/[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 500 by dpavlin, Sun Oct 10 11:04:52 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;
# Line 31  my $result = GetOptions( Line 32  my $result = GetOptions(
32          "low_mem!"      => \$low_mem,          "low_mem!"      => \$low_mem,
33  );  );
34    
35    my $filter = {
36            'CROVOC' => sub {
37                    my $tmp = shift || return;
38                    return undef unless ($tmp =~ s/CROVOC.*$/ */);
39                    return $tmp;
40            },
41    };
42    
43  # create WebPAC object  # create WebPAC object
44  #  #
45  my $webpac = new WebPAC(  my $webpac = new WebPAC(
# Line 39  my $webpac = new WebPAC( Line 48  my $webpac = new WebPAC(
48          start_mfn => $start_mfn,          start_mfn => $start_mfn,
49          debug => $debug,          debug => $debug,
50          low_mem => $low_mem,          low_mem => $low_mem,
51            filter => $filter,
52  ) || die;  ) || die;
53    
54  my $log = $webpac->_get_logger() || die "can't get logger";  my $log = $webpac->_get_logger() || die "can't get logger";
# Line 64  my $maxmfn = $webpac->open_isis( Line 74  my $maxmfn = $webpac->open_isis(
74          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },          { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
75          { 'key' => '900_mfn:v900', 'val' => 'v000' },          { 'key' => '900_mfn:v900', 'val' => 'v000' },
76          # tree structure          # tree structure
77          { 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' },          { 'eval' => 'length("v251") == 2 && "v800" =~ m/EUROVOC/ || "v800" =~ m/CROVOC/ && "v251" =~ m/^(H|HD|L|Z|P)$/', 'key' => 'root:v251', 'val' => 'v900' },
78          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },          { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },
79          { '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' },
80          ],          ],
# Line 115  while (my $rec = $webpac->fetch_rec) { Line 125  while (my $rec = $webpac->fetch_rec) {
125          foreach my $ds (@ds) {          foreach my $ds (@ds) {
126                  next if (! $ds->{'swish'});                  next if (! $ds->{'swish'});
127    
128                    # strip all non word characters from beginning or end
129                    # of word
130                    my $words = join(" ",@{$ds->{'swish'}});
131                    $words =~ s/^\W+//;
132                    $words =~ s/\W*\s+\W*/ /g;
133                    $words =~ s/\W+$//;
134    
135                  $index->insert(                  $index->insert(
136                          index_name => $ds->{'tag'},                          index_name => $ds->{'tag'},
137                          path => $f,                          #path => $f,
138                            path => $webpac->mfn,
139                          headline => $headline,                          headline => $headline,
140                          words => join(" ",@{$ds->{'swish'}})                          words => $words,
141                  );                  );
142          }          }
143    
# Line 192  my @tree = ({ Line 210  my @tree = ({
210          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[1]}} },
211          have_children   => sub { return $l->{"a:".$_[0]."::"} },          have_children   => sub { return $l->{"a:".$_[0]."::"} },
212          have_children_at_level => sub {          have_children_at_level => sub {
213                                  my $l2 = shift;                                  return unless (defined($l->{"code:".$_[1]}));
214                                  my $l1 = substr($l2,0,1);                                  my $code = shift @{$l->{"code:".$_[1]}};
215                                  return(9, "a:$l1:$l2:" ) if ($l->{"a:$l1:$l2"});                                  print STDERR "## $_[1] -> $code\n";
216                                    return undef unless($code);
217                                    return(9, $l->{"a:$code:"} ) if (defined($l->{"a:$code:"}));
218                          },                          },
219          style           => 'display: none',          style           => 'display: none',
220          },{          },{
# Line 266  my @tree = ({ Line 286  my @tree = ({
286          have_children   => sub { 0 },          have_children   => sub { 0 },
287          },{          },{
288          # 9 - level which is never reached except explicitly          # 9 - level which is never reached except explicitly
289          code_arr        => sub { @{$l->{$_[0]}} },          code_arr        => sub { @{$_[0]} },
290          filter_code     => sub { shift },          filter_code     => sub { shift },
291          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },          lookup_v900     => sub { shift @{$l->{"code:".$_[0]}} },
292          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },          lookup_term     => sub { shift @{$l->{"d:".$_[0]}} },
293          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },          lookup_mfn      => sub { shift @{$l->{"900_mfn:".$_[0]}} },
294          have_children   => sub { 0 },          have_children   => sub { 0 },
295          have_children_at_level => sub { defined($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]}) },
296          },{          },{
297  });  });
298    
Legend:
Removed from v.460  
changed lines
  Added in v.500
  ViewVC Help
Powered by ViewVC 1.1.26