| 6 |
use base qw/ |
use base qw/ |
| 7 |
Catalyst::Model |
Catalyst::Model |
| 8 |
/; |
/; |
| 9 |
use Data::Dumper; |
use WebPAC::Store 0.08; |
| 10 |
use WebPAC::DB; |
use Search::Estraier 0.04; |
|
use WebPAC::Output::TT; |
|
|
use WebPAC::Search::Estraier 0.02; |
|
| 11 |
use File::Slurp; |
use File::Slurp; |
| 12 |
use Time::HiRes; |
use Time::HiRes qw/time/; |
| 13 |
|
use Encode qw/encode decode from_to/; |
| 14 |
|
use Template; |
| 15 |
|
|
| 16 |
=head1 NAME |
=head1 NAME |
| 17 |
|
|
| 34 |
|
|
| 35 |
# configuration for hyper estraier full text search engine |
# configuration for hyper estraier full text search engine |
| 36 |
hyperestraier: |
hyperestraier: |
| 37 |
url: 'http://localhost:1978/node/webpac2' |
masterurl: 'http://localhost:1978/node/webpac2' |
| 38 |
|
defaultnode: 'webpac2' |
| 39 |
|
defaultdepth: 1 |
| 40 |
user: 'admin' |
user: 'admin' |
| 41 |
passwd: 'admin' |
passwd: 'admin' |
| 42 |
hits_on_page: 100 |
hits_on_page: 100 |
| 43 |
|
hits_for_pager: 1000 |
| 44 |
|
|
| 45 |
webpac: |
webpac: |
| 46 |
db_path: '/data/webpac2/db' |
db_path: '/data/webpac2/db' |
| 48 |
template: 'html_ffzg_results_short.tt' |
template: 'html_ffzg_results_short.tt' |
| 49 |
# encoding comming from webpac |
# encoding comming from webpac |
| 50 |
webpac_encoding: 'iso-8859-2' |
webpac_encoding: 'iso-8859-2' |
|
# encoding expected by Catalyst |
|
|
out_encoding: 'UTF-8' |
|
| 51 |
|
|
| 52 |
=cut |
=cut |
| 53 |
|
|
| 63 |
my $est_cfg = $c->config->{hyperestraier}; |
my $est_cfg = $c->config->{hyperestraier}; |
| 64 |
$est_cfg->{'log'} = $log; |
$est_cfg->{'log'} = $log; |
| 65 |
|
|
| 66 |
$est_cfg->{encoding} = $est_cfg->{catalyst_encoding}; |
$est_cfg->{encoding} = $est_cfg->{catalyst_encoding} || $c->config->{catalyst_encoding} or $c->log->fatal("can't find catalyst_encoding"); |
| 67 |
|
|
| 68 |
$log->debug("using config:" . Dumper($est_cfg) ); |
$log->dumper($est_cfg, 'est_cfg'); |
| 69 |
|
|
| 70 |
$self->{est} = new WebPAC::Search::Estraier( %{ $est_cfg } ); |
if (! $est_cfg->{database}) { |
| 71 |
|
my $defaultnode = $est_cfg->{defaultnode} || $log->logdie("can't find defaultnode in estraier configuration"); |
| 72 |
|
$log->info("using default node $defaultnode"); |
| 73 |
|
$est_cfg->{database} = $defaultnode; |
| 74 |
|
} |
| 75 |
|
|
| 76 |
|
my $url = $est_cfg->{masterurl} . '/node/' . $est_cfg->{database}; |
| 77 |
|
|
| 78 |
|
$log->info("opening Hyper Estraier index $url as $est_cfg->{'user'}"); |
| 79 |
|
|
| 80 |
my $db_path = $c->config->{webpac}->{db_path}; |
$self->{est_node} = Search::Estraier::Node->new( |
| 81 |
my $template_path = $c->config->{webpac}->{template_path}; |
url => $url, |
| 82 |
$self->{template_path} = $template_path; |
user => $est_cfg->{user}, |
| 83 |
|
passwd => $est_cfg->{passwd}, |
| 84 |
|
); |
| 85 |
|
|
| 86 |
|
$log->fatal("can't create Search::Estraier::Node $url") unless ($self->{est_node}); |
| 87 |
|
|
| 88 |
|
# save config parametars in object |
| 89 |
|
foreach my $f (qw/ |
| 90 |
|
db_path template_path hits_on_page webpac_encoding defaultdepth |
| 91 |
|
masterurl defaultnode |
| 92 |
|
/) { |
| 93 |
|
$self->{$f} = $c->config->{hyperestraier}->{$f} || |
| 94 |
|
$c->config->{webpac}->{$f}; |
| 95 |
|
$log->debug("self->{$f} = " . $self->{$f}); |
| 96 |
|
} |
| 97 |
|
my $db_path = $self->{db_path}; |
| 98 |
|
my $template_path = $self->{template_path}; |
| 99 |
|
|
| 100 |
$log->debug("using db path '$db_path', template path '$template_path'"); |
$log->debug("using db path '$db_path', template path '$template_path'"); |
| 101 |
|
|
| 102 |
$self->{db} = new WebPAC::DB( |
$self->{db} = new WebPAC::Store( |
| 103 |
path => $db_path, |
path => $db_path, |
| 104 |
read_only => 1, |
read_only => 1, |
| 105 |
); |
database => $est_cfg->{database}, |
|
|
|
|
$self->{out} = new WebPAC::Output::TT( |
|
|
include_path => $template_path, |
|
|
filters => { foo => sub { shift } }, |
|
| 106 |
); |
); |
| 107 |
|
|
| 108 |
# default template from config.yaml |
# default template from config.yaml |
| 109 |
$self->{template} ||= $c->config->{webpac}->{template}; |
$self->{template} ||= $c->config->{webpac}->{template}; |
| 110 |
|
|
|
$self->{iconv} = new Text::Iconv( |
|
|
$c->config->{webpac}->{webpac_encoding}, |
|
|
$c->config->{webpac}->{out_encoding} |
|
|
); |
|
|
|
|
| 111 |
$log->debug("converting encoding from webpac_encoding '" . |
$log->debug("converting encoding from webpac_encoding '" . |
| 112 |
$c->config->{webpac}->{webpac_encoding} . |
$c->config->{webpac}->{webpac_encoding} . |
|
"' to '" . |
|
|
$c->config->{webpac}->{out_encoding} . |
|
| 113 |
"'" |
"'" |
| 114 |
); |
); |
| 115 |
|
|
| 116 |
# save config parametars in object |
$self->{databases} = $c->config->{databases} || $log->fatal("can't find databases in config"); |
| 117 |
foreach my $f (qw/hits_on_page/) { |
|
| 118 |
$self->{$_} = $c->config->{$_}; |
# create Template toolkit instance |
| 119 |
$log->debug("self->{$_} = " . $c->config->{$_}); |
$self->{'tt'} = Template->new( |
| 120 |
} |
INCLUDE_PATH => $template_path, |
| 121 |
|
FILTERS => { |
| 122 |
|
dump_html => sub { |
| 123 |
|
return unless (@_); |
| 124 |
|
my $out; |
| 125 |
|
my $i = 1; |
| 126 |
|
foreach my $v (@_) { |
| 127 |
|
$out .= qq{<div id="dump_$i">} . |
| 128 |
|
Data::HTMLDumper->Dump([ $v ],[ "v$i" ]) . |
| 129 |
|
qq{</div>}; |
| 130 |
|
$i++; |
| 131 |
|
} |
| 132 |
|
$out =~ s!<table[^>/]*>!<table class="dump">!gis if ($out); |
| 133 |
|
return $out; |
| 134 |
|
} |
| 135 |
|
}, |
| 136 |
|
EVAL_PERL => 1, |
| 137 |
|
); |
| 138 |
|
|
| 139 |
return $self; |
return $self; |
| 140 |
|
|
| 141 |
} |
} |
| 142 |
|
|
| 143 |
=head2 iconv_on_save |
=head2 setup_site |
| 144 |
|
|
| 145 |
my $out = $m->iconv_on_save( $content ); |
$self->setup_site('site_name'); |
| 146 |
|
|
| 147 |
Convert data saved to disk in Webpac encoding. |
Change node URL and database name according to site name (if available) or fallback |
| 148 |
|
to C<defaultnode> from configuration. |
| 149 |
|
|
| 150 |
=cut |
=cut |
| 151 |
|
|
| 152 |
sub iconv_on_save { |
sub setup_site { |
| 153 |
my $self = shift; |
my $self = shift; |
| 154 |
|
|
| 155 |
$self->{iconv_save} ||= new Text::Iconv( |
my $site = shift || $self->{defaultnode}; |
| 156 |
$self->config->{webpac}->{out_encoding}, |
if (! $site) { |
| 157 |
$self->config->{webpac}->{webpac_encoding}, |
$self->{log}->warn("not changing site from " . $self->{est_node}->{url}); |
| 158 |
); |
return; |
| 159 |
|
} |
| 160 |
|
|
| 161 |
$self->{iconv_save}->convert( @_ ); |
$self->{log}->fatal("setup_site can't find site or defaultnode") unless ($site); |
|
} |
|
| 162 |
|
|
| 163 |
|
my $url = $self->{masterurl} . '/node/' . $site; |
| 164 |
|
$self->{est_node}->set_url( $url ); |
| 165 |
|
$self->{log}->debug("setup_site '$site' using $url"); |
| 166 |
|
} |
| 167 |
|
|
| 168 |
=head2 search |
=head2 search |
| 169 |
|
|
| 173 |
get_attr => [ '@uri' ], |
get_attr => [ '@uri' ], |
| 174 |
max => 42, |
max => 42, |
| 175 |
template => 'result_template.tt', |
template => 'result_template.tt', |
| 176 |
|
depth => 1, |
| 177 |
); |
); |
| 178 |
|
|
| 179 |
All fields are standard C<WebPAC::Search::Estraier> parametars except |
All fields are standard C<WebPAC::Search::Estraier> parametars except |
| 185 |
sub search { |
sub search { |
| 186 |
my $self = shift; |
my $self = shift; |
| 187 |
|
|
| 188 |
|
my $search_start_t = time(); |
| 189 |
|
|
| 190 |
my $args = {@_}; |
my $args = {@_}; |
| 191 |
|
|
| 192 |
my $log = $self->{log}; |
my $log = $self->{log}; |
| 193 |
|
|
| 194 |
$log->debug("args: " . Dumper( $args )); |
$log->dumper($args, 'args'); |
| 195 |
|
|
| 196 |
my $query = $args->{phrase} || $log->warn("no query phrase") && return; |
my $query = $args->{phrase} || $log->warn("no query phrase") && return; |
| 197 |
|
|
|
$log->debug("search model query: '$query'"); |
|
|
if ($args->{add_attr}) { |
|
|
$log->debug(" + add_attr: " . |
|
|
join("','", @{ $args->{add_attr} }) |
|
|
); |
|
|
} |
|
|
|
|
| 198 |
my $template_filename = $args->{template} || $self->{template}; |
my $template_filename = $args->{template} || $self->{template}; |
| 199 |
|
|
| 200 |
$args->{max} ||= $self->{'hits_on_page'}; |
$args->{hits_on_page} ||= $self->{'hits_for_pager'}; |
| 201 |
if (! $args->{max}) { |
if (! $args->{hits_on_page}) { |
| 202 |
$args->{max} = 10; |
$args->{hits_on_page} = 100; |
| 203 |
$log->warn("max not set when calling model. Using default of 10"); |
$log->warn("max not set when calling model. Using default of $args->{hits_on_page}"); |
| 204 |
} |
} |
| 205 |
|
|
| 206 |
my $times; # store some times for benchmarking |
my $times; # store some times for benchmarking |
| 207 |
|
|
| 208 |
my $t = time(); |
my $t = time(); |
| 209 |
|
|
| 210 |
my @results = $self->{est}->search( %{ $args } ); |
# transfer depth of search |
| 211 |
|
if (! $args->{depth}) { |
| 212 |
|
my $default = $self->{defaultdepth} || $log->logdie("can't find defaultdepth in estraier configuration"); |
| 213 |
|
$args->{depth} = $default; |
| 214 |
|
$log->warn("using default search depth $default"); |
| 215 |
|
} |
| 216 |
|
$args->{depth} ||= 0; |
| 217 |
|
|
| 218 |
$times->{est} += time() - $t; |
$log->debug("searching " . $self->{est_node}->{url} . " hits on page: $args->{hits_on_page} depth: $args->{depth} phrase: " . ($query || '[none]') ); |
| 219 |
|
|
| 220 |
|
# |
| 221 |
|
# construct condition for Hyper Estraier |
| 222 |
|
# |
| 223 |
|
my $cond = Search::Estraier::Condition->new(); |
| 224 |
|
if ( ref($args->{add_attr}) eq 'ARRAY' ) { |
| 225 |
|
$log->debug("adding search attributes: " . join(", ", @{ $args->{add_attr} }) ); |
| 226 |
|
map { |
| 227 |
|
$cond->add_attr( $_ ); |
| 228 |
|
$log->debug(" + $_"); |
| 229 |
|
} @{ $args->{add_attr} }; |
| 230 |
|
}; |
| 231 |
|
|
| 232 |
|
$cond->set_phrase( $query ) if ($query); |
| 233 |
|
$cond->set_options( $args->{options} ) if ($args->{options}); |
| 234 |
|
$cond->set_order( $args->{order} ) if ($args->{order}); |
| 235 |
|
|
| 236 |
|
my $hits_on_page = $args->{hits_on_page} || 7; |
| 237 |
|
my $page = $args->{page} || 1; |
| 238 |
|
if ($page < 1) { |
| 239 |
|
$log->warn("page number $page < 1"); |
| 240 |
|
$page = 1; |
| 241 |
|
} |
| 242 |
|
|
| 243 |
|
$cond->set_max( my $max = $page * $hits_on_page ); |
| 244 |
|
$cond->set_skip( my $skip = ( $page - 1 ) * $hits_on_page ); |
| 245 |
|
|
| 246 |
|
$log->debug("search max: $max, skip: $skip"); |
| 247 |
|
|
| 248 |
|
my $result = $self->{est_node}->search($cond, $args->{depth}); |
| 249 |
|
if (! $result) { |
| 250 |
|
$self->{log}->fatal("search didn't return result"); |
| 251 |
|
return; |
| 252 |
|
} |
| 253 |
|
my $hits = $result->doc_num; |
| 254 |
|
|
| 255 |
my $hits = $#results + 1; |
$times->{est} += time() - $t; |
| 256 |
|
|
| 257 |
$log->debug( sprintf("search took %.2fs and returned $hits hits.", $times->{est}) ); |
$log->debug( sprintf("search took %.6fs and returned $hits hits.", $times->{est}) ); |
| 258 |
|
|
| 259 |
# just return results? |
$self->{hints} = $result->{hints}; |
| 260 |
return @results unless ($args->{'template'}); |
#$log->dumper($self->{hints}, 'original hints' ); |
| 261 |
|
|
| 262 |
# |
# |
| 263 |
# construct HTML results |
# fetch results |
| 264 |
# |
# |
| 265 |
|
|
| 266 |
my @html_results; |
my @results; |
|
|
|
|
for my $i ( 0 .. $#results ) { |
|
| 267 |
|
|
| 268 |
my $mfn = $1 if ( $results[$i]->{'@uri'} =~ m#/(\d+)$#); |
|
| 269 |
|
|
| 270 |
#$log->debug("load_ds( $mfn )"); |
for my $i ( 0 .. ( $hits < $max ? ($hits-1) : ($max-1) ) ) { |
| 271 |
|
|
| 272 |
$t = time(); |
$t = time(); |
| 273 |
|
|
| 274 |
my $ds = $self->{db}->load_ds( $mfn ) || $log->error("can't load_ds( $mfn )") && next; |
#$log->debug("get_doc($i)"); |
| 275 |
|
my $doc = $result->get_doc( $i ); |
| 276 |
|
if (! $doc) { |
| 277 |
|
$log->warn("can't find result $i"); |
| 278 |
|
next; |
| 279 |
|
} |
| 280 |
|
|
| 281 |
$times->{db} += time() - $t; |
my $hash; |
| 282 |
|
|
| 283 |
#$log->debug( "ds = " . Dumper( \@html_results ) ); |
foreach my $attr (@{ $args->{get_attr} }) { |
| 284 |
|
my $val = $doc->attr( $attr ); |
| 285 |
|
#$log->debug("attr $attr = ", $val || 'undef'); |
| 286 |
|
$hash->{$attr} = $val if (defined($val)); |
| 287 |
|
} |
| 288 |
|
|
| 289 |
$t = time(); |
$times->{hash} += time() - $t; |
| 290 |
|
|
| 291 |
my $html = $self->{out}->apply( |
next unless ($hash); |
|
template => $template_filename, |
|
|
data => $ds, |
|
|
); |
|
| 292 |
|
|
| 293 |
$times->{out} += time() - $t; |
if (! $args->{'template'}) { |
| 294 |
|
push @results, $hash; |
| 295 |
|
} else { |
| 296 |
|
my ($database, $prefix, $id); |
| 297 |
|
|
| 298 |
$t = time(); |
if ( $hash->{'@uri'} =~ m!/([^/]+)/([^/]+)/(\d+)$!) { |
| 299 |
|
($database, $prefix,$id) = ($1,$2,$3); |
| 300 |
|
} else { |
| 301 |
|
$log->warn("can't decode database/prefix/id from " . $hash->{'@uri'}); |
| 302 |
|
next; |
| 303 |
|
} |
| 304 |
|
|
| 305 |
$html = $self->{iconv}->convert( $html ) || $log->error("can't convert: $html"); |
#$log->debug("load_ds( id => $id, prefix => '$prefix' )"); |
| 306 |
|
|
| 307 |
$times->{iconv} += time() - $t; |
$t = time(); |
| 308 |
|
|
| 309 |
push @html_results, $html; |
my $ds = $self->{db}->load_ds( database => $database, prefix => $prefix, id => $id ); |
| 310 |
|
if (! $ds) { |
| 311 |
|
$log->error("can't load_ds( ${database}/${prefix}/${id} )"); |
| 312 |
|
next; |
| 313 |
|
} |
| 314 |
|
|
| 315 |
} |
$times->{db} += time() - $t; |
| 316 |
|
|
| 317 |
|
$t = time(); |
| 318 |
|
|
| 319 |
|
my $html = $self->apply( |
| 320 |
|
template => $template_filename, |
| 321 |
|
data => $ds, |
| 322 |
|
record_uri => "${database}/${prefix}/${id}", |
| 323 |
|
config => $self->{databases}->{$database}, |
| 324 |
|
); |
| 325 |
|
|
| 326 |
|
$times->{apply} += time() - $t; |
| 327 |
|
|
| 328 |
|
$t = time(); |
| 329 |
|
|
| 330 |
|
$html = decode($self->{webpac_encoding}, $html); |
| 331 |
|
|
| 332 |
|
$times->{decode} += time() - $t; |
| 333 |
|
|
| 334 |
#$log->debug( '@html_results = ' . Dumper( \@html_results ) ); |
push @results, $html; |
| 335 |
|
} |
| 336 |
|
|
| 337 |
|
} |
| 338 |
|
|
| 339 |
$log->debug( sprintf( |
$log->debug( sprintf( |
| 340 |
"time spent: db = %.2f, out = %.2f, iconv = %.2f", |
"duration breakdown: estraier %.6fs, hash %.6fs, store %.6fs, apply %.6fs, decode %.06f, total: %.6fs", |
| 341 |
$times->{db}, $times->{out}, $times->{iconv}, |
$times->{est}, $times->{hash}, $times->{db}, $times->{apply}, $times->{decode}, time() - $search_start_t, |
| 342 |
) ); |
) ); |
| 343 |
|
|
| 344 |
return \@html_results; |
return \@results; |
| 345 |
|
} |
| 346 |
|
|
| 347 |
|
=head2 hints |
| 348 |
|
|
| 349 |
|
my $hints = $m->hints; |
| 350 |
|
|
| 351 |
|
Return various useful hints about result |
| 352 |
|
|
| 353 |
|
=cut |
| 354 |
|
|
| 355 |
|
sub hints { |
| 356 |
|
my $self = shift; |
| 357 |
|
|
| 358 |
|
unless ($self->{hints}) { |
| 359 |
|
$self->{log}->fatal("no hints found!"); |
| 360 |
|
return; |
| 361 |
|
} |
| 362 |
|
|
| 363 |
|
my $hints; |
| 364 |
|
|
| 365 |
|
while (my ($key,$val) = each %{ $self->{hints} }) { |
| 366 |
|
|
| 367 |
|
#$self->{log}->debug("current hint $key = $val"); |
| 368 |
|
|
| 369 |
|
if ($key =~ m/^(?:HITS*|TIME|DOCNUM|WORDNUM)$/) { |
| 370 |
|
$hints->{ lc($key) } = $val; |
| 371 |
|
} elsif ($key =~ m/^HINT#/) { |
| 372 |
|
my ($word,$count) = split(/\t/,$val,2); |
| 373 |
|
$hints->{words}->{$word} = $count; |
| 374 |
|
} elsif ($key =~ m/^LINK#/) { |
| 375 |
|
my ($url,undef,undef,undef,undef,undef,$results) = split(/\t/,$val,7); |
| 376 |
|
if ($url =~ m#/node/(.+)$#) { |
| 377 |
|
$hints->{node}->{$1} = $results; |
| 378 |
|
} else { |
| 379 |
|
$self->{log}->debug("url $url doesn't have /node/ in it!"); |
| 380 |
|
} |
| 381 |
|
} else { |
| 382 |
|
$self->{log}->debug("unknown hint $key = $val"); |
| 383 |
|
} |
| 384 |
|
|
| 385 |
|
} |
| 386 |
|
|
| 387 |
|
$self->{log}->dumper($hints, 'model hints' ); |
| 388 |
|
|
| 389 |
|
return $hints; |
| 390 |
|
} |
| 391 |
|
|
| 392 |
|
|
| 393 |
|
=head2 record |
| 394 |
|
|
| 395 |
|
my $html = $m->record( |
| 396 |
|
mfn => 42, |
| 397 |
|
template => 'foo.tt', |
| 398 |
|
); |
| 399 |
|
|
| 400 |
|
This will load one record, convert it to html using C<template> and return |
| 401 |
|
it. |
| 402 |
|
|
| 403 |
|
=cut |
| 404 |
|
|
| 405 |
|
sub record { |
| 406 |
|
my $self = shift; |
| 407 |
|
|
| 408 |
|
my $args = {@_}; |
| 409 |
|
my $log = $self->{log}; |
| 410 |
|
$log->dumper( $args, 'args' ); |
| 411 |
|
|
| 412 |
|
foreach my $f (qw/record_uri template/) { |
| 413 |
|
$log->fatal("need $f") unless ($args->{$f}); |
| 414 |
|
} |
| 415 |
|
|
| 416 |
|
my ($database, $prefix, $id); |
| 417 |
|
|
| 418 |
|
if ($args->{record_uri} =~ m#^([^/]+)/([^/]+)/([^/]+)$#) { |
| 419 |
|
($database, $prefix, $id) = ($1,$2,$3); |
| 420 |
|
} else { |
| 421 |
|
$log->error("can't parse $args->{record_uri} into prefix, database and uri"); |
| 422 |
|
return; |
| 423 |
|
} |
| 424 |
|
|
| 425 |
|
my $ds = $self->{db}->load_ds( id => $id, prefix => $prefix, database => $database ); |
| 426 |
|
if (! $ds) { |
| 427 |
|
$log->error("can't load_ds( $database/$prefix/$id )"); |
| 428 |
|
return; |
| 429 |
|
} |
| 430 |
|
|
| 431 |
|
my $html = $self->apply( |
| 432 |
|
template => $args->{template}, |
| 433 |
|
data => $ds, |
| 434 |
|
record_uri => $args->{record_uri}, |
| 435 |
|
config => $self->{databases}->{$database}, |
| 436 |
|
); |
| 437 |
|
|
| 438 |
|
$html = decode($self->{webpac_encoding}, $html); |
| 439 |
|
|
| 440 |
|
return $html; |
| 441 |
|
} |
| 442 |
|
|
| 443 |
|
|
| 444 |
|
=head2 list_nodes |
| 445 |
|
|
| 446 |
|
my @nodes = $m->list_nodes( 'site' ); |
| 447 |
|
|
| 448 |
|
Return all databases which have records for selected site. Returned array of |
| 449 |
|
hashes has elements C<name> and C<label>. |
| 450 |
|
|
| 451 |
|
This function does cacheing inside C<< $self->{nodes_in_site} >>, but you |
| 452 |
|
probably didn't want to know that. |
| 453 |
|
|
| 454 |
|
=cut |
| 455 |
|
|
| 456 |
|
sub list_nodes { |
| 457 |
|
my $self = shift; |
| 458 |
|
|
| 459 |
|
my $site = shift || $self->{defaultnode}; |
| 460 |
|
|
| 461 |
|
# cache? |
| 462 |
|
if ($self->{nodes_in_site}->{$site} && ref($self->{nodes_in_site}->{$site}) eq 'ARRAY') { |
| 463 |
|
$self->{log}->debug("list_nodes for site $site and returns from cache"); |
| 464 |
|
return @{ $self->{nodes_in_site}->{$site} }; |
| 465 |
|
}; |
| 466 |
|
|
| 467 |
|
$self->{log}->debug("list_nodes for site $site"); |
| 468 |
|
|
| 469 |
|
$self->setup_site( $site ); |
| 470 |
|
|
| 471 |
|
my @nodes; |
| 472 |
|
|
| 473 |
|
if ($self->{est_node}->doc_num > 0) { |
| 474 |
|
push @nodes, { |
| 475 |
|
name => $self->{est_node}->name, |
| 476 |
|
label => $self->{est_node}->label, |
| 477 |
|
doc_num => $self->{est_node}->doc_num, |
| 478 |
|
} |
| 479 |
|
} |
| 480 |
|
|
| 481 |
|
# refresh set info |
| 482 |
|
$self->{est_node}->_set_info; |
| 483 |
|
|
| 484 |
|
my $links = $self->{est_node}->links || return @nodes; |
| 485 |
|
|
| 486 |
|
$self->{log}->dumper( $links, 'links' ); |
| 487 |
|
|
| 488 |
|
foreach my $link (@{ $links }) { |
| 489 |
|
my ($url, $label, $credit) = split(/\t/, $link, 3); |
| 490 |
|
if ($url =~ m#/node/(.+)$#) { |
| 491 |
|
my $node = $1; |
| 492 |
|
$self->setup_site( $node ); |
| 493 |
|
$self->{est_node}->_set_info; |
| 494 |
|
$label = decode('UTF-8', $label); |
| 495 |
|
push @nodes, { |
| 496 |
|
name => $node, |
| 497 |
|
label => $label, |
| 498 |
|
doc_num => $self->{est_node}->doc_num, |
| 499 |
|
} |
| 500 |
|
} else { |
| 501 |
|
$self->{log}->warn("can't find node name in link $link"); |
| 502 |
|
} |
| 503 |
|
} |
| 504 |
|
|
| 505 |
|
$self->setup_site( $site ); |
| 506 |
|
$self->{est_node}->_set_info; |
| 507 |
|
|
| 508 |
|
$self->{nodes_in_site}->{$site} = \@nodes; |
| 509 |
|
|
| 510 |
|
return @nodes; |
| 511 |
} |
} |
| 512 |
|
|
| 513 |
=head2 save_html |
=head2 save_html |
| 514 |
|
|
| 515 |
$m->save_html( '/full/path/to/file', $content ); |
$m->save_html( '/full/path/to/file', $content ); |
| 516 |
|
|
| 517 |
It will use C<iconv_on_save> to convert content encoding back to |
It will use C<Encode> to convert content encoding back to |
| 518 |
Webpac codepage, recode JavaScript Unicode entities (%u1234), |
Webpac codepage, recode JavaScript Unicode entities (%u1234), |
| 519 |
strip extra newlines at beginning and end, and save to |
strip extra newlines at beginning and end, and save to |
| 520 |
C</full/path/to/file.new> and if that succeeds, just rename |
C</full/path/to/file.new> and if that succeeds, just rename |
| 525 |
sub save_html { |
sub save_html { |
| 526 |
my ($self, $path, $content) = @_; |
my ($self, $path, $content) = @_; |
| 527 |
|
|
| 528 |
$content = $self->iconv_on_save( $content ) || die "no content?"; |
# FIXME Should this be UTF-8 or someting? |
| 529 |
|
my $js_encoding = $self->{webpac_encoding}; |
| 530 |
|
$js_encoding = 'UTF-16'; |
| 531 |
|
|
| 532 |
sub _conv_js { |
sub _conv_js { |
| 533 |
my $t = shift || return; |
return '0x' . $_[1]; |
| 534 |
return $self->{iconv}->convert(chr(hex($t))); |
return encode($_[0], chr(hex($_[1]))); |
| 535 |
} |
} |
| 536 |
$content =~ s/%u([a-fA-F0-9]{4})/_conv_js($1)/gex; |
#$content =~ s/%u([a-fA-F0-9]{4})/_conv_js($js_encoding,$1)/gex; |
| 537 |
$content =~ s/^[\n\r]+//s; |
$content =~ s/^[\n\r]+//s; |
| 538 |
$content =~ s/[\n\r]+$/\n/s; |
$content =~ s/[\n\r]+$/\n/s; |
| 539 |
|
$content =~ s/\n\r/\n/gs; |
| 540 |
|
|
| 541 |
write_file($path . '.new', $content) || die "can't save ${path}.new $!"; |
my $disk_encoding = $self->{webpac_encoding} || 'utf-8'; |
| 542 |
|
$self->{log}->debug("convert encoding to $disk_encoding"); |
| 543 |
|
from_to($content, 'utf-8', $disk_encoding) || $self->{log}->warn("encoding from utf-8 to $disk_encoding failed for: $content"); |
| 544 |
|
|
| 545 |
|
write_file($path . '.new', {binmode => ':raw' }, $content) || die "can't save ${path}.new $!"; |
| 546 |
rename $path . '.new', $path || die "can't rename to $path: $!"; |
rename $path . '.new', $path || die "can't rename to $path: $!"; |
| 547 |
} |
} |
| 548 |
|
|
| 561 |
|
|
| 562 |
die "no path?" unless ($path); |
die "no path?" unless ($path); |
| 563 |
|
|
| 564 |
my $content = read_file($path) || die "can't read $path: $!"; |
my $content = read_file($path, {binmode => ':raw' }) || die "can't read $path: $!"; |
|
#$content = $q->escapeHTML($iconv_utf8->convert($content)); |
|
|
$content = $self->{iconv}->convert($content); |
|
| 565 |
|
|
| 566 |
return $content; |
return decode($self->{webpac_encoding}, $content); |
| 567 |
} |
} |
| 568 |
|
|
| 569 |
|
|
| 570 |
|
=head2 apply |
| 571 |
|
|
| 572 |
|
Create output from in-memory data structure using Template Toolkit template. |
| 573 |
|
|
| 574 |
|
my $text = $tt->apply( |
| 575 |
|
template => 'text.tt', |
| 576 |
|
data => $ds, |
| 577 |
|
record_uri => 'database/prefix/mfn', |
| 578 |
|
); |
| 579 |
|
|
| 580 |
|
It also has follwing template toolikit filter routies defined: |
| 581 |
|
|
| 582 |
|
=cut |
| 583 |
|
|
| 584 |
|
# Escape <, >, & and ", and to produce valid XML |
| 585 |
|
my %escape = ('<'=>'<', '>'=>'>', '&'=>'&', '"'=>'"'); |
| 586 |
|
my $escape_re = join '|' => keys %escape; |
| 587 |
|
|
| 588 |
|
sub apply { |
| 589 |
|
my $self = shift; |
| 590 |
|
|
| 591 |
|
my $args = {@_}; |
| 592 |
|
|
| 593 |
|
my $log = $self->{log} || die "no log?"; |
| 594 |
|
|
| 595 |
|
foreach my $a (qw/template data/) { |
| 596 |
|
$log->fatal("need $a") unless ($args->{$a}); |
| 597 |
|
} |
| 598 |
|
|
| 599 |
|
=head3 tt_filter_type |
| 600 |
|
|
| 601 |
|
filter to return values of specified from $ds, usage from TT template is in form |
| 602 |
|
C<d('FieldName','delimiter')>, where C<delimiter> is optional, like this: |
| 603 |
|
|
| 604 |
|
[% d('Title') %] |
| 605 |
|
[% d('Author',', ' %] |
| 606 |
|
|
| 607 |
|
=cut |
| 608 |
|
|
| 609 |
|
sub tt_filter_type { |
| 610 |
|
my ($data,$type) = @_; |
| 611 |
|
|
| 612 |
|
die "no data?" unless ($data); |
| 613 |
|
$type ||= 'display'; |
| 614 |
|
|
| 615 |
|
my $default_delimiter = { |
| 616 |
|
'display' => '¶<br/>', |
| 617 |
|
'index' => '\n', |
| 618 |
|
}; |
| 619 |
|
|
| 620 |
|
return sub { |
| 621 |
|
|
| 622 |
|
my ($name,$join) = @_; |
| 623 |
|
|
| 624 |
|
die "no data hash" unless ($data->{'data'} && ref($data->{'data'}) eq 'HASH'); |
| 625 |
|
# Hm? Should we die here? |
| 626 |
|
return unless ($name); |
| 627 |
|
|
| 628 |
|
my $item = $data->{'data'}->{$name} || return; |
| 629 |
|
|
| 630 |
|
my $v = $item->{$type} || return; |
| 631 |
|
|
| 632 |
|
if (ref($v) eq 'ARRAY') { |
| 633 |
|
if ($#{$v} == 0) { |
| 634 |
|
$v = $v->[0]; |
| 635 |
|
$v =~ s/($escape_re)/$escape{$1}/g; |
| 636 |
|
} else { |
| 637 |
|
$join = $default_delimiter->{$type} unless defined($join); |
| 638 |
|
$v = join($join, map { |
| 639 |
|
s/($escape_re)/$escape{$1}/g; |
| 640 |
|
} @{$v}); |
| 641 |
|
} |
| 642 |
|
} else { |
| 643 |
|
warn("TT filter $type(): field $name values aren't ARRAY, ignoring"); |
| 644 |
|
} |
| 645 |
|
|
| 646 |
|
return $v; |
| 647 |
|
} |
| 648 |
|
} |
| 649 |
|
|
| 650 |
|
$args->{'d'} = tt_filter_type($args, 'display'); |
| 651 |
|
$args->{'display'} = tt_filter_type($args, 'display'); |
| 652 |
|
|
| 653 |
|
=head3 tt_filter_search |
| 654 |
|
|
| 655 |
|
filter to return links to search, usage in TT: |
| 656 |
|
|
| 657 |
|
[% search('FieldToDisplay','FieldToSearch','optional delimiter', 'optional_template.tt') %] |
| 658 |
|
|
| 659 |
|
=cut |
| 660 |
|
|
| 661 |
|
sub tt_filter_search { |
| 662 |
|
|
| 663 |
|
my ($data) = @_; |
| 664 |
|
|
| 665 |
|
die "no data?" unless ($data); |
| 666 |
|
|
| 667 |
|
return sub { |
| 668 |
|
|
| 669 |
|
my ($display,$search,$delimiter,$template) = @_; |
| 670 |
|
|
| 671 |
|
# default delimiter |
| 672 |
|
$delimiter ||= '¶<br/>', |
| 673 |
|
|
| 674 |
|
die "no data hash" unless ($data->{'data'} && ref($data->{'data'}) eq 'HASH'); |
| 675 |
|
# Hm? Should we die here? |
| 676 |
|
return unless ($display); |
| 677 |
|
|
| 678 |
|
my $item = $data->{'data'}->{$display} || return; |
| 679 |
|
|
| 680 |
|
return unless($item->{'display'}); |
| 681 |
|
if (! $item->{'search'}) { |
| 682 |
|
warn "error in TT template: field $display didn't insert anything into search, use d('$display') and not search('$display'...)"; |
| 683 |
|
return; |
| 684 |
|
} |
| 685 |
|
|
| 686 |
|
my @warn; |
| 687 |
|
foreach my $type (qw/display search/) { |
| 688 |
|
push @warn, "field $display type $type values aren't ARRAY" unless (ref($item->{$type}) eq 'ARRAY'); |
| 689 |
|
} |
| 690 |
|
|
| 691 |
|
if (@warn) { |
| 692 |
|
warn("TT filter search(): " . join(",", @warn) . ", skipping"); |
| 693 |
|
return; |
| 694 |
|
} |
| 695 |
|
my @html; |
| 696 |
|
|
| 697 |
|
my $d_el = $#{ $item->{'display'} }; |
| 698 |
|
my $s_el = $#{ $item->{'search'} }; |
| 699 |
|
|
| 700 |
|
# easy, both fields have same number of elements or there is just |
| 701 |
|
# one search and multiple display |
| 702 |
|
if ( $d_el == $s_el || $s_el == 0 ) { |
| 703 |
|
|
| 704 |
|
foreach my $i ( 0 .. $d_el ) { |
| 705 |
|
|
| 706 |
|
my $s; |
| 707 |
|
if ($s_el > 0) { |
| 708 |
|
$s = $item->{'search'}->[$i] or warn "can't find value $i for type search in field $search"; |
| 709 |
|
} else { |
| 710 |
|
$s = $item->{'search'}->[0]; |
| 711 |
|
} |
| 712 |
|
#$s =~ s/([^\w.-])/sprintf("%%%02X",ord($1))/eg; |
| 713 |
|
$s = __quotemeta( $s ); |
| 714 |
|
|
| 715 |
|
my $d = $item->{'display'}->[$i] or warn "can't find value $i for type display in field $display"; |
| 716 |
|
|
| 717 |
|
my $template_arg = ''; |
| 718 |
|
$template_arg = qq{,'$template'} if ($template); |
| 719 |
|
|
| 720 |
|
push @html, qq{<a href="#" onclick="return search_via_link('$search','$s'${template_arg})">$d</a>}; |
| 721 |
|
} |
| 722 |
|
|
| 723 |
|
return join($delimiter, @html); |
| 724 |
|
} else { |
| 725 |
|
my $html = qq{<div class="notice">WARNING: we should really support if there is $d_el display elements and $s_el search elements, but currently there is no nice way to do so, so we will just display values</div>}; |
| 726 |
|
my $v = $item->{'display'}; |
| 727 |
|
|
| 728 |
|
if ($#{$v} == 0) { |
| 729 |
|
$html .= $v->[0]; |
| 730 |
|
} else { |
| 731 |
|
$html .= join($delimiter, @{$v}); |
| 732 |
|
} |
| 733 |
|
return $html; |
| 734 |
|
} |
| 735 |
|
} |
| 736 |
|
} |
| 737 |
|
|
| 738 |
|
$args->{'search'} = tt_filter_search($args); |
| 739 |
|
|
| 740 |
|
=head3 load_rec |
| 741 |
|
|
| 742 |
|
Used mostly for onClick events like this: |
| 743 |
|
|
| 744 |
|
<a href="#" onClick="[% load_rec( record_uri, 'template_name.tt') %]>foo</a> |
| 745 |
|
|
| 746 |
|
It will automatically do sanity checking and create correct JavaScript code. |
| 747 |
|
|
| 748 |
|
=cut |
| 749 |
|
|
| 750 |
|
$args->{'load_rec'} = sub { |
| 751 |
|
my @errors; |
| 752 |
|
|
| 753 |
|
my $record_uri = shift or push @errors, "record_uri missing"; |
| 754 |
|
my $template = shift or push @errors, "template missing"; |
| 755 |
|
|
| 756 |
|
if ($record_uri !~ m#^[^/]+/[^/]+/[^/]+$#) { |
| 757 |
|
push @errors, "invalid format of record_uri: $record_uri"; |
| 758 |
|
} |
| 759 |
|
|
| 760 |
|
if (@errors) { |
| 761 |
|
return "Logger.error('errors in load_rec: " . join(", ", @errors) . "'); return false;"; |
| 762 |
|
} else { |
| 763 |
|
return "load_rec('$record_uri','$template'); return false;"; |
| 764 |
|
} |
| 765 |
|
}; |
| 766 |
|
|
| 767 |
|
=head3 load_template |
| 768 |
|
|
| 769 |
|
Used to re-submit search request and load results in different template |
| 770 |
|
|
| 771 |
|
<a href="#" onClick="[% load_template( 'template_name.tt' ) %]">bar</a> |
| 772 |
|
|
| 773 |
|
=cut |
| 774 |
|
|
| 775 |
|
$args->{'load_template'} = sub { |
| 776 |
|
my $template = shift or return "Logger.error('load_template missing template name!'); return false;"; |
| 777 |
|
return "load_template($template); return false;"; |
| 778 |
|
}; |
| 779 |
|
|
| 780 |
|
my $out; |
| 781 |
|
|
| 782 |
|
$self->{'tt'}->process( |
| 783 |
|
$args->{'template'}, |
| 784 |
|
$args, |
| 785 |
|
\$out |
| 786 |
|
) || $log->error( "apply can't process template: ", $self->{'tt'}->error() ); |
| 787 |
|
|
| 788 |
|
return $out; |
| 789 |
|
} |
| 790 |
|
|
| 791 |
|
|
| 792 |
|
=head2 __quotemeta |
| 793 |
|
|
| 794 |
|
Helper to quote JavaScript-friendly characters |
| 795 |
|
|
| 796 |
|
=cut |
| 797 |
|
|
| 798 |
|
sub __quotemeta { |
| 799 |
|
local $_ = shift; |
| 800 |
|
$_ = decode('iso-8859-2', $_); |
| 801 |
|
|
| 802 |
|
s<([\x{0080}-\x{fffd}]+)>{sprintf '\u%0*v4X', '\u', $1}ge if ( Encode::is_utf8($_) ); |
| 803 |
|
{ |
| 804 |
|
use bytes; |
| 805 |
|
s<((?:[^ \x21-\x7E]|(?:\\(?!u)))+)>{sprintf '\x%0*v2X', '\x', $1}ge; |
| 806 |
|
} |
| 807 |
|
|
| 808 |
|
s/\\x09/\\t/g; |
| 809 |
|
s/\\x0A/\\n/g; |
| 810 |
|
s/\\x0D/\\r/g; |
| 811 |
|
s/"/\\"/g; |
| 812 |
|
s/\\x5C/\\\\/g; |
| 813 |
|
|
| 814 |
|
return $_; |
| 815 |
|
} |
| 816 |
|
|
| 817 |
|
|
| 818 |
|
|
| 819 |
=head1 AUTHOR |
=head1 AUTHOR |
| 820 |
|
|
| 821 |
Dobrica Pavlinusic |
Dobrica Pavlinusic C<< <dpavlin@rot13.org> >> |
| 822 |
|
|
| 823 |
=head1 LICENSE |
=head1 LICENSE |
| 824 |
|
|
Parent Directory
| 
Revision Log
| 
Patch