/[swish]/trunk/make_config.pl
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Diff of /trunk/make_config.pl

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 33 by dpavlin, Sun May 4 01:31:31 2003 UTC revision 53 by dpavlin, Tue Jan 20 18:41:38 2004 UTC
# Line 6  Line 6 
6  #  #
7    
8  use strict;  use strict;
9    use Getopt::Long;
10    
11    my $stripurl;
12    
13    my $result = GetOptions ("stripurl=s" => \$stripurl);
14    
15  my $name = shift @ARGV;  my $name = shift @ARGV;
16  my $url = shift @ARGV;  my $url = shift @ARGV;
17  my $stripurl = shift @ARGV;  my $title = join(" ",@ARGV);
18    
19  if (!$name || !$url) {  if (!$name || !$url) {
20          print "Usage: $0 name URL [strip from url]\n";          print "Usage: $0 [--stripurl=path] name URL title\n";
21          exit;          exit;
22  }  }
23    
# Line 27  my ($host,$urlpath); Line 32  my ($host,$urlpath);
32    
33  if ($url =~ m#(http://[^/]+)(/.*)$#) {  if ($url =~ m#(http://[^/]+)(/.*)$#) {
34          ($host,$urlpath) = ($1,$2);          ($host,$urlpath) = ($1,$2);
35          if ($stripurl && $urlpath =~ m/^(.*$stripurl)/) {          if ($stripurl && $url =~ m/^(.*$stripurl)/) {
36                  $stripurl = $1;                  $stripurl = $1;
37          } else {          } else {
38                  $stripurl = "";                  $stripurl = $host;
39          }          }
40  } else {  } else {
41          print "Can't parse URL $url into host and path!\n";          print "Can't parse URL $url into host and path!\n";
# Line 41  my $baseurl = $url; Line 46  my $baseurl = $url;
46  $baseurl =~ s#/[^/]+$#/#g;  $baseurl =~ s#/[^/]+$#/#g;
47    
48  print "Config name: $name\nCrawling url: $url [strip $stripurl]\n";  print "Config name: $name\nCrawling url: $url [strip $stripurl]\n";
49    print "Title: $title\n" if ($title);
50    
51    
52  my $config = <<"EOF";  my $config = <<"EOF";
# Line 49  my $config = <<"EOF"; Line 55  my $config = <<"EOF";
55  IncludeConfigFile $dir/common.config  IncludeConfigFile $dir/common.config
56    
57  # this is a cludge to implement no parent URL feature in swish indexer  # this is a cludge to implement no parent URL feature in swish indexer
58  IndexDir "$url $baseurl"  IndexDir "$baseurl $url"
59  ReplaceRules replace " $baseurl" ""  ReplaceRules replace "$baseurl " ""
60  ReplaceRules replace "${host}${stripurl}" ""  ReplaceRules replace "${stripurl}" ""
61    
62  IndexFile $dir/index/$name  IndexFile $dir/index/$name
63    
# Line 71  my $xml = << "EOF"; Line 77  my $xml = << "EOF";
77          prog="/usr/bin/swish-e"          prog="/usr/bin/swish-e"
78          index="$dir/index/$name"          index="$dir/index/$name"
79          charset="iso-8859-2"          charset="iso-8859-2"
         title="$name search"  
80          affix="/usr/lib/ispell/american.aff"          affix="/usr/lib/ispell/american.aff"
81          url="$stripurl"  EOF
82    if ($title) {
83            $xml .= "\ttitle=\"$title\"\n";
84    } else {
85            $xml .= "\ttitle=\"$name search\"\n";
86    }
87    
88    my $xml_title = $title || $url;
89    
90    my $xml_url = $url;
91    $xml_url =~ s/^$host//;
92    my $xml_urlprefix = $stripurl;
93    $xml_urlprefix =~ s/^$host//;
94    
95    $xml .= "\turl=\"$xml_urlprefix\"\n" if ($stripurl);
96    $xml .= << "EOF";
97          >          >
98          <labels>          <labels>
99                  <label value="10"> 10</label>                  <label value="10"> 10</label>
# Line 88  my $xml = << "EOF"; Line 108  my $xml = << "EOF";
108                  <no_properties>don't display results details</no_properties>                  <no_properties>don't display results details</no_properties>
109                  <hits>Showing %s documents (of maximum %d)... (%s)</hits>                  <hits>Showing %s documents (of maximum %d)... (%s)</hits>
110                  <no_hits>Can't find any documents (%s, %s)</no_hits>                  <no_hits>Can't find any documents (%s, %s)</no_hits>
111                  <footer>Searcher will try to create different variations of words using spelling dictionary. If you enter a word with minus (-) before it, it will exclude documents with this word and plus (+) will do the opposite (e.g. -work +play)</footer>                  <footer><![CDATA[
112                    Searcher will try to create different variations of words
113                    using spelling dictionary. If you enter a word with minus
114                    (<tt>-</tt>) before it, it will exclude documents with this
115                    word and plus (<tt>+</tt>) will do the opposite
116                    (e.g. <tt>-work +play</tt>)
117                    <p><small>Content indexed:
118                    <a href="$xml_url">$xml_title</small></p>
119                    ]]></footer>
120          </text>          </text>
121          <!-- additional data to show summary -->          <!-- additional data to show summary -->
122          <properties>swishdescription swishdocsize</properties>          <properties>swishdescription swishdocsize</properties>
# Line 99  my $xml = << "EOF"; Line 127  my $xml = << "EOF";
127  </config>  </config>
128  EOF  EOF
129    
130  print "Creating $dir/$name.config\n";  print "Creating $dir/config/$name\n";
131  open(C,"> $dir/$name.config") || die "can't open $dir/$name.config: $!";  open(C,"> $dir/config/$name") || die "can't open $dir/config/$name: $!";
132  print C $config;  print C $config;
133  close(C);  close(C);
134    
# Line 112  close(C); Line 140  close(C);
140  print "Creating symlink to $dir/html/$name.cgi\n";  print "Creating symlink to $dir/html/$name.cgi\n";
141  symlink "$dir/html/swish.cgi","$dir/html/$name.cgi" || die "can't create symlink $dir/html/$name.cgi: $!";  symlink "$dir/html/swish.cgi","$dir/html/$name.cgi" || die "can't create symlink $dir/html/$name.cgi: $!";
142    
143  #print "Index this with:\nswish-e -S http -c $dir/$name.config\n";  #print "Index this with:\nswish-e -S http -c $dir/config/$name\n";
144  #print "Search using $name.cgi\n";  #print "Search using $name.cgi\n";
Legend:
Removed from v.33  
changed lines
  Added in v.53
  ViewVC Help
Powered by ViewVC 1.1.26