/[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 466 by dpavlin, Thu Sep 23 23:17:42 2004 UTC revision 549 by dpavlin, Tue Oct 26 20:53:03 2004 UTC
# Line 7  all2all.pl - basic script for all WebPAC Line 7  all2all.pl - basic script for all WebPAC
7  =cut  =cut
8    
9  use strict;  use strict;
10    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 31  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            return $string;
57    }
58    sub unac_2 {
59            my $string = shift || return;
60            if (length($string) > 2) {
61                    my $pr = substr($string,0,2);
62                    $string = unac_string($code_page,substr($string,2));
63                    $string =~ tr/đĐ/dD/;
64                    $string = $pr . $string;
65            }
66            $string =~ s/&(\w)(acute|cedil|circ|grave|ring|slash|tilde|uml);/$1/gi;
67            $string =~ s/ð/d/g;
68            $string =~ s/&E[tT][hH];/D/g;
69            $string =~ s/&(\w\w)lig;/$1/gi;
70            return lc($string);
71    }
72    
73  # create WebPAC object  # create WebPAC object
74  #  #
75  my $webpac = new WebPAC(  my $webpac = new WebPAC(
# Line 39  my $webpac = new WebPAC( Line 78  my $webpac = new WebPAC(
78          start_mfn => $start_mfn,          start_mfn => $start_mfn,
79          debug => $debug,          debug => $debug,
80          low_mem => $low_mem,          low_mem => $low_mem,
81            filter => $filter,
82  ) || die;  ) || die;
83    
84  my $log = $webpac->_get_logger() || die "can't get logger";  my $log = $webpac->_get_logger() || die "can't get logger";
# Line 57  $|=1; Line 97  $|=1;
97  my $maxmfn = $webpac->open_isis(  my $maxmfn = $webpac->open_isis(
98          filename => shift @ARGV || '/data/hidra/THS/THS',          filename => shift @ARGV || '/data/hidra/THS/THS',
99          lookup => [          lookup => [
100          { 'key' => 'd:v900', 'val' => 'v250^a' },          { 'key' => 'd:v900', 'val' => 'filter{CROVOC_tree}v250^a v800' },
101  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
102  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
103  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
104          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
105          { 'key' => '900_mfn:v900', 'val' => 'v000' },          { 'key' => '900_mfn:v900', 'val' => 'v000' },
106          # tree structure          # tree structure
107          { 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' },          { 'eval' => 'length("v251") == 2 && "v800" =~ m/EUROVOC/ || "v800" =~ m/CROVOC/ && "v251" =~ m/^(H|HD|L|Z|P)$/', 'key' => 'root:v251', 'val' => 'v900' },
108          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },
109          { 'eval' => '"v561^4" && "v562^4"', 'key' => 'code:v900', 'val' => 'v561^4:v562^4' },          { 'eval' => '"v561^4" && "v562^4"', 'key' => 'code:v900', 'val' => 'v561^4:v562^4' },
110          ],          ],
# Line 122  while (my $rec = $webpac->fetch_rec) { Line 162  while (my $rec = $webpac->fetch_rec) {
162                  $words =~ s/\W*\s+\W*/ /g;                  $words =~ s/\W*\s+\W*/ /g;
163                  $words =~ s/\W+$//;                  $words =~ s/\W+$//;
164    
165                    # first try to generate headline for this entry from index
166                    my $h = $ds->{'index'}->[0];
167                    # then, from display
168                    $h ||= $ds->{'display'}->[0];
169                    # and as last resport, fallback to headline
170                    $h ||= $headline;
171    
172                  $index->insert(                  $index->insert(
173                          index_name => $ds->{'tag'},                          index_name => $ds->{'tag'},
174                          #path => $f,                          #path => $f,
175                          path => $webpac->mfn,                          path => $webpac->mfn,
176                          headline => $headline,                          headline => $h,
177                          words => $words,                          words => unac($words),
178                  );                  );
179          }          }
180    
# Line 135  while (my $rec = $webpac->fetch_rec) { Line 182  while (my $rec = $webpac->fetch_rec) {
182          foreach my $ds (@ds) {          foreach my $ds (@ds) {
183                  next if (! $ds->{'index'});                  next if (! $ds->{'index'});
184    
185                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index;                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index( name => $ds->{'tag'} );
186    
187                  foreach my $h (@{$ds->{'index'}}) {                  foreach my $h (@{$ds->{'index'}}) {
188                          $thes->{$ds->{'tag'}}->insert(                          $thes->{$ds->{'tag'}}->insert(
189                                  headline => $h,                                  sort_by => unac_2($h),
190                                  mfn => $webpac->mfn,                                  mfn => $webpac->mfn,
191                                    headline => $h,
192                          );                          );
193                  }                  }
194          }          }
# Line 178  if (0 && $log->is_debug) { Line 226  if (0 && $log->is_debug) {
226    
227  } # XXX if(0)  } # XXX if(0)
228    
229  $log->debug("lookup hash: ",Dump($webpac->{'lookup'}));  #$log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
230    
231  $log->info("creating tree");  $log->info("creating tree");
232    
# Line 200  my @tree = ({ Line 248  my @tree = ({
248          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },
249          have_children   => sub { return $l->{"a:".$_[0]."::"} },          have_children   => sub { return $l->{"a:".$_[0]."::"} },
250          have_children_at_level => sub {          have_children_at_level => sub {
251                                  my $l2 = shift;                                  return unless (defined($l->{"code:".$_[1]}));
252                                  my $l1 = substr($l2,0,1);                                  my $code = shift @{$l->{"code:".$_[1]}};
253                                  return(9, "a:$l1:$l2:" ) if ($l->{"a:$l1:$l2"});                                  print STDERR "## $_[1] -> $code\n";
254                                    return undef unless($code);
255                                    return(9, $l->{"a:$code:"} ) if (defined($l->{"a:$code:"}));
256                          },                          },
257          style           => 'display: none',          style           => 'display: none',
258          },{          },{
# Line 274  my @tree = ({ Line 324  my @tree = ({
324          have_children   => sub { 0 },          have_children   => sub { 0 },
325          },{          },{
326          # 9 - level which is never reached except explicitly          # 9 - level which is never reached except explicitly
327          code_arr        => sub { @{$l->{$_[0]}} },          code_arr        => sub { @{$_[0]} },
328          filter_code     => sub { shift },          filter_code     => sub { shift },
329          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },
330          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },
331          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },
332          have_children   => sub { 0 },          have_children   => sub { 0 },
333          have_children_at_level => sub { defined($l->{"a:".$_[1].":".$_[0]}) && return (9,"a:".$_[1].":".$_[0]) },          have_children_at_level => sub { defined($l->{"a:".$_[1].":".$_[0]}) && return (9,$l->{"a:".$_[1].":".$_[0]}) },
334          },{          },{
335  });  });
336    
337  my $tree = new WebPAC::Tree(  my $tree = new WebPAC::Tree(
338          dir => './out',          tree => \@tree,
339          html => 'browse.html',  );
340    
341    $tree->output(
342            dir => './out',
343            html => 'browse.html',
344          template => './output_template/tree.tt',          template => './output_template/tree.tt',
345          js => 'tree-ids.js',          js => 'tree-ids.js',
         tree => \@tree,  
346  );  );
347    
348    $tree->output(
349            dir => './eurovoc',
350            html => 'hijerarhija.html',
351            template => './output_template/hijerarhija.tt',
352            js => 'tree-ids.js',
353    );
354    
355    
356  $log->info("closing index");  $log->info("closing index");
357  $index->close;  $index->close;
358    
Legend:
Removed from v.466  
changed lines
  Added in v.549
  ViewVC Help
Powered by ViewVC 1.1.26