| 1 |
# -*- Mode: Perl -*- |
| 2 |
# Table.pm -- |
| 3 |
# ITIID : $ITI$ $Header $__Header$ |
| 4 |
# Author : Ulrich Pfeifer |
| 5 |
# Created On : Thu Aug 8 13:05:10 1996 |
| 6 |
# Last Modified By: Ulrich Pfeifer |
| 7 |
# Last Modified On: Sun Nov 22 18:44:37 1998 |
| 8 |
# Language : CPerl |
| 9 |
# Update Count : 51 |
| 10 |
# Status : Unknown, Use with caution! |
| 11 |
# |
| 12 |
# Copyright (c) 1996-1997, Ulrich Pfeifer |
| 13 |
# |
| 14 |
|
| 15 |
=head1 NAME |
| 16 |
|
| 17 |
WAIT::Table -- Module for maintaining Tables / Relations |
| 18 |
|
| 19 |
=head1 SYNOPSIS |
| 20 |
|
| 21 |
require WAIT::Table; |
| 22 |
|
| 23 |
=head1 DESCRIPTION |
| 24 |
|
| 25 |
=cut |
| 26 |
|
| 27 |
package WAIT::Table; |
| 28 |
require WAIT::Parse::Base; |
| 29 |
use strict; |
| 30 |
use Carp; |
| 31 |
use DB_File; |
| 32 |
use Fcntl; |
| 33 |
|
| 34 |
my $USE_RECNO = 0; |
| 35 |
|
| 36 |
=head2 Creating a Table. |
| 37 |
|
| 38 |
The constructor WAIT::Table-<gt>new is normally called via the |
| 39 |
create_table method of a database handle. This is not enforced, but |
| 40 |
creating a table doesn not make any sense unless the table is |
| 41 |
registered by the database because the latter implements persistence |
| 42 |
of the meta data. Registering is done automatically by letting the |
| 43 |
database handle create a table. |
| 44 |
|
| 45 |
my $db = create WAIT::Database name => 'sample'; |
| 46 |
my $tb = $db->create_table (name => 'test', |
| 47 |
attr => ['docid', 'headline'], |
| 48 |
layout => $layout, |
| 49 |
access => $access, |
| 50 |
); |
| 51 |
|
| 52 |
The constructor returns a handle for the table. This handle is hidden by the |
| 53 |
table module, to prevent direct access if called via Table. |
| 54 |
|
| 55 |
=over 10 |
| 56 |
|
| 57 |
=item C<access> => I<accesobj> |
| 58 |
|
| 59 |
A reference to a acces object for the external parts (attributes) of |
| 60 |
tuples. As you may remember, the WAIT System does not enforce that |
| 61 |
objects are completely stored inside the system to avoid duplication. |
| 62 |
There is no (strong) point in storing all you HTML-Documents inside |
| 63 |
the system when indexing your WWW-Server. |
| 64 |
|
| 65 |
=item C<file> => I<fname> |
| 66 |
|
| 67 |
The filename of the records file. Files for indexes will have I<fname> |
| 68 |
as prefix. I<Mandatory> |
| 69 |
|
| 70 |
=item C<name> => I<name> |
| 71 |
|
| 72 |
The name of this table. I<Mandatory> |
| 73 |
|
| 74 |
=item C<attr> => [ I<attr> ... ] |
| 75 |
|
| 76 |
A reference to an array of attribute names. I<Mandatory> |
| 77 |
|
| 78 |
=item C<djk> => [ I<attr> ... ] |
| 79 |
|
| 80 |
A reference to an array of attribute names which make up the |
| 81 |
I<disjointness key>. Don't think about it - i's of no use yet; |
| 82 |
|
| 83 |
=item C<layout> => I<layoutobj> |
| 84 |
|
| 85 |
A reference to an external parser object. Defaults to anew instance of |
| 86 |
C<WAIT::Parse::Base> |
| 87 |
|
| 88 |
=item C<access> => I<accesobj> |
| 89 |
|
| 90 |
A reference to a acces object for the external parts of tuples. |
| 91 |
|
| 92 |
=back |
| 93 |
|
| 94 |
=cut |
| 95 |
|
| 96 |
sub new { |
| 97 |
my $type = shift; |
| 98 |
my %parm = @_; |
| 99 |
my $self = {}; |
| 100 |
|
| 101 |
# Do that before we eventually add '_weight' to attributes. |
| 102 |
$self->{keyset} = $parm{keyset} || [[@{$parm{attr}}]]; |
| 103 |
$self->{mode} = O_CREAT | O_RDWR; |
| 104 |
# Determine and set up subclass |
| 105 |
$type = ref($type) || $type; |
| 106 |
if (defined $parm{djk}) { |
| 107 |
if (@{$parm{djk}} == @{$parm{attr}}) { |
| 108 |
# All attributes in DK (sloppy test here!) |
| 109 |
$type .= '::Independent'; |
| 110 |
require WAIT::Table::Independent; |
| 111 |
} else { |
| 112 |
$type .= '::Disjoint'; |
| 113 |
require WAIT::Table::Disjoint; |
| 114 |
} |
| 115 |
# Add '_weight' to attributes |
| 116 |
my %attr; |
| 117 |
@attr{@{$parm{attr}}} = (1) x @{$parm{attr}}; |
| 118 |
unshift @{$parm{attr}}, '_weight' unless $attr{'_weight'}; |
| 119 |
} |
| 120 |
|
| 121 |
$self->{file} = $parm{file} or croak "No file specified"; |
| 122 |
if (-d $self->{file} or !mkdir($self->{file}, 0775)) { |
| 123 |
croak "Could not 'mkdir $self->{file}': $!\n"; |
| 124 |
} |
| 125 |
$self->{name} = $parm{name} or croak "No name specified"; |
| 126 |
$self->{attr} = $parm{attr} or croak "No attributes specified"; |
| 127 |
$self->{djk} = $parm{djk} if defined $parm{djk}; |
| 128 |
$self->{layout} = $parm{layout} || new WAIT::Parse::Base; |
| 129 |
$self->{access} = $parm{access} if defined $parm{access}; |
| 130 |
$self->{nextk} = 1; # next record to insert; first record unused |
| 131 |
$self->{deleted} = {}; # no deleted records yet |
| 132 |
$self->{indexes} = {}; |
| 133 |
|
| 134 |
bless $self, $type; |
| 135 |
# Call create_index() and create_index() for compatibility |
| 136 |
for (@{$self->{keyset}||[]}) { |
| 137 |
#carp "Specification of indexes at table create time is deprecated"; |
| 138 |
$self->create_index(@$_); |
| 139 |
} |
| 140 |
while (@{$parm{invindex}||[]}) { |
| 141 |
# carp "Specification of inverted indexes at table create time is deprecated"; |
| 142 |
my $att = shift @{$parm{invindex}}; |
| 143 |
my @spec = @{shift @{$parm{invindex}}}; |
| 144 |
my @opt; |
| 145 |
|
| 146 |
if (ref($spec[0])) { |
| 147 |
carp "Secondary pipelines are deprecated\n"; |
| 148 |
@opt = %{shift @spec}; |
| 149 |
} |
| 150 |
$self->create_inverted_index(attribute => $att, pipeline => \@spec, @opt); |
| 151 |
} |
| 152 |
$self; |
| 153 |
# end of backwarn compatibility stuff |
| 154 |
} |
| 155 |
|
| 156 |
=head2 Creating an index |
| 157 |
|
| 158 |
$tb->create_index('docid'); |
| 159 |
|
| 160 |
=item C<create_index> |
| 161 |
|
| 162 |
must be called with a list of attributes. This must be a subset of the |
| 163 |
attributes specified when the table was created. Currently this |
| 164 |
method must be called before the first tuple is inserted in the |
| 165 |
table! |
| 166 |
|
| 167 |
=cut |
| 168 |
|
| 169 |
sub create_index { |
| 170 |
my $self= shift; |
| 171 |
|
| 172 |
croak "Cannot create index for table aready populated" |
| 173 |
if $self->{nextk} > 1; |
| 174 |
|
| 175 |
require WAIT::Index; |
| 176 |
|
| 177 |
my $name = join '-', @_; |
| 178 |
$self->{indexes}->{$name} = |
| 179 |
new WAIT::Index file => $self->{file}.'/'.$name, attr => $_; |
| 180 |
} |
| 181 |
|
| 182 |
=head2 Creating an inverted index |
| 183 |
|
| 184 |
$tb->create_inverted_index |
| 185 |
(attribute => 'au', |
| 186 |
pipeline => ['detex', 'isotr', 'isolc', 'split2', 'stop'], |
| 187 |
predicate => 'plain', |
| 188 |
); |
| 189 |
|
| 190 |
=over 5 |
| 191 |
|
| 192 |
=item C<attribute> |
| 193 |
|
| 194 |
The attribute to build the index on. This attribute may not be in the |
| 195 |
set attributes specified when the table was created. |
| 196 |
|
| 197 |
=item C<pipeline> |
| 198 |
|
| 199 |
A piplines specification is a reference to and array of method names |
| 200 |
(from package C<WAIT::Filter>) which are to applied in sequence to the |
| 201 |
contents of the named attribute. The attribute name may not be in the |
| 202 |
attribute list. |
| 203 |
|
| 204 |
=item C<predicate> |
| 205 |
|
| 206 |
An indication which predicate the index implements. This may be |
| 207 |
e.g. 'plain', 'stemming' or 'soundex'. The indicator will be used for |
| 208 |
query processing. Currently there is no standard set of predicate |
| 209 |
names. The predicate defaults to the last member of the ppline if |
| 210 |
omitted. |
| 211 |
|
| 212 |
=back |
| 213 |
|
| 214 |
Currently this method must be called before the first tuple is |
| 215 |
inserted in the table! |
| 216 |
|
| 217 |
=cut |
| 218 |
|
| 219 |
sub create_inverted_index { |
| 220 |
my $self = shift; |
| 221 |
my %parm = @_; |
| 222 |
|
| 223 |
croak "No attribute specified" unless $parm{attribute}; |
| 224 |
croak "No pipeline specified" unless $parm{pipeline}; |
| 225 |
|
| 226 |
$parm{predicate} ||= $parm{pipeline}->[-1]; |
| 227 |
|
| 228 |
croak "Cannot create index for table aready populated" |
| 229 |
if $self->{nextk} > 1; |
| 230 |
|
| 231 |
require WAIT::InvertedIndex; |
| 232 |
|
| 233 |
# backward compatibility stuff |
| 234 |
my %opt = %parm; |
| 235 |
for (qw(attribute pipeline predicate)) { |
| 236 |
delete $opt{$_}; |
| 237 |
} |
| 238 |
|
| 239 |
my $name = join '_', ($parm{attribute}, @{$parm{pipeline}}); |
| 240 |
my $idx = new WAIT::InvertedIndex(file => $self->{file}.'/'.$name, |
| 241 |
filter => [@{$parm{pipeline}}], # clone |
| 242 |
name => $name, |
| 243 |
attr => $parm{attribute}, |
| 244 |
%opt, # backward compatibility stuff |
| 245 |
); |
| 246 |
# We will have to use $parm{predicate} here |
| 247 |
push @{$self->{inverted}->{$parm{attribute}}}, $idx; |
| 248 |
} |
| 249 |
|
| 250 |
sub dir { |
| 251 |
$_[0]->{file}; |
| 252 |
} |
| 253 |
|
| 254 |
=head2 C<$tb-E<gt>layout> |
| 255 |
|
| 256 |
Returns the reference to the associated parser object. |
| 257 |
|
| 258 |
=cut |
| 259 |
|
| 260 |
sub layout { $_[0]->{layout} } |
| 261 |
|
| 262 |
=head2 C<$tb-E<gt>fields> |
| 263 |
|
| 264 |
Returns the array of attribute names. |
| 265 |
|
| 266 |
=cut |
| 267 |
|
| 268 |
|
| 269 |
sub fields { keys %{$_[0]->{inverted}}} |
| 270 |
|
| 271 |
=head2 C<$tb-E<gt>drop> |
| 272 |
|
| 273 |
Must be called via C<WAIT::Database::drop_table> |
| 274 |
|
| 275 |
=cut |
| 276 |
|
| 277 |
sub drop { |
| 278 |
my $self = shift; |
| 279 |
if ((caller)[0] eq 'WAIT::Database') { # database knows about this |
| 280 |
$self->close; # just make sure |
| 281 |
my $file = $self->{file}; |
| 282 |
|
| 283 |
for (values %{$self->{indexes}}) { |
| 284 |
$_->drop; |
| 285 |
} |
| 286 |
unlink "$file/records"; |
| 287 |
! (!-e $file or rmdir $file); |
| 288 |
} else { |
| 289 |
croak ref($self)."::drop called directly"; |
| 290 |
} |
| 291 |
} |
| 292 |
|
| 293 |
sub mrequire ($) { |
| 294 |
my $module = shift; |
| 295 |
|
| 296 |
$module =~ s{::}{/}g; |
| 297 |
$module .= '.pm'; |
| 298 |
require $module; |
| 299 |
} |
| 300 |
|
| 301 |
sub open { |
| 302 |
my $self = shift; |
| 303 |
my $file = $self->{file} . '/records'; |
| 304 |
|
| 305 |
mrequire ref($self); # that's tricky eh? |
| 306 |
if (defined $self->{'layout'}) { |
| 307 |
mrequire ref($self->{'layout'}); |
| 308 |
} |
| 309 |
if (defined $self->{'access'}) { |
| 310 |
mrequire ref($self->{'access'}); |
| 311 |
} |
| 312 |
if (exists $self->{indexes}) { |
| 313 |
require WAIT::Index; |
| 314 |
for (values %{$self->{indexes}}) { |
| 315 |
$_->{mode} = $self->{mode}; |
| 316 |
} |
| 317 |
} |
| 318 |
if (exists $self->{inverted}) { |
| 319 |
my ($att, $idx); |
| 320 |
for $att (keys %{$self->{inverted}}) { |
| 321 |
for $idx (@{$self->{inverted}->{$att}}) { |
| 322 |
$idx->{mode} = $self->{mode}; |
| 323 |
} |
| 324 |
} |
| 325 |
require WAIT::InvertedIndex; |
| 326 |
} |
| 327 |
unless (defined $self->{dbh}) { |
| 328 |
if ($USE_RECNO) { |
| 329 |
$self->{dbh} = tie(@{$self->{db}}, 'DB_File', $file, |
| 330 |
$self->{mode}, 0664, $DB_RECNO); |
| 331 |
} else { |
| 332 |
$self->{dbh} = |
| 333 |
tie(%{$self->{db}}, 'DB_File', $file, |
| 334 |
$self->{mode}, 0664, $DB_BTREE); |
| 335 |
} |
| 336 |
} |
| 337 |
$self; |
| 338 |
} |
| 339 |
|
| 340 |
sub fetch_extern { |
| 341 |
my $self = shift; |
| 342 |
|
| 343 |
print "#@_", $self->{'access'}->{Mode}, "\n"; |
| 344 |
if (exists $self->{'access'}) { |
| 345 |
mrequire ref($self->{'access'}); |
| 346 |
$self->{'access'}->FETCH(@_); |
| 347 |
} |
| 348 |
} |
| 349 |
|
| 350 |
sub fetch_extern_by_id { |
| 351 |
my $self = shift; |
| 352 |
|
| 353 |
$self->fetch_extern($self->fetch(@_)); |
| 354 |
} |
| 355 |
|
| 356 |
sub _find_index { |
| 357 |
my $self = shift; |
| 358 |
my (@att) = @_; |
| 359 |
my %att; |
| 360 |
my $name; |
| 361 |
|
| 362 |
@att{@att} = @att; |
| 363 |
|
| 364 |
KEY: for $name (keys %{$self->{indexes}}) { |
| 365 |
my @iat = split /-/, $name; |
| 366 |
for (@iat) { |
| 367 |
next KEY unless exists $att{$_}; |
| 368 |
} |
| 369 |
return $self->{indexes}->{$name}; |
| 370 |
} |
| 371 |
return undef; |
| 372 |
} |
| 373 |
|
| 374 |
sub have { |
| 375 |
my $self = shift; |
| 376 |
my %parm = @_; |
| 377 |
|
| 378 |
my $index = $self->_find_index(keys %parm); |
| 379 |
croak "No index found" unless $index; |
| 380 |
defined $self->{db} or $self->open; |
| 381 |
return $index->have(@_); |
| 382 |
} |
| 383 |
|
| 384 |
sub insert { |
| 385 |
my $self = shift; |
| 386 |
my %parm = @_; |
| 387 |
|
| 388 |
defined $self->{db} or $self->open; |
| 389 |
|
| 390 |
my $tuple = join($;, map($parm{$_} || '', @{$self->{attr}})); |
| 391 |
my $key; |
| 392 |
my @deleted = keys %{$self->{deleted}}; |
| 393 |
|
| 394 |
if (@deleted) { |
| 395 |
$key = pop @deleted; |
| 396 |
delete $self->{deleted}->{$key}; |
| 397 |
} else { |
| 398 |
$key = $self->{nextk}++; |
| 399 |
} |
| 400 |
if ($USE_RECNO) { |
| 401 |
$self->{db}->[$key] = $tuple; |
| 402 |
} else { |
| 403 |
$self->{db}->{$key} = $tuple; |
| 404 |
} |
| 405 |
for (values %{$self->{indexes}}) { |
| 406 |
unless ($_->insert($key, %parm)) { |
| 407 |
# duplicate key, undo changes |
| 408 |
if ($key == $self->{nextk}-1) { |
| 409 |
$self->{nextk}--; |
| 410 |
} else { |
| 411 |
$self->{deleted}->{$key}=1; |
| 412 |
} |
| 413 |
my $idx; |
| 414 |
for $idx (values %{$self->{indexes}}) { |
| 415 |
last if $idx eq $_; |
| 416 |
$idx->remove($key, %parm); |
| 417 |
} |
| 418 |
return undef; |
| 419 |
} |
| 420 |
} |
| 421 |
if (defined $self->{inverted}) { |
| 422 |
my $att; |
| 423 |
for $att (keys %{$self->{inverted}}) { |
| 424 |
if (defined $parm{$att}) { |
| 425 |
map $_->insert($key, $parm{$att}), @{$self->{inverted}->{$att}}; |
| 426 |
#map $_->sync, @{$self->{inverted}->{$att}} |
| 427 |
} |
| 428 |
} |
| 429 |
} |
| 430 |
$key |
| 431 |
} |
| 432 |
|
| 433 |
sub sync { |
| 434 |
my $self = shift; |
| 435 |
|
| 436 |
for (values %{$self->{indexes}}) { |
| 437 |
map $_->sync, $_; |
| 438 |
} |
| 439 |
if (defined $self->{inverted}) { |
| 440 |
my $att; |
| 441 |
for $att (keys %{$self->{inverted}}) { |
| 442 |
map $_->sync, @{$self->{inverted}->{$att}} |
| 443 |
} |
| 444 |
} |
| 445 |
} |
| 446 |
|
| 447 |
sub fetch { |
| 448 |
my $self = shift; |
| 449 |
my $key = shift; |
| 450 |
|
| 451 |
return () if exists $self->{deleted}->{$key}; |
| 452 |
|
| 453 |
defined $self->{db} or $self->open; |
| 454 |
if ($USE_RECNO) { |
| 455 |
$self->unpack($self->{db}->[$key]); |
| 456 |
} else { |
| 457 |
$self->unpack($self->{db}->{$key}); |
| 458 |
} |
| 459 |
} |
| 460 |
|
| 461 |
sub delete_by_key { |
| 462 |
my $self = shift; |
| 463 |
my $key = shift; |
| 464 |
|
| 465 |
return $self->{deleted}->{$key} if defined $self->{deleted}->{$key}; |
| 466 |
my %tuple = $self->fetch($key); |
| 467 |
for (values %{$self->{indexes}}) { |
| 468 |
$_->delete($key, %tuple); |
| 469 |
} |
| 470 |
if (defined $self->{inverted}) { |
| 471 |
# User *must* provide the full record for this or the entries |
| 472 |
# in the inverted index will not be removed |
| 473 |
%tuple = (%tuple, @_); |
| 474 |
my $att; |
| 475 |
for $att (keys %{$self->{inverted}}) { |
| 476 |
if (defined $tuple{$att}) { |
| 477 |
map $_->delete($key, $tuple{$att}), @{$self->{inverted}->{$att}} |
| 478 |
} |
| 479 |
} |
| 480 |
} |
| 481 |
++$self->{deleted}->{$key}; |
| 482 |
} |
| 483 |
|
| 484 |
sub delete { |
| 485 |
my $self = shift; |
| 486 |
my $tkey = $self->have(@_); |
| 487 |
|
| 488 |
defined $tkey && $self->delete_by_key($tkey, @_); |
| 489 |
} |
| 490 |
|
| 491 |
sub unpack { |
| 492 |
my $self = shift; |
| 493 |
my $tuple = shift; |
| 494 |
|
| 495 |
my $att; |
| 496 |
my @result; |
| 497 |
my @tuple = split /$;/, $tuple; |
| 498 |
|
| 499 |
for $att (@{$self->{attr}}) { |
| 500 |
push @result, $att, shift @tuple; |
| 501 |
} |
| 502 |
@result; |
| 503 |
} |
| 504 |
|
| 505 |
sub close { |
| 506 |
my $self = shift; |
| 507 |
|
| 508 |
if (exists $self->{'access'}) { |
| 509 |
eval {$self->{'access'}->close}; # dont bother if not opened |
| 510 |
} |
| 511 |
for (values %{$self->{indexes}}) { |
| 512 |
$_->close(); |
| 513 |
} |
| 514 |
if (defined $self->{inverted}) { |
| 515 |
my $att; |
| 516 |
for $att (keys %{$self->{inverted}}) { |
| 517 |
if ($] > 5.003) { # avoid bug in perl up to 5.003_05 |
| 518 |
my $idx; |
| 519 |
for $idx (@{$self->{inverted}->{$att}}) { |
| 520 |
$idx->close; |
| 521 |
} |
| 522 |
} else { |
| 523 |
map $_->close(), @{$self->{inverted}->{$att}}; |
| 524 |
} |
| 525 |
} |
| 526 |
} |
| 527 |
if ($self->{dbh}) { |
| 528 |
delete $self->{dbh}; |
| 529 |
|
| 530 |
if ($USE_RECNO) { |
| 531 |
untie @{$self->{db}}; |
| 532 |
} else { |
| 533 |
untie %{$self->{db}}; |
| 534 |
} |
| 535 |
delete $self->{db}; |
| 536 |
} |
| 537 |
|
| 538 |
1; |
| 539 |
} |
| 540 |
|
| 541 |
sub open_scan { |
| 542 |
my $self = shift; |
| 543 |
my $code = shift; |
| 544 |
|
| 545 |
$self->{dbh} or $self->open; |
| 546 |
require WAIT::Scan; |
| 547 |
new WAIT::Scan $self, $self->{nextk}-1, $code; |
| 548 |
} |
| 549 |
|
| 550 |
sub open_index_scan { |
| 551 |
my $self = shift; |
| 552 |
my $attr = shift; |
| 553 |
my $code = shift; |
| 554 |
my $name = join '-', @$attr; |
| 555 |
|
| 556 |
if (defined $self->{indexes}->{$name}) { |
| 557 |
$self->{indexes}->{$name}->open_scan($code); |
| 558 |
} else { |
| 559 |
croak "No such index '$name'"; |
| 560 |
} |
| 561 |
} |
| 562 |
|
| 563 |
eval {sub WAIT::Query::Raw::new} unless defined \&WAIT::Query::Raw::new; |
| 564 |
|
| 565 |
sub prefix { |
| 566 |
my ($self , $attr, $prefix) = @_; |
| 567 |
my %result; |
| 568 |
|
| 569 |
defined $self->{db} or $self->open; # require layout |
| 570 |
|
| 571 |
for (@{$self->{inverted}->{$attr}}) { |
| 572 |
my $result = $_->prefix($prefix); |
| 573 |
if (defined $result) { |
| 574 |
$result{$_->name} = $result; |
| 575 |
} |
| 576 |
} |
| 577 |
bless \%result, 'WAIT::Query::Raw'; |
| 578 |
} |
| 579 |
|
| 580 |
sub intervall { |
| 581 |
my ($self, $attr, $lb, $ub) = @_; |
| 582 |
my %result; |
| 583 |
|
| 584 |
defined $self->{db} or $self->open; # require layout |
| 585 |
|
| 586 |
for (@{$self->{inverted}->{$attr}}) { |
| 587 |
my $result = $_->intervall($lb, $ub); |
| 588 |
if (defined $result) { |
| 589 |
$result{$_->name} = $result; |
| 590 |
} |
| 591 |
} |
| 592 |
bless \%result, 'WAIT::Query::Raw'; |
| 593 |
} |
| 594 |
|
| 595 |
sub search { |
| 596 |
my $self = shift; |
| 597 |
my $attr = shift; |
| 598 |
my $cont = shift; |
| 599 |
my $raw = shift; |
| 600 |
my %result; |
| 601 |
|
| 602 |
defined $self->{db} or $self->open; # require layout |
| 603 |
|
| 604 |
if ($raw) { |
| 605 |
for (@{$self->{inverted}->{$attr}}) { |
| 606 |
my $name = $_->name; |
| 607 |
if (exists $raw->{$name} and @{$raw->{$name}}) { |
| 608 |
my $scale = 1/scalar(@{$raw->{$name}}); |
| 609 |
my %r = $_->search_raw(@{$raw->{$name}}); |
| 610 |
my ($key, $val); |
| 611 |
while (($key, $val) = each %r) { |
| 612 |
if (exists $result{$key}) { |
| 613 |
$result{$key} += $val*$scale; |
| 614 |
} else { |
| 615 |
$result{$key} = $val*$scale; |
| 616 |
} |
| 617 |
} |
| 618 |
} |
| 619 |
} |
| 620 |
} |
| 621 |
if (defined $cont and $cont ne '') { |
| 622 |
for (@{$self->{inverted}->{$attr}}) { |
| 623 |
my %r = $_->search($cont); |
| 624 |
my ($key, $val); |
| 625 |
while (($key, $val) = each %r) { |
| 626 |
if (exists $result{$key}) { |
| 627 |
$result{$key} += $val; |
| 628 |
} else { |
| 629 |
$result{$key} = $val; |
| 630 |
} |
| 631 |
} |
| 632 |
} |
| 633 |
} |
| 634 |
# sanity check for deleted documents. |
| 635 |
# this should not be necessary !@#$ |
| 636 |
for (keys %result) { |
| 637 |
delete $result{$_} if $self->{deleted}->{$_} |
| 638 |
} |
| 639 |
%result; |
| 640 |
} |
| 641 |
|
| 642 |
sub hilight_positions { |
| 643 |
my ($self, $attr, $text, $query, $raw) = @_; |
| 644 |
my %pos; |
| 645 |
|
| 646 |
if (defined $raw) { |
| 647 |
for (@{$self->{inverted}->{$attr}}) { |
| 648 |
my $name = $_->name; |
| 649 |
if (exists $raw->{$name}) { |
| 650 |
my %qt; |
| 651 |
grep $qt{$_}++, @{$raw->{$name}}; |
| 652 |
for ($_->parse_pos($text)) { |
| 653 |
if (exists $qt{$_->[0]}) { |
| 654 |
$pos{$_->[1]} = max($pos{$_->[1]}, length($_->[0])); |
| 655 |
} |
| 656 |
} |
| 657 |
} |
| 658 |
} |
| 659 |
} |
| 660 |
if (defined $query) { |
| 661 |
for (@{$self->{inverted}->{$attr}}) { |
| 662 |
my %qt; |
| 663 |
|
| 664 |
grep $qt{$_}++, $_->parse($query); |
| 665 |
for ($_->parse_pos($text)) { |
| 666 |
if (exists $qt{$_->[0]}) { |
| 667 |
if (exists $pos{$_->[1]}) { # perl -w ;-) |
| 668 |
$pos{$_->[1]} = max($pos{$_->[1]}, length($_->[0])); |
| 669 |
} else { |
| 670 |
$pos{$_->[1]} = length($_->[0]); |
| 671 |
} |
| 672 |
} |
| 673 |
} |
| 674 |
} |
| 675 |
} |
| 676 |
|
| 677 |
\%pos; |
| 678 |
} |
| 679 |
|
| 680 |
sub hilight { |
| 681 |
my ($tb, $text, $query, $raw) = @_; |
| 682 |
my $type = $tb->layout(); |
| 683 |
my @result; |
| 684 |
|
| 685 |
$query ||= {}; |
| 686 |
$raw ||= {}; |
| 687 |
my @ttxt = $type->tag($text); |
| 688 |
while (@ttxt) { |
| 689 |
no strict 'refs'; |
| 690 |
my %tag = %{shift @ttxt}; |
| 691 |
my $txt = shift @ttxt; |
| 692 |
my $fld; |
| 693 |
|
| 694 |
my %hl; |
| 695 |
for $fld (grep defined $tag{$_}, keys %$query, keys %$raw) { |
| 696 |
my $hp = $tb->hilight_positions($fld, $txt, |
| 697 |
$query->{$fld}, $raw->{$fld}); |
| 698 |
for (keys %$hp) { |
| 699 |
if (exists $hl{$_}) { # -w ;-( |
| 700 |
$hl{$_} = max($hl{$_}, $hp->{$_}); |
| 701 |
} else { |
| 702 |
$hl{$_} = $hp->{$_}; |
| 703 |
} |
| 704 |
} |
| 705 |
} |
| 706 |
my $pos; |
| 707 |
my $qt = {_qt => 1, %tag}; |
| 708 |
my $pl = \%tag; |
| 709 |
my $last = length($txt); |
| 710 |
my @tmp; |
| 711 |
for $pos (sort {$b <=> $a} keys %hl) { |
| 712 |
unshift @tmp, $pl, substr($txt,$pos+$hl{$pos},$last-$pos-$hl{$pos}); |
| 713 |
unshift @tmp, $qt, substr($txt,$pos,$hl{$pos}); |
| 714 |
$last = $pos; |
| 715 |
} |
| 716 |
push @result, $pl, substr($txt,0,$last); |
| 717 |
push @result, @tmp; |
| 718 |
} |
| 719 |
@result; # no speed necessary |
| 720 |
} |
| 721 |
|
| 722 |
1; |
| 723 |
|
Parent Directory
| 
Revision Log