git.rot13.org Git - webpac/blob - feeds/sciencedirect2.pl

   1 #!/usr/bin/perl -w
   2
   3 # This script will fatch list of articles on which you have access
   4 # (using IP authorisation) from ScienceDirect
   5 #
   6 # This version requires CSV dumps from ScienceDirect for Holdings data
   7 # and categories, but can output much more data about each record
   8
   9 use LWP::UserAgent;
  10 use HTML::TreeBuilder;
  11 require Text::CSV;
  12 use Text::Unaccent;
  13 use strict;
  14
  15 my $debug=1;
  16
  17 # configure ScienceDirect CVS files location
  18 my $csv_dir="/data/isis_data/sciencedirect";
  19 my $j_holdings="sd_JournalHoldingsRpt.txt";
  20 my $j_category="sd_Journal_Category.txt";
  21
  22 # URL to list of subscribed journals
  23 my $url = 'http://www.sciencedirect.com/science?_ob=JournalListURL&_type=subscribed&_stype=title&subjColl=all&_auth=y&_update=y&_frameSeg=M&_title=all&_acct=C000050661&_version=1&_urlVersion=0&_userid=1034703&md5=6d4b6e263318a1d7d2a3b523d861f920';
  24 my $html_codepage="iso-8859-1";
  25
  26 my $csv = Text::CSV->new();
  27 my $journal;
  28 my $c_wo_h = 0;         # category without holding record
  29 my $c_nr = 0;           # number of categories assigned
  30
  31 my $j_basic = 0;
  32 my $j_detailed = 0;
  33
  34 print STDERR "unrolling $j_holdings\n";
  35
  36 sub nuc {
  37         # normalizing UC
  38         my $s=shift @_ || return "";
  39         $s=unac_string($html_codepage,$s);
  40         $s=~s/[^\w]/ /g;
  41         $s=~s/  +/ /g;
  42         return uc($s);
  43 }
  44
  45 open(H,"$csv_dir/$j_holdings") || die "can't open $csv_dir/$j_holdings: $!";
  46 my $line = <H>;         # skip header line
  47 while(<H>) {
  48         chomp;
  49         $csv->parse($_) || warn "can't parse '$_': ".$csv->error_input;
  50         my @data = $csv->fields;
  51         my $key = nuc($data[0]);
  52         push @data,"";                  # for categories later...
  53         $journal->{$key} = \@data;
  54 }
  55 close(H);
  56
  57 print STDERR "unrolling $j_category\n";
  58
  59 open(C,"$csv_dir/$j_category") || die "can't open $csv_dir/$j_category: $!";
  60 $line = <C>;            # skip header line
  61 while(<C>) {
  62         chomp;
  63         $csv->parse($_) || warn "can't parse '$_': ".$csv->error_input;
  64         my @data = $csv->fields;
  65         my $key = nuc($data[1]);
  66         if (! $journal->{$key}) {
  67                 $c_wo_h++;
  68                 next;
  69         }
  70
  71         foreach my $i (4, 6, 8, 10) {
  72                 push @{$journal->{$key}},$data[$i] || "";
  73                 if ($data[$i]) {
  74                         $c_nr++;
  75                 }
  76         }
  77 }
  78 close(C);
  79
  80 print STDERR "$c_nr categories assigned, $c_wo_h categories with holdings\n";
  81
  82 $debug++ if (lc($ARGV[0]) eq "-d");
  83
  84 my $ua = new LWP::UserAgent;
  85 $ua->agent("Mjesec educational harvester -- contact mglavica\@ffzg.hr 0.0");
  86 $ua->timeout(60);
  87 #$ua->env_proxy();
  88 #$ua->proxy(['http', 'ftp'], 'http://proxy.carnet.hr:8001/');
  89
  90 print STDERR "getting '$url'...\n" if ($debug);
  91 my $req = HTTP::Request->new(GET => $url);
  92
  93 my @out;
  94
  95 my $res = $ua->request($req);
  96 if ($res->is_success) {
  97         print STDERR "parsing html...\n" if ($debug);
  98         my $tree = HTML::TreeBuilder->new;
  99 #       $tree->parse_file("list.html");   # !
 100         $tree->parse($res->content);
 101
 102         foreach my $tr ($tree->look_down('_tag', 'tr')) {
 103                 my $link;
 104                 if ($link = $tr->look_down('_tag','a')) {
 105                         if ($link->attr('href') =~ m{/science\?_ob=JournalURL}) {
 106                                 my $j=nuc($link->as_text);
 107                                 if ($journal->{$j}) {
 108                                         my $i=0;
 109                                         foreach my $line (@{$journal->{$j}}) {
 110                                                 print $i++,": $line\n";
 111                                         }
 112                                         $j_detailed++;
 113                                 } else {
 114                                         print "0: ",$link->as_text."\n";
 115                                         print "7: http://www.sciencedirect.com",$link->attr('href')."\n";
 116                                         $j_basic++;
 117                                         print STDERR "can't find details for $j\n" if ($debug);
 118                                 }
 119
 120                                 print "\n";
 121                         }
 122                 }
 123         }
 124
 125         $tree->delete; # clear memory!
 126
 127 } else {
 128     warn "can't fetch web page from '$url'";
 129 }
 130
 131 print STDERR "Processed ",($j_basic+$j_detailed)," journals, $j_basic with basic data and $j_detailed detailed\n";
 132