/[webpac]/trunk2/all2all.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 /trunk2/all2all.pl

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

revision 490 by dpavlin, Sat Oct 9 21:44:25 2004 UTC revision 550 by dpavlin, Tue Oct 26 21:44:57 2004 UTC
# Line 11  use locale; Line 11  use locale;
11  use YAML;  use YAML;
12  use Carp;  use Carp;
13  use Getopt::Long;  use Getopt::Long;
14    use Text::Unaccent 1.02;
15    
16  use lib './lib';  use lib './lib';
17  use WebPAC;  use WebPAC;
# Line 32  my $result = GetOptions( Line 33  my $result = GetOptions(
33          "low_mem!"      => \$low_mem,          "low_mem!"      => \$low_mem,
34  );  );
35    
36    my $filter = {
37            'CROVOC' => sub {
38                    my $tmp = shift || return;
39                    return undef unless ($tmp =~ s/\s*CROVOC.*$/ */);
40                    return $tmp;
41            },
42            'CROVOC_tree' => sub {
43                    my $tmp = shift || return;
44                    $tmp =~ s/\s*CROVOC.*$/ */;
45                    $tmp =~ s/\s*EUROVOC.*//;
46                    return $tmp;
47            },
48    };
49    
50    ## remove accented characters
51    #
52    sub unac {
53            my $string = shift || return;
54            $string = unac_string($code_page,$string);
55            $string =~ tr/đĐ/dD/;
56            $string = unac_entities($string);
57            return $string;
58    }
59    sub unac_2 {
60            my $string = shift || return;
61            if (length($string) > 2) {
62                    my $pr = substr($string,0,2);
63                    $string = unac_string($code_page,substr($string,2));
64                    $string =~ tr/đĐ/dD/;
65                    $string = $pr . $string;
66            }
67            $string = unac_entities($string);
68            return lc($string);
69    }
70    
71    sub unac_entities {
72            my $ent = shift || return;
73    
74            $ent =~ s/&(\w)(acute|cedil|circ|grave|ring|slash|tilde|uml);/$1/gi;
75            $ent =~ s/ð/d/g;
76            $ent =~ s/&E[tT][hH];/D/g;
77            $ent =~ s/&(\w\w)lig;/$1/gi;
78    
79            return $ent;
80    }
81    
82  # create WebPAC object  # create WebPAC object
83  #  #
84  my $webpac = new WebPAC(  my $webpac = new WebPAC(
# Line 40  my $webpac = new WebPAC( Line 87  my $webpac = new WebPAC(
87          start_mfn => $start_mfn,          start_mfn => $start_mfn,
88          debug => $debug,          debug => $debug,
89          low_mem => $low_mem,          low_mem => $low_mem,
90            filter => $filter,
91  ) || die;  ) || die;
92    
93  my $log = $webpac->_get_logger() || die "can't get logger";  my $log = $webpac->_get_logger() || die "can't get logger";
# Line 58  $|=1; Line 106  $|=1;
106  my $maxmfn = $webpac->open_isis(  my $maxmfn = $webpac->open_isis(
107          filename => shift @ARGV || '/data/hidra/THS/THS',          filename => shift @ARGV || '/data/hidra/THS/THS',
108          lookup => [          lookup => [
109          { 'key' => 'd:v900', 'val' => 'v250^a' },          { 'key' => 'd:v900', 'val' => 'filter{CROVOC_tree}v250^a v800' },
110  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
111  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
112  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
# Line 123  while (my $rec = $webpac->fetch_rec) { Line 171  while (my $rec = $webpac->fetch_rec) {
171                  $words =~ s/\W*\s+\W*/ /g;                  $words =~ s/\W*\s+\W*/ /g;
172                  $words =~ s/\W+$//;                  $words =~ s/\W+$//;
173    
174                    # first try to generate headline for this entry from index
175                    my $h = $ds->{'index'}->[0];
176                    # then, from display
177                    $h ||= $ds->{'display'}->[0];
178                    # and as last resport, fallback to headline
179                    $h ||= $headline;
180    
181                  $index->insert(                  $index->insert(
182                          index_name => $ds->{'tag'},                          index_name => $ds->{'tag'},
183                          #path => $f,                          #path => $f,
184                          path => $webpac->mfn,                          path => $webpac->mfn,
185                          headline => $headline,                          headline => $h,
186                          words => $words,                          words => unac($words),
187                  );                  );
188          }          }
189    
# Line 136  while (my $rec = $webpac->fetch_rec) { Line 191  while (my $rec = $webpac->fetch_rec) {
191          foreach my $ds (@ds) {          foreach my $ds (@ds) {
192                  next if (! $ds->{'index'});                  next if (! $ds->{'index'});
193    
194                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index;                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index( name => $ds->{'tag'} );
195    
196                  foreach my $h (@{$ds->{'index'}}) {                  foreach my $h (@{$ds->{'index'}}) {
197                          $thes->{$ds->{'tag'}}->insert(                          $thes->{$ds->{'tag'}}->insert(
198                                  headline => $h,                                  sort_by => unac_2($h),
199                                  mfn => $webpac->mfn,                                  mfn => $webpac->mfn,
200                                    headline => $h,
201                          );                          );
202                  }                  }
203          }          }
# Line 179  if (0 && $log->is_debug) { Line 235  if (0 && $log->is_debug) {
235    
236  } # XXX if(0)  } # XXX if(0)
237    
238  $log->debug("lookup hash: ",Dump($webpac->{'lookup'}));  #$log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
239    
240  $log->info("creating tree");  $log->info("creating tree");
241    
# Line 288  my @tree = ({ Line 344  my @tree = ({
344  });  });
345    
346  my $tree = new WebPAC::Tree(  my $tree = new WebPAC::Tree(
347          dir => './out',          tree => \@tree,
348          html => 'browse.html',  );
349    
350    $tree->output(
351            dir => './out',
352            html => 'browse.html',
353          template => './output_template/tree.tt',          template => './output_template/tree.tt',
354          js => 'tree-ids.js',          js => 'tree-ids.js',
         tree => \@tree,  
355  );  );
356    
357    $tree->output(
358            dir => './eurovoc',
359            html => 'hijerarhija.html',
360            template => './output_template/hijerarhija.tt',
361            js => 'tree-ids.js',
362    );
363    
364    
365  $log->info("closing index");  $log->info("closing index");
366  $index->close;  $index->close;
367    
Legend:
Removed from v.490  
changed lines
  Added in v.550
  ViewVC Help
Powered by ViewVC 1.1.26