User:HBC Searchbot/Source

From Wikipedia, the free encyclopedia

See also templates.pl

use strict;
use LWP::UserAgent;
use HTTP::Request::Common;
use XML::Simple;
use URI::Escape;

my %revisions = get_complete_history('User talk:HighInBC');
open(PL,'templates.pl');
sysread(PL,my $templates,-s(PL));
close(PL);
my(%templates);
eval($templates) || die;

my $searches;
warn ("Checking ".scalar(keys(%revisions))." revisions.\n");
my(%warnings);
foreach my $revision (sort {$a <=> $b} keys(%revisions))
  {
  my $rh_rev = $revisions{$revision};
  my $text = ${$rh_rev}{'text'}{'content'};
  foreach my $template (sort keys(%templates))
    {
    $searches++;
    my $start = index(lc($text), lc($templates{$template}));
    if ($start > 0)
      {
      my $end = (index($text,"\n",$start) - $start-1);
      $end = (length($text)-$start) if ($end < 0);
      my $string = substr($text,$start,$end);
      $warnings{$string}{'template'}    = $template;
      $warnings{$string}{'regex'}       = $templates{$template};
      push(@{$warnings{$string}{'revisions'}},$revision);
      }
    }
  }
warn "$searches searches performed.\n";
warn Dumper(\%warnings);

sub get_complete_history
  {
  mkdir('cache') unless (-d('cache'));
  my $page = shift;
  my(%revisions);
  my $count;
  my $offset;
  my $fname = 'cache/'.uri_escape($page);
  if (-f($fname))
    {
    warn "found $fname in cache, loading\n";
    open(IN,$fname);
    sysread(IN,my $code,-s(IN));
    close(IN);
    my $VAR1; eval($code); %revisions = %{$VAR1};
    my(@keys) = sort {$a <=> $b} keys(%revisions);
    $offset = ($revisions{$keys[scalar(@keys)-1]}{'timestamp'});
    warn (scalar(keys(%revisions))." loaded from cache.\n");
    }
  else
    {
    warn "No cache, starting fresh.\n";
    $offset = '0';
    }
  my $total;
  GETMORE:
  warn "Downloading 100 revisions.\n";
  my $ua = LWP::UserAgent->new('agent' => 'HighInBC warning checker .01b');
  my $index = 'http://en.wikipedia.org/w/index.php';
  my $res = $ua->request
        (
         POST $index."?title=Special:Export",
         Content_Type  => 'application/x-www-form-urlencoded',
         Content       =>    [(
                                  'pages'       => $page,
                                  'action'      => 'submit',
                                  'submit'      => 'Export',
                                  'limit'       => 100,
                                  'offset'      => $offset
                                )]
        );
  my $current = $res->content();
  unless ($current =~ m|^<mediawiki|)
    {
    warn "Failed somehow, trying again\n";
    goto GETMORE;
    }
  my $index = rindex($current, '<timestamp>');
  my $string = substr($current,$index,43);
  $string =~ m|<timestamp>(.+?)</timestamp>|;
  $offset = $1;
  my $xml_data = XMLin($current);
  $count = 0;
  if (${$xml_data}{'page'}{'revision'}{'timestamp'} eq $offset)
    {
    # do nothing
    }
  elsif (${$xml_data}{'page'}{'revision'}{'comment'})
    {
    ($count++ && $total++) unless ($revisions{${$xml_data}{'page'}{'revision'}{'id'}});
    $revisions{${$xml_data}{'page'}{'revision'}{'id'}} = ${$xml_data}{'page'}{'revision'};
    }
  else
    {
    foreach my $revision (sort {$a <=> $b} keys(%{${$xml_data}{'page'}{'revision'}}))
      {
      ($count++ && $total++) unless ($revisions{$revision});
      $revisions{$revision} = ${$xml_data}{'page'}{'revision'}{$revision};
      }
    }
  warn "Got $count revisions\n";
  if ($count == 100)
    {
    warn "Still more past $offset to get, waiting 5 seconds between hits\n";
    sleep(5);
    goto GETMORE;
    }
  if ($total > 0)
    {
    warn "Saving cache...\n";
    open(OUT, '>'.$fname);
    print OUT (Dumper(\%revisions));
    close(OUT);
    warn "done.\n";
    }
  return %revisions;
  }