/[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 460 by dpavlin, Tue Sep 21 20:43:43 2004 UTC revision 546 by dpavlin, Tue Oct 26 18:59:38 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 . $string;
65            }
66            return lc($string);
67    }
68    
69  # create WebPAC object  # create WebPAC object
70  #  #
71  my $webpac = new WebPAC(  my $webpac = new WebPAC(
# Line 39  my $webpac = new WebPAC( Line 74  my $webpac = new WebPAC(
74          start_mfn => $start_mfn,          start_mfn => $start_mfn,
75          debug => $debug,          debug => $debug,
76          low_mem => $low_mem,          low_mem => $low_mem,
77            filter => $filter,
78  ) || die;  ) || die;
79    
80  my $log = $webpac->_get_logger() || die "can't get logger";  my $log = $webpac->_get_logger() || die "can't get logger";
# Line 57  $|=1; Line 93  $|=1;
93  my $maxmfn = $webpac->open_isis(  my $maxmfn = $webpac->open_isis(
94          filename => shift @ARGV || '/data/hidra/THS/THS',          filename => shift @ARGV || '/data/hidra/THS/THS',
95          lookup => [          lookup => [
96          { 'key' => 'd:v900', 'val' => 'v250^a' },          { 'key' => 'd:v900', 'val' => 'filter{CROVOC_tree}v250^a v800' },
97  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
98  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
99  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },  #       { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
100          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
101          { 'key' => '900_mfn:v900', 'val' => 'v000' },          { 'key' => '900_mfn:v900', 'val' => 'v000' },
102          # tree structure          # tree structure
103          { 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' },          { 'eval' => 'length("v251") == 2 && "v800" =~ m/EUROVOC/ || "v800" =~ m/CROVOC/ && "v251" =~ m/^(H|HD|L|Z|P)$/', 'key' => 'root:v251', 'val' => 'v900' },
104          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },
105          { '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' },
106          ],          ],
# Line 115  while (my $rec = $webpac->fetch_rec) { Line 151  while (my $rec = $webpac->fetch_rec) {
151          foreach my $ds (@ds) {          foreach my $ds (@ds) {
152                  next if (! $ds->{'swish'});                  next if (! $ds->{'swish'});
153    
154                    # strip all non word characters from beginning or end
155                    # of word
156                    my $words = join(" ",@{$ds->{'swish'}});
157                    $words =~ s/^\W+//;
158                    $words =~ s/\W*\s+\W*/ /g;
159                    $words =~ s/\W+$//;
160    
161                    # first try to generate headline for this entry from index
162                    my $h = $ds->{'index'}->[0];
163                    # then, from display
164                    $h ||= $ds->{'display'}->[0];
165                    # and as last resport, fallback to headline
166                    $h ||= $headline;
167    
168                  $index->insert(                  $index->insert(
169                          index_name => $ds->{'tag'},                          index_name => $ds->{'tag'},
170                          path => $f,                          #path => $f,
171                          headline => $headline,                          path => $webpac->mfn,
172                          words => join(" ",@{$ds->{'swish'}})                          headline => $h,
173                            words => unac($words),
174                  );                  );
175          }          }
176    
# Line 127  while (my $rec = $webpac->fetch_rec) { Line 178  while (my $rec = $webpac->fetch_rec) {
178          foreach my $ds (@ds) {          foreach my $ds (@ds) {
179                  next if (! $ds->{'index'});                  next if (! $ds->{'index'});
180    
181                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index;                  $thes->{$ds->{'tag'}} ||= new WebPAC::Index( name => $ds->{'tag'} );
182    
183                  foreach my $h (@{$ds->{'index'}}) {                  foreach my $h (@{$ds->{'index'}}) {
184                          $thes->{$ds->{'tag'}}->insert(                          $thes->{$ds->{'tag'}}->insert(
185                                  headline => $h,                                  sort_by => unac_2($h),
186                                  mfn => $webpac->mfn,                                  mfn => $webpac->mfn,
187                                    headline => $h,
188                          );                          );
189                  }                  }
190          }          }
# Line 170  if (0 && $log->is_debug) { Line 222  if (0 && $log->is_debug) {
222    
223  } # XXX if(0)  } # XXX if(0)
224    
225  $log->debug("lookup hash: ",Dump($webpac->{'lookup'}));  #$log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
226    
227  $log->info("creating tree");  $log->info("creating tree");
228    
# Line 192  my @tree = ({ Line 244  my @tree = ({
244          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },
245          have_children   => sub { return $l->{"a:".$_[0]."::"} },          have_children   => sub { return $l->{"a:".$_[0]."::"} },
246          have_children_at_level => sub {          have_children_at_level => sub {
247                                  my $l2 = shift;                                  return unless (defined($l->{"code:".$_[1]}));
248                                  my $l1 = substr($l2,0,1);                                  my $code = shift @{$l->{"code:".$_[1]}};
249                                  return(9, "a:$l1:$l2:" ) if ($l->{"a:$l1:$l2"});                                  print STDERR "## $_[1] -> $code\n";
250                                    return undef unless($code);
251                                    return(9, $l->{"a:$code:"} ) if (defined($l->{"a:$code:"}));
252                          },                          },
253          style           => 'display: none',          style           => 'display: none',
254          },{          },{
# Line 266  my @tree = ({ Line 320  my @tree = ({
320          have_children   => sub { 0 },          have_children   => sub { 0 },
321          },{          },{
322          # 9 - level which is never reached except explicitly          # 9 - level which is never reached except explicitly
323          code_arr        => sub { @{$l->{$_[0]}} },          code_arr        => sub { @{$_[0]} },
324          filter_code     => sub { shift },          filter_code     => sub { shift },
325          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },
326          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },
327          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },
328          have_children   => sub { 0 },          have_children   => sub { 0 },
329          have_children_at_level => sub { defined($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]}) },
330          },{          },{
331  });  });
332    
333  my $tree = new WebPAC::Tree(  my $tree = new WebPAC::Tree(
334          dir => './out',          tree => \@tree,
335          html => 'browse.html',  );
336    
337    $tree->output(
338            dir => './out',
339            html => 'browse.html',
340          template => './output_template/tree.tt',          template => './output_template/tree.tt',
341          js => 'tree-ids.js',          js => 'tree-ids.js',
         tree => \@tree,  
342  );  );
343    
344    $tree->output(
345            dir => './eurovoc',
346            html => 'hijerarhija.html',
347            template => './output_template/hijerarhija.tt',
348            js => 'tree-ids.js',
349    );
350    
351    
352  $log->info("closing index");  $log->info("closing index");
353  $index->close;  $index->close;
354    
Legend:
Removed from v.460  
changed lines
  Added in v.546
  ViewVC Help
Powered by ViewVC 1.1.26