View | Details | Raw Unified | Return to bug 7963
Collapse All | Expand All

(-)a/misc/cronjobs/checkurl (-1 / +179 lines)
Line 0 Link Here
0
- 
1
#!/usr/bin/perl
2
3
# Copyright 2012 Tamil s.a.r.l.
4
#
5
# This file is part of Koha.
6
#
7
# Koha is free software; you can redistribute it and/or modify it under the
8
# terms of the GNU General Public License as published by the Free Software
9
# Foundation; either version 2 of the License, or (at your option) any later
10
# version.
11
#
12
# Koha is distributed in the hope that it will be useful, but WITHOUT ANY
13
# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
14
# A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
15
#
16
# You should have received a copy of the GNU General Public License along
17
# with Koha; if not, write to the Free Software Foundation, Inc.,
18
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
19
20
use 5.010;
21
use utf8;
22
use strict;
23
use warnings;
24
use Pod::Usage;
25
use Getopt::Long;
26
use C4::Context;
27
use C4::Biblio;
28
use AnyEvent;
29
use AnyEvent::HTTP;
30
31
my ($verbose, $help, $html) = (0, 0, 0);
32
my ($host, $host_pro)       = ('', '');
33
my ($timeout, $maxconn)     = (10, 200);
34
my $uriedit                 = "/cgi-bin/koha/cataloguing/addbiblio.pl?biblionumber=";
35
GetOptions(
36
    'verbose'       => \$verbose,
37
    'html'          => \$html,
38
    'help'          => \$help,
39
    'host=s'        => \$host,
40
    'host-pro=s'    => \$host_pro,
41
    'timeout=i'     => \$timeout,
42
    'maxconn=i'     => \$maxconn,
43
);
44
45
46
sub usage {
47
    pod2usage( -verbose => 2 );
48
    exit;
49
}
50
51
52
# Check all URLs from all current Koha biblio records
53
54
sub check_all_url {
55
    my $sth = C4::Context->dbh->prepare(
56
        "SELECT biblionumber FROM biblioitems WHERE url <> '' LIMIT 20" );
57
    $sth->execute;
58
59
    my $countconn = 0;
60
    my $cv = AnyEvent->condvar;
61
    say "<html>\n<body>\n<table>" if $html;
62
    my $idle = AnyEvent->idle(
63
        cb => sub {
64
            return if $countconn > $maxconn;
65
            while ( my ($biblionumber) = $sth->fetchrow ) {
66
                my $record = GetMarcBiblio( $biblionumber );
67
                next unless $record->field('856');
68
                foreach my $field ( $record->field('856') ) {
69
                    my $url = $field->subfield('u');
70
                    next unless $url;
71
                    $url = "$host/$url" unless $url =~ /^http/;
72
                    $countconn++;
73
                    http_request(
74
                        HEAD => $url,
75
                        headers => { 'user-agent' => 'Mozilla/5.0 (compatible; U; Koha checkurl)' },
76
                        timeout => $timeout,
77
                        sub {
78
                            my (undef, $hdr) = @_;
79
                            $countconn--;
80
                            if ( $hdr->{Status} !~ /^2/ || $verbose ) {
81
                                print $html
82
                                ? "<tr>\n<td><a href=\"" .
83
                                  $host_pro . $uriedit . $biblionumber.
84
                                  "\">$biblionumber</a>" .
85
                                  "</td>\n<td>" . $url . "</td>\n<td>" .
86
                                  "$hdr->{Status} $hdr->{Reason}</td>\n</tr>\n\n"
87
                                : "$biblionumber\t" . $url . "\t" .
88
                                  "$hdr->{Status} $hdr->{Reason} \n";
89
                            }
90
                        },
91
                    );
92
                }
93
                return;
94
            }
95
            $cv->send;
96
        }
97
    );
98
    $cv->recv;
99
    $idle = undef;
100
101
    # Give $timeout to pending requests to send their result
102
    $cv = AnyEvent->condvar;
103
    my $timer = AnyEvent->timer( after => $timeout + 1, cb => sub { $cv->send } );
104
    $cv->recv;
105
    say "</table>\n</body>\n</html>" if $html;
106
}
107
108
109
usage() if $help;
110
111
if ( $html && !$host_pro ) {
112
    if ( $host ) {
113
        $host_pro = $host;
114
    }
115
    else {
116
        say "Error: host-pro parameter or host must be provided in html mode";
117
        exit;
118
    }
119
}
120
121
check_all_url();
122
123
124
=head1 NAME
125
126
checkurl - Check URLs from 856$u field.
127
128
=head1 USAGE
129
130
=over
131
132
=item checkurl [--verbose|--help] [--host=http://default.tld]
133
134
Scan all URLs found in 856$u of bib records and display if resources are
135
available or not. HTTP requests are sent in parallel for efficiency.
136
137
=back
138
139
=head1 PARAMETERS
140
141
=over
142
143
=item B<--host=http://default.tld>
144
145
Server host used when URL doesn't have one, ie doesn't begin with 'http:'.
146
For example, if --host=http://www.mylib.com, then when 856$u contains
147
'img/image.jpg', the url checked is: http://www.mylib.com/image.jpg'.
148
149
=item B<--verbose|-v>
150
151
Outputs both successful and failed URLs.
152
153
=item B<--html>
154
155
Formats output in HTML. The result can be redirected to a file
156
accessible by http. This way, it's possible to link directly to biblio
157
record in edit mode. With this parameter B<--host-pro> is required.
158
159
=item B<--host-pro=http://koha-pro.tld>
160
161
Server host used to link to biblio record editing page.
162
163
=item B<--timeout=10>
164
165
Timeout for fetching URLs. By default 10 seconds.
166
167
=item B<--maxconn=1000>
168
169
Number of simulaneous HTTP requests. By default 200 connexions.
170
171
=item B<--help|-h>
172
173
Print this help page.
174
175
=back
176
177
=cut
178
179

Return to bug 7963