185 my $r = $m->get($frame->url_abs());
186 my $pdf_m = find_pdf_link($m,
190 if (defined $pdf_m) {
199 my ($mech,$guess,$call) = @_;
200 $guess = 1 unless defined $guess;
201 $call = 0 unless defined $call;
202 # avoid looping endlessly
203 return undef if $call > 5;
204 my $m = $mech->clone();
205 if ($m->content =~ /select\s*a\s*website\s*below/i) {
206 print STDERR $m->uri() if $DEBUG;
207 print STDERR $m->content() if $DEBUG > 1;
208 my @inputs = $m->find_all_inputs(type => 'hidden',
209 name => q(urls['sd']),
211 return unless @inputs;
212 $m->get($inputs[0]->value);
213 print STDERR $m->content() if $DEBUG > 1;
216 # this brings forward the actual link at Science
217 push @possible_links,
218 grep {my $temp = $_->attrs();
219 exists $temp->{rel} and $temp->{rel} =~ qr/view-/i and
220 defined $_->text() and $_->text() =~ qr/Full\s*Text.*PDF/i
222 $m->find_all_links(text_regex => qr/PDF/i);
223 # this is to prioritize the real link at science direct
224 push @possible_links,
225 grep {my $temp = $_->attrs();
227 print STDERR Dumper($temp);
228 (exists $temp->{title} and $temp->{title} =~ qr/(Download|Full\s*Text)\s*PDF/i) or
229 (defined $_->text() and $_->text() =~ qr/(Full\s*Text|Download).*PDF/i)
231 $m->find_all_links(text_regex => qr/PDF/i);
232 my $possible_links = 0;
235 print STDERR "possible links[$possible_links]:\n";
238 push @possible_links, grep { $_->url_abs() !~ /_orig(?:in)?=article/} $m->find_all_links(text_regex => qr/PDF/i);
241 print STDERR "possible links[$possible_links]:\n";
244 push @possible_links, $m->find_all_links(tag_regex => qr/meta/,
245 url_regex => qr/(reprint|\.pdf)/i,
249 print STDERR "possible links[$possible_links]:\n";
252 # The masthead grep here is to handle PNAS, which has a link to their masthead in every article.
253 push @possible_links,
254 grep {my $temp = $_->attrs(); (not defined $temp->{title}) or $temp->{title} !~ qr/Masthead/i;}
255 $m->find_all_links(text_regex => qr/pdf/i);
258 print STDERR "possible links[$possible_links]:\n";
261 push @possible_links,$m->find_all_links(text_regex => qr/manual\s*download/i);
264 print STDERR "possible links[$possible_links]:\n";
267 print STDERR $m->uri() if $DEBUG;
268 print STDERR $m->content() if $DEBUG > 1;
269 print STDERR map{"possible pdf link: ".$_->url_abs().qq(\n)} @possible_links if $DEBUG;
270 if (not @possible_links and $DEBUG) {
271 print STDERR $m->content();
273 my $best_guess = $possible_links[0] if @possible_links;
274 for my $link (@possible_links) {
275 print STDERR "trying ".$link->url_abs()."..." if $DEBUG;
276 my $r = $m->get($link->url_abs());
277 my $content = $m->content();
278 if ($r->header('Content-Type') =~ /pdf/) {
279 print STDERR "success\n" if $DEBUG;
282 my $ret = check_subframes($m,$call);
283 return $ret if defined $ret;
284 print STDERR "failure; content type ".$r->header('Content-Type')."\n" if $DEBUG;
285 print STDERR $content if $DEBUG;
287 my $ret = check_subframes($m,$call);
288 return $ret if defined $ret;
289 # if ($guess and defined $best_guess) {
290 # $m->get($best_guess->url_abs());