1 |
package SWISH::PlusPlus; |
2 |
|
3 |
use 5.008004; |
4 |
use strict; |
5 |
use warnings; |
6 |
|
7 |
our $VERSION = '0.10'; |
8 |
|
9 |
use Carp; |
10 |
use File::Temp qw/ tempdir /; |
11 |
use BerkeleyDB; |
12 |
#use YAML; |
13 |
|
14 |
=head1 NAME |
15 |
|
16 |
SWISH::PlusPlus - Perl extension SWISH++ |
17 |
|
18 |
=head1 SYNOPSIS |
19 |
|
20 |
use SWISH::PlusPlus; |
21 |
blah blah blah |
22 |
|
23 |
=head1 DESCRIPTION |
24 |
|
25 |
This is perl module to use SWISH++ indexer by Paul J. Lucas. SWISH++ is |
26 |
rewrite of swish-e in C++ which is extremly fast (thank to mmap), but without |
27 |
support for properties (which this module tries to fix). |
28 |
|
29 |
Implementation of this module is crafted after L<Plucene::Simple> and it |
30 |
should be easy to replace Plucene with this module for increased |
31 |
performance. However, this module is not plug-in replacement. |
32 |
|
33 |
=head1 METHODS |
34 |
|
35 |
=head2 new |
36 |
|
37 |
Create new indexing object. |
38 |
|
39 |
my $i = SWISH::PlusPlus->new( |
40 |
index_dir => '/path/to/index', |
41 |
index => 'index++', |
42 |
search => 'search++', |
43 |
debug => 1, |
44 |
meta_in_body => 1, |
45 |
use_stopwords => 1, |
46 |
); |
47 |
|
48 |
Options to new are following: |
49 |
|
50 |
=over 5 |
51 |
|
52 |
=item C<index_dir> |
53 |
|
54 |
Path to directory in which index will be created. |
55 |
|
56 |
=item C<index> |
57 |
|
58 |
Full or partial path to SWISH++ index executable. By default, it's B<index> |
59 |
for self-compiled version. If you use Debian GNU/Linux package specify |
60 |
B<index++>. See C<Debian>. |
61 |
|
62 |
=item C<search> |
63 |
|
64 |
Full or partial path to SWISH++ search executable. By default, it's B<search>. |
65 |
|
66 |
=item C<debug> |
67 |
|
68 |
This option (off by default) will produce a lot of debugging output on |
69 |
C<STDERR> prefixed by C<##>. |
70 |
|
71 |
=item C<meta_in_body> |
72 |
|
73 |
This option (off by default) enables to search content of meta fields |
74 |
without specifing them (like they are in body of document). This will |
75 |
somewhat increate index size. |
76 |
|
77 |
=item C<use_stopwords> |
78 |
|
79 |
Use built-in SWISH++ stop words. By default, they are disabled. |
80 |
|
81 |
=back |
82 |
|
83 |
=cut |
84 |
|
85 |
sub new { |
86 |
my $class = shift; |
87 |
my $self = {@_}; |
88 |
bless($self, $class); |
89 |
|
90 |
foreach (qw(index_dir)) { |
91 |
croak "need $_" unless $self->{$_}; |
92 |
} |
93 |
|
94 |
my $index_dir = $self->{'index_dir'}; |
95 |
|
96 |
my $cwd; |
97 |
chomp($cwd = `pwd`); |
98 |
$self->{'cwd'} = $cwd || carp "can't get cwd!"; |
99 |
|
100 |
if ($index_dir !~ m#^/#) { |
101 |
$index_dir = "$cwd/$index_dir"; |
102 |
print STDERR "## full path to index_dir: $index_dir\n" if ($self->{'debug'}); |
103 |
$self->{'index_dir'} = $index_dir; |
104 |
} |
105 |
|
106 |
if (! -e $index_dir) { |
107 |
mkdir $index_dir || confess "can't create index ",$self->{'index'},": $!"; |
108 |
} |
109 |
|
110 |
# default executables |
111 |
$self->{'index'} ||= 'index'; |
112 |
$self->{'search'} ||= 'search'; |
113 |
|
114 |
print STDERR "## new index_dir: ",$index_dir," index: ",$self->{'index'}, " search: ",$self->{'search'},"\n" if ($self->{'debug'}); |
115 |
|
116 |
$self ? return $self : return undef; |
117 |
} |
118 |
|
119 |
|
120 |
=head2 check_bin |
121 |
|
122 |
Check if swish++ binaries specified in L<new> are available and verify |
123 |
version signature. |
124 |
|
125 |
if ($i->check_bin) { |
126 |
print "swish++ binaries found\n"; |
127 |
}; |
128 |
|
129 |
It will also setup property |
130 |
|
131 |
$i->{'version'} |
132 |
|
133 |
which you can examine to see version. |
134 |
|
135 |
=cut |
136 |
|
137 |
sub check_bin { |
138 |
my $self = shift; |
139 |
|
140 |
my $i = `$self->{'index'} -V 2>&1` || confess "can't find '",$self->{'index'},"' binary"; |
141 |
my $s = `$self->{'search'} -V 2>&1` || confess "can't find '",$self->{'search'},"' binary"; |
142 |
|
143 |
chomp $i; |
144 |
chomp $s; |
145 |
|
146 |
confess $self->{'index'}," binary is not SWISH++" unless ($i =~ m/^SWISH\+\+/); |
147 |
confess $self->{'search'}," binary is not SWISH++" unless ($s =~ m/^SWISH\+\+/); |
148 |
|
149 |
if ($i eq $s) { |
150 |
$i =~ s/^SWISH\+\+\s+// || confess "can't strip SWISH++ from version"; |
151 |
$self->{'version'} = $i; |
152 |
return 1; |
153 |
} else { |
154 |
carp "version difference: index is $i while search is $s"; |
155 |
return; |
156 |
} |
157 |
|
158 |
} |
159 |
|
160 |
=head2 index_document |
161 |
|
162 |
Quick way to add simple data to index. |
163 |
|
164 |
$i->index_document($key, $data); |
165 |
$i->index_document( 42 => 'meaning of life' ); |
166 |
|
167 |
=cut |
168 |
|
169 |
sub index_document { |
170 |
my $self = shift; |
171 |
|
172 |
my %doc = @_; |
173 |
|
174 |
foreach my $id (keys %doc) { |
175 |
$self->_create_doc( |
176 |
path => $id, |
177 |
body => $doc{$id}, |
178 |
); |
179 |
} |
180 |
|
181 |
return 1; |
182 |
} |
183 |
|
184 |
=head2 add |
185 |
|
186 |
Add document with metadata to index. |
187 |
|
188 |
$i->add( |
189 |
path => 'path/to/document', |
190 |
title => 'this is result title', |
191 |
meta => { |
192 |
description => 'this is description meta tag', |
193 |
date => '2004-11-04', |
194 |
author => 'Dobrica Pavlinusic', |
195 |
} |
196 |
body => 'this is text without meta data', |
197 |
); |
198 |
|
199 |
This is thin wrapper round L<_create_doc>. |
200 |
|
201 |
=cut |
202 |
|
203 |
sub add { |
204 |
my $self = shift; |
205 |
|
206 |
$self->_create_doc(@_); |
207 |
|
208 |
return 1; |
209 |
} |
210 |
=head2 search |
211 |
|
212 |
Search your index. |
213 |
|
214 |
my @results = $i->search("swhish query"); |
215 |
|
216 |
Returns array with result IDs. |
217 |
|
218 |
=cut |
219 |
|
220 |
sub search { |
221 |
my $self = shift; |
222 |
|
223 |
my $query = shift || return; |
224 |
|
225 |
$self->finish_update; |
226 |
$self->_tie_meta_db(DB_RDONLY); |
227 |
|
228 |
my @results; |
229 |
|
230 |
# escape double quotes in query for shell |
231 |
$query =~ s/"/\\"/g; |
232 |
|
233 |
my $open_cmd = $self->{'search'} . |
234 |
' -i ' . $self->{'index_dir'}.'/index' . |
235 |
' "' . $query . '"'. |
236 |
' |'; |
237 |
print STDERR "## search: $open_cmd\n" if ($self->{'debug'}); |
238 |
|
239 |
open(SEARCH, $open_cmd) || confess "can't start $open_cmd: $!"; |
240 |
my $l; |
241 |
while($l = <SEARCH>) { |
242 |
next if ($l =~ /^#/); |
243 |
chomp($l); |
244 |
print STDERR "## $l\n" if ($self->{'debug'}); |
245 |
my ($rank,$path,$size,$title) = split(/ /,$l,4); |
246 |
$path =~ s#^\./##; # strip from path |
247 |
push @results, { |
248 |
rank => $rank, |
249 |
path => $path, |
250 |
size => $size, |
251 |
title => $title, |
252 |
} |
253 |
} |
254 |
|
255 |
close(SEARCH) || confess "can't close search"; |
256 |
|
257 |
#print STDERR "## results: ",Dump(@results),"\n" if ($self->{'debug'}); |
258 |
|
259 |
return @results; |
260 |
} |
261 |
|
262 |
=head2 property |
263 |
|
264 |
Return stored meta property from result or result path. |
265 |
|
266 |
print $i->property('path', 'title'); |
267 |
print $i->property($res->{'path'}, 'title'); |
268 |
|
269 |
=cut |
270 |
|
271 |
sub property { |
272 |
my $self = shift; |
273 |
|
274 |
my ($path,$meta) = @_; |
275 |
|
276 |
if ($path =~ m/^HASH/) { |
277 |
$path = $path->{'path'} || confess "can't find path in input data"; |
278 |
} |
279 |
|
280 |
my $val = $self->{'meta_db'}->{"$path-$meta"}; |
281 |
|
282 |
print STDERR "## property $path-$meta: ",($val || 'undef'),"\n" if ($self->{'debug'}); |
283 |
return $val; |
284 |
} |
285 |
|
286 |
=head2 finish_update |
287 |
|
288 |
This method will close index. |
289 |
|
290 |
$i->finish_update; |
291 |
|
292 |
It will be called on DESTROY when $i goes out of scope. |
293 |
|
294 |
=cut |
295 |
|
296 |
sub finish_update { |
297 |
my $self = shift; |
298 |
|
299 |
print STDERR "## finish_update\n" if ($self->{'debug'}); |
300 |
|
301 |
$self->_close_index && $self->_untie_meta_db; |
302 |
} |
303 |
|
304 |
sub DESTROY { |
305 |
my $self = shift; |
306 |
$self->finish_update; |
307 |
} |
308 |
|
309 |
=head1 PRIVATE METHODS |
310 |
|
311 |
Private methods implement internals for creating temporary file needed for |
312 |
swish++. You should have no need to call them directly, and they are here |
313 |
just to have documentation. |
314 |
|
315 |
=head2 _init_indexer |
316 |
|
317 |
Create temporary directory in which files for indexing will be created and |
318 |
start index process. |
319 |
|
320 |
my $i->_init_indexer || die "can't start indexer"; |
321 |
|
322 |
It will also create empty file C<_stopwords_> to disable stop words. |
323 |
|
324 |
=cut |
325 |
|
326 |
sub _init_indexer { |
327 |
my $self = shift; |
328 |
|
329 |
return if ($self->{'_index_fh'}); |
330 |
|
331 |
my $tmp_dir = tempdir( CLEANUP => 1 ) || confess "can't create temporary directory: $!"; |
332 |
$self->{'tmp_dir'} = $tmp_dir; |
333 |
|
334 |
chdir $tmp_dir || confess "can't chdir to ".$tmp_dir.": $!"; |
335 |
|
336 |
print STDERR "## tmp_dir: $tmp_dir" if ($self->{'debug'}); |
337 |
|
338 |
my $opt = "-v " . ($self->{'debug'} || '0'); |
339 |
|
340 |
unless ($self->{'use_stopwrods'}) { |
341 |
open(STOP, '>', "_stopwords_") || carp "can't create empty stopword file, skipping\n"; |
342 |
print STOP " "; |
343 |
close(STOP); |
344 |
$opt .= " -s _stopwords_"; |
345 |
} |
346 |
|
347 |
my $index_dir = $self->{'index_dir'} || confess "no index_dir?"; |
348 |
|
349 |
my $open_cmd = '| '.$self->{'index'}.' '.$opt.' -e "html:*" -i '.$index_dir.'/index -'; |
350 |
|
351 |
print STDERR "## init_indexer: $open_cmd\n" if ($self->{'debug'}); |
352 |
|
353 |
open($self->{'_index_fh'}, $open_cmd) || confess "can't start index with $open_cmd: $!"; |
354 |
|
355 |
chdir $self->{'cwd'} || confess "can't chdir to ".$self->{'cwd'}.": $!"; |
356 |
|
357 |
$self->_tie_meta_db(DB_CREATE); |
358 |
|
359 |
return $self->{'_index_fh'}; |
360 |
} |
361 |
|
362 |
=head2 _tie_meta_db |
363 |
|
364 |
Open BerkeleyDB database with meta properties. |
365 |
|
366 |
$i->_tie_meta_db(DB_CREATE); |
367 |
$i->_tie_meta_db(DB_RDONLY); |
368 |
|
369 |
} |
370 |
|
371 |
=cut |
372 |
|
373 |
sub _tie_meta_db { |
374 |
my $self = shift; |
375 |
|
376 |
my $flags = shift || confess "need DB_CREATE or DB_RDONLY"; |
377 |
|
378 |
return if ($self->{'_meta_db_flags'} && $self->{'_meta_db_flags'} == $flags); |
379 |
|
380 |
print STDERR "## _tie_meta_db($flags)\n" if ($self->{'debug'}); |
381 |
|
382 |
$self->_untie_meta_db; |
383 |
$self->{'_meta_db_flags'} = $flags; |
384 |
|
385 |
my $file = $self->{'index_dir'}.'/meta.db'; |
386 |
|
387 |
tie %{$self->{'meta_db'}}, "BerkeleyDB::Hash", |
388 |
-Filename => $file, |
389 |
-Flags => $flags |
390 |
or confess "cannot open $file: $! $BerkeleyDB::Error\n" ; |
391 |
|
392 |
return 1; |
393 |
} |
394 |
|
395 |
=head2 _untie_meta_db |
396 |
|
397 |
Close BerkeleyDB database with meta properties. |
398 |
|
399 |
$i->_untie_meta_db |
400 |
|
401 |
=cut |
402 |
|
403 |
sub _untie_meta_db { |
404 |
my $self = shift; |
405 |
|
406 |
return unless ($self->{'meta_db'}); |
407 |
|
408 |
print STDERR "## _untie_meta_db\n" if ($self->{'debug'}); |
409 |
untie %{$self->{'meta_db'}} || confess "can't untie!"; |
410 |
undef $self->{'meta_db'}; |
411 |
undef $self->{'_meta_db_flags'}; |
412 |
|
413 |
return 1; |
414 |
} |
415 |
|
416 |
=head2 _create_doc |
417 |
|
418 |
Create temporary file and pass it's name to swish++ |
419 |
|
420 |
$i->_create_doc( |
421 |
path => 'path/to/store/in/index', |
422 |
title => 'this is title in results', |
423 |
body => 'data to story in body tag', |
424 |
meta => { |
425 |
'meta name' => 'data for this meta', |
426 |
'another' => 'again more data', |
427 |
} |
428 |
); |
429 |
|
430 |
To delete document, just omit body and meta data. |
431 |
|
432 |
=cut |
433 |
|
434 |
sub _create_doc { |
435 |
my $self = shift; |
436 |
|
437 |
my $arg = {@_}; |
438 |
|
439 |
# open indexer if needed |
440 |
$self->_init_indexer; |
441 |
|
442 |
my $path = $self->{'tmp_dir'} || confess "no tmp_dir?"; |
443 |
my $id = $arg->{'path'} || confess "no path?"; |
444 |
$path .= "/$id"; |
445 |
|
446 |
print STDERR "## _create_doc: $path\n" if ($self->{'debug'}); |
447 |
|
448 |
open(TMP, '>', $path) || die "can't create temp file $path: $!"; |
449 |
|
450 |
print TMP '<html><head>'; |
451 |
|
452 |
$arg->{'body'} ||= ''; |
453 |
|
454 |
if ($arg->{'meta'}) { |
455 |
foreach my $name (keys %{$arg->{'meta'}}) { |
456 |
my $content = $arg->{'meta'}->{$name}; |
457 |
print TMP qq{<meta name="$name" content="$content">}; |
458 |
$arg->{'body'} .= " $content" if ($self->{'meta_in_body'}); |
459 |
$self->{'meta_db'}->{"$id-$name"} = $content; |
460 |
} |
461 |
} |
462 |
|
463 |
my $title = $arg->{'title'}; |
464 |
if (defined($title)) { |
465 |
print TMP "<title>$title</title>"; |
466 |
$arg->{'body'} .= " $title" if ($self->{'meta_in_body'}); |
467 |
$self->{'meta_db'}->{"$id-title"} = $title; |
468 |
} |
469 |
|
470 |
print TMP '</head><body>' . $arg->{'body'} . '</body></html>'; |
471 |
|
472 |
close(TMP) || confess "can't close tmp file ".$arg->{'path'}.": $!"; |
473 |
|
474 |
print { $self->{'_index_fh'} } "$id\n"; |
475 |
} |
476 |
|
477 |
=head2 _close_index |
478 |
|
479 |
Close index after indexing. |
480 |
|
481 |
$i->_close_index; |
482 |
|
483 |
You have to close index before searching. |
484 |
|
485 |
=cut |
486 |
|
487 |
sub _close_index { |
488 |
my $self = shift; |
489 |
|
490 |
return unless ($self->{'_index_fh'}); |
491 |
|
492 |
print STDERR "## close index\n" if ($self->{'debug'}); |
493 |
|
494 |
close($self->{'_index_fh'}) || confess "can't close index: $!"; |
495 |
undef $self->{'_index_fh'}; |
496 |
|
497 |
return 1; |
498 |
} |
499 |
|
500 |
1; |
501 |
__END__ |
502 |
|
503 |
=head2 EXPORT |
504 |
|
505 |
None by default. |
506 |
|
507 |
=head1 RELATED |
508 |
|
509 |
=head2 Debian |
510 |
|
511 |
Debian version of swish++ is often old (version 5 at moment of this writing |
512 |
while version 6 is available in source code), so this module by default |
513 |
uses executable names B<index> and B<search> for self-compiled version |
514 |
instead of one from Debian package. See L<new> how to specify Debian |
515 |
default binaries B<index++> and B<search++>. |
516 |
|
517 |
=head2 SWISH++ |
518 |
|
519 |
Aside from very good rewrite in C++, SWISH++ is fatster because it has |
520 |
claver heuristics about which data in input files are words to index and |
521 |
which are not. It's based on English language and might be best choice if |
522 |
you plan to install large amount of long text documents. |
523 |
|
524 |
However, if you plan to index all data from structured storage (e.g. RDBMS) |
525 |
you might want B<all> words from data to end up in index as opposed to just |
526 |
those which look like English words. This is especially important if you |
527 |
don't plan to index English texts with this module. |
528 |
|
529 |
With distribution build versions of SWISH++ you might have problems with |
530 |
disepearing words. To overcome this problem, you will have to compile and |
531 |
configure SWISH++ yourself (because language characteristics are |
532 |
compilation-time option). |
533 |
|
534 |
Compilation of SWISH++ is easy process well described on project's web |
535 |
pages. To see my very relaxed sample configuration take a look at C<swish++> |
536 |
directory included in distribution. |
537 |
|
538 |
=head2 SWISH++ config |
539 |
|
540 |
C<config.h> located in C<swish++> directory of this distribution is relaxed |
541 |
SWISH++ configuration that will index all words passed to it. This |
542 |
configuration is needed for B<date test> because default configuration |
543 |
doesn't recognize 2004-12-05 as date. Have in mind that your index size |
544 |
might explode. |
545 |
|
546 |
=head1 SEE ALSO |
547 |
|
548 |
C<swish++> web site L<http://homepage.mac.com/pauljlucas/software/swish/> |
549 |
|
550 |
=head1 AUTHOR |
551 |
|
552 |
Dobrica Pavlinusic, E<lt>dpavlin@rot13.orgE<gt> |
553 |
|
554 |
=head1 COPYRIGHT AND LICENSE |
555 |
|
556 |
Copyright (C) 2004 by Dobrica Pavlinusic |
557 |
|
558 |
This library is free software; you can redistribute it and/or modify |
559 |
it under the same terms as Perl itself, either Perl version 5.8.4 or, |
560 |
at your option, any later version of Perl 5 you may have available. |
561 |
|
562 |
|
563 |
=cut |