/[webpac2]/Webpacus/lib/Webpacus/Model/WebPAC.pm
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 /Webpacus/lib/Webpacus/Model/WebPAC.pm

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

revision 96 by dpavlin, Tue Nov 22 12:57:30 2005 UTC revision 178 by dpavlin, Sun Nov 27 05:07:01 2005 UTC
# Line 5  use warnings; Line 5  use warnings;
5  use lib '/data/webpac2/lib';  use lib '/data/webpac2/lib';
6  use base qw/  use base qw/
7          Catalyst::Model          Catalyst::Model
         WebPAC::Search::Estraier  
8  /;  /;
9  use Data::Dumper;  use Data::Dumper;
10  use WebPAC::DB;  use WebPAC::DB;
11  use WebPAC::Output::TT;  use WebPAC::Output::TT;
12    use WebPAC::Search::Estraier 0.02;
13    use File::Slurp;
14    use Time::HiRes;
15    
16  =head1 NAME  =head1 NAME
17    
# Line 35  Configuration for hyperestraier in C<con Line 37  Configuration for hyperestraier in C<con
37    url: 'http://localhost:1978/node/webpac2'    url: 'http://localhost:1978/node/webpac2'
38    user: 'admin'    user: 'admin'
39    passwd: 'admin'    passwd: 'admin'
40      hits_on_page: 100
41    
42   webpac:   webpac:
43    db_path: '/data/webpac2/db'    db_path: '/data/webpac2/db'
# Line 59  sub new { Line 62  sub new {
62          my $est_cfg = $c->config->{hyperestraier};          my $est_cfg = $c->config->{hyperestraier};
63          $est_cfg->{'log'} = $log;          $est_cfg->{'log'} = $log;
64    
65            $est_cfg->{encoding} = $est_cfg->{catalyst_encoding};
66    
67          $log->debug("using config:" . Dumper($est_cfg) );          $log->debug("using config:" . Dumper($est_cfg) );
68    
69          $self->{est} = new WebPAC::Search::Estraier( %{ $est_cfg } );          $self->{est} = new WebPAC::Search::Estraier( %{ $est_cfg } );
70    
71          my $db_path = $c->config->{webpac}->{db_path};          # save config parametars in object
72          my $template_path = $c->config->{webpac}->{template_path};          foreach my $f (qw/db_path template_path hits_on_page/) {
73                    $self->{$f} = $c->config->{hyperestraier}->{$f} ||
74                            $c->config->{webpac}->{$f};
75                    $log->debug("self->{$f} = " . $self->{$f});
76            }
77            my $db_path = $self->{db_path};
78            my $template_path = $self->{template_path};
79    
80          $log->debug("using db path '$db_path', template path '$template_path'");          $log->debug("using db path '$db_path', template path '$template_path'");
81    
# Line 78  sub new { Line 89  sub new {
89                  filters => { foo => sub { shift } },                  filters => { foo => sub { shift } },
90          );          );
91    
92            # default template from config.yaml
93          $self->{template} ||= $c->config->{webpac}->{template};          $self->{template} ||= $c->config->{webpac}->{template};
94    
95          $self->{iconv} = new Text::Iconv(          $self->{iconv} = new Text::Iconv(
# Line 85  sub new { Line 97  sub new {
97                  $c->config->{webpac}->{out_encoding}                  $c->config->{webpac}->{out_encoding}
98          );          );
99    
100            $log->debug("converting encoding from webpac_encoding '" .
101                    $c->config->{webpac}->{webpac_encoding} .
102                    "' to '" .
103                    $c->config->{webpac}->{out_encoding} .
104                    "'"
105            );
106    
107    
108          return $self;          return $self;
109    
110  }  }
111    
112    =head2 iconv_on_save
113    
114      my $out = $m->iconv_on_save( $content );
115    
116    Convert data saved to disk in Webpac encoding.
117    
118    =cut
119    
120    sub iconv_on_save {
121            my $self = shift;
122            my $content = shift || return;
123    
124            $self->{iconv_save} ||= new Text::Iconv(
125                    $self->config->{webpac}->{out_encoding},
126                    $self->config->{webpac}->{webpac_encoding},
127            );
128    
129            return $self->{iconv_save}->convert( $content );
130    }
131    
132    
133    =head2 search
134    
135      my $m->search(
136            phrase => 'query phrase',
137            add_attr => \@add_attr
138            get_attr => [ '@uri' ],
139            max => 42,
140            template => 'result_template.tt',
141      );
142    
143    All fields are standard C<WebPAC::Search::Estraier> parametars except
144    C<template> which will (if specified) return results in HTML using
145    selected template.
146    
147    =cut
148    
149  sub search {  sub search {
150          my ( $self, $query ) = @_;          my $self = shift;
151    
152            my $args = {@_};
153    
154          my $log = $self->{log};          my $log = $self->{log};
155    
156          $log->debug("search got query: $query<--");          $log->debug("args: " . Dumper( $args ));
157    
158          my $template_filename = $self->{template};          my $query = $args->{phrase} || $log->warn("no query phrase") && return;
159    
160          my @results = $self->{est}->search(          $log->debug("search model query: '$query'");
161                  query => $query,          if ($args->{add_attr}) {
162                  attr => [ '@uri' ],                  $log->debug(" + add_attr: " .
163                  max => 100,                          join("','", @{ $args->{add_attr} })
164          );                  );
165            }
166    
167            my $template_filename = $args->{template} || $self->{template};
168    
169            $args->{max} ||= $self->{'hits_on_page'};
170            if (! $args->{max}) {
171                    $args->{max} = 10;
172                    $log->warn("max not set when calling model. Using default of 10");
173            }
174    
175            my $times;      # store some times for benchmarking
176    
177            my $t = time();
178    
179            my @results = $self->{est}->search( %{ $args } );
180    
181            $times->{est} += time() - $t;
182    
183            my $hits = $#results + 1;
184    
185            $log->debug( sprintf("search took %.2fs and returned $hits hits.", $times->{est}) );
186    
187            # just return results?
188            return @results unless ($args->{'template'});
189    
190            #
191            # construct HTML results
192            #
193    
194            my @html_results;
195    
196          for my $i ( 0 .. $#results ) {          for my $i ( 0 .. $#results ) {
197    
198                  my $mfn = $1 if ( $results[$i]->{'@uri'} =~ m#/(\d+)$#);                  my $mfn = $1 if ( $results[$i]->{'@uri'} =~ m#/(\d+)$#);
199    
200                  $log->debug("load_ds( $mfn )");                  #$log->debug("load_ds( $mfn )");
201    
202                    $t = time();
203    
204                    my $ds = $self->{db}->load_ds( $mfn ) || $log->error("can't load_ds( $mfn )") && next;
205    
206                    $times->{db} += time() - $t;
207    
208                    #$log->debug( "ds = " . Dumper( \@html_results ) );
209    
210                  my $ds = $self->{db}->load_ds( $mfn ) || next;                  $t = time();
211                  $results[$i]->{ html } = $self->{iconv}->convert(  
212                  $self->{out}->apply(                  my $html = $self->{out}->apply(
213                          template => $template_filename,                          template => $template_filename,
214                          data => $ds,                          data => $ds,
215                  ) );                  );
216    
217                    $times->{out} += time() - $t;
218    
219                    $t = time();
220    
221                    $html = $self->{iconv}->convert( $html ) || $log->error("can't convert: $html");
222    
223                    $times->{iconv} += time() - $t;
224    
225                    push @html_results, $html;
226    
227          }          }
228    
229          return \@results;          #$log->debug( '@html_results = ' . Dumper( \@html_results ) );
230    
231            $log->debug( sprintf(
232                    "time spent: db = %.2f, out = %.2f, iconv = %.2f",
233                    $times->{db}, $times->{out}, $times->{iconv},
234            ) );
235    
236            return \@html_results;
237    }
238    
239    =head2 record
240    
241      my $html = $m->record(
242            mfn => 42,
243            template => 'foo.tt',
244      );
245    
246    This will load one record, convert it to html using C<template> and return
247    it.
248    
249    =cut
250    
251    sub record {
252            my $self = shift;
253    
254            my $args = {@_};
255            my $log = $self->{log};
256            $log->debug("args: " . Dumper( $args ));
257    
258            foreach my $f (qw/mfn template/) {
259                    $log->die("need $f") unless ($args->{$f});
260            }
261    
262            my $mfn = $args->{mfn};
263    
264            my $ds = $self->{db}->load_ds( $mfn ) || $log->error("can't load_ds( $mfn )") && next;
265    
266            my $html = $self->{out}->apply(
267                    template => $args->{template},
268                    data => $ds,
269            );
270    
271            $html = $self->{iconv}->convert( $html ) || $log->error("can't convert: $html");
272    
273            return $html;
274  }  }
275    
276    =head2 save_html
277    
278              $m->save_html( '/full/path/to/file', $content );
279    
280    It will use C<iconv_on_save> to convert content encoding back to
281    Webpac codepage, recode JavaScript Unicode entities (%u1234),
282    strip extra newlines at beginning and end, and save to
283    C</full/path/to/file.new> and if that succeeds, just rename
284    it over original file which should be atomic on filesystem level.
285    
286    =cut
287    
288    sub save_html {
289            my ($self, $path, $content) = @_;
290    
291            sub _conv_js {
292                    my $t = shift || return;
293                    return $self->{iconv}->convert(chr(hex($t)));
294            }
295            $content =~ s/%u([a-fA-F0-9]{4})/_conv_js($1)/gex;
296            $content =~ s/^[\n\r]+//s;
297            $content =~ s/[\n\r]+$/\n/s;
298    
299            my $iconv_on_save = new Text::Iconv(
300                    $self->config->{webpac}->{out_encoding},
301                    $self->config->{webpac}->{webpac_encoding},
302            );
303            $self->{log}->debug( "content BEFORE : $content" );
304            no utf8;
305            $content = $iconv_on_save->convert( $content ) || die "no content?";
306    
307            $self->{log}->debug( "content AFTER: $content" );
308    
309            write_file($path . '.new', $content) || die "can't save ${path}.new $!";
310            rename $path . '.new', $path || die "can't rename to $path: $!";
311    }
312    
313    =head2 load_html
314    
315      my $html = $m->load_html('/full/path/to/file');
316    
317    This will convert file from Webpac encoding to Catalyst and
318    convert that data to escaped HTML (for sending into
319    C<< <textarea/> >> tags in html.
320    
321    =cut
322    
323    sub load_html {
324            my ($self, $path) = @_;
325    
326            die "no path?" unless ($path);
327    
328            my $content = read_file($path) || die "can't read $path: $!";
329            #$content = $q->escapeHTML($iconv_utf8->convert($content));
330            $content = $self->{iconv}->convert($content);
331    
332            return $content;
333    }
334    
335  =head1 AUTHOR  =head1 AUTHOR
336    
Legend:
Removed from v.96  
changed lines
  Added in v.178
  ViewVC Help
Powered by ViewVC 1.1.26