[BACK]Return to errata_scraper.pl CVS log [TXT][DIR] Up to [local] / openbsd / errata_scraper

Diff for /openbsd/errata_scraper/errata_scraper.pl between version 1.1 and 1.4

version 1.1, 2011/03/21 17:26:58 version 1.4, 2011/05/29 02:04:30
Line 1 
Line 1 
 #!/usr/bin/perl  #!/usr/bin/perl
 # $AFresh1$  # $AFresh1: errata_scraper.pl,v 1.3 2011/03/23 18:46:16 andrew Exp $
 ########################################################################  ########################################################################
 # Copyright (c) 2011 Andrew Fresh <andrew@afresh1.com>  # Copyright (c) 2011 Andrew Fresh <andrew@afresh1.com>
 #  #
Line 18 
Line 18 
 use strict;  use strict;
 use warnings;  use warnings;
   
 use Mojo::Client;  use Mojo::UserAgent;
 my $client = Mojo::Client->new;  
   my $base_uri = 'http://www.openbsd.org/';
 $client->get(  
     'http://www.openbsd.org/errata48.html' => sub {  my $ua = Mojo::UserAgent->new;
         shift->res->dom('li')->each(  
             sub {  my $ls = $ua->get( $base_uri . 'errata.html' )->res->dom('a[href^="errata"]');
                 my $e = shift;  
   foreach my $l ( @{$ls}[ -2, -1 ] ) {
                 my $patch = $e->at('a[href$=".patch"]')->attrs->{href};      print 'Errata for OpenBSD ', $l->text, "\n";
       foreach my $e (
                 my $title = $e->at('strong')->replace('')->all_text;          reverse @{ $ua->get( $base_uri . $l->attrs->{'href'} )->res->dom('li')
                 $title =~ s/\s+/ /gxms;          } )
       {
                 my $arch = $e->at('i')->replace('')->all_text;          my $patch = $e->at('a[href$=".patch"]')->replace('')->{href};
                 $arch =~ s/\s+/ /gxms;          my $title = $e->at('strong')->replace('')->all_text;
           my $arch  = $e->at('i')->replace('')->all_text;
                 # the li ends at p, but the parser expects a /li          my $descr = $e->all_text;
                 $e->at('p')->replace('');          $descr =~ s/\s+/ /gs;
           $descr =~ s/\s(\.(?:\s|$))/$1/gs;
                 my $descr = $e->all_text;          $descr =~ s/\.+$/./gs;
                 $descr =~ s/\s+/ /gxms;  
                 $descr =~ s/^\s+|\s+$//gxms;          print 'Title: ', $title, "\n";
           print 'Arch:  ', $arch,  "\n";
                 print 'Title: ', $title, "\n";          print 'Patch: ', $patch, "\n";
                 print 'Arch: ',  $arch,  "\n";          print 'Descr: ', $descr, "\n";
                 print 'Patch: ', $patch, "\n";          print "\n";
                 print 'Descr: ', $descr, "\n";  
                 print "\n";  
             }  
         );  
     }      }
 );  }
   
 $client->start;  

Legend:
Removed from v.1.1  
changed lines
  Added in v.1.4

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>