1 |
dpavlin |
1 |
package SWISH::PlusPlus; |
2 |
|
|
|
3 |
|
|
use 5.008004; |
4 |
|
|
use strict; |
5 |
|
|
use warnings; |
6 |
|
|
|
7 |
dpavlin |
3 |
our $VERSION = '0.02'; |
8 |
dpavlin |
1 |
|
9 |
|
|
use Carp; |
10 |
dpavlin |
4 |
use File::Temp qw/ tempdir /; |
11 |
dpavlin |
1 |
|
12 |
|
|
=head1 NAME |
13 |
|
|
|
14 |
|
|
SWISH::PlusPlus - Perl extension SWISH++ |
15 |
|
|
|
16 |
|
|
=head1 SYNOPSIS |
17 |
|
|
|
18 |
|
|
use SWISH::PlusPlus; |
19 |
|
|
blah blah blah |
20 |
|
|
|
21 |
|
|
=head1 DESCRIPTION |
22 |
|
|
|
23 |
|
|
This is perl module to use SWISH++ indexer by Paul J. Lucas. SWISH++ is |
24 |
dpavlin |
3 |
rewrite of swish-e in C++ which is extremly fast (thank to mmap), but without |
25 |
|
|
support for properties (which this module tries to fix). |
26 |
dpavlin |
1 |
|
27 |
dpavlin |
3 |
Implementation of this module is crafted after L<Plucene::Simple> and it |
28 |
|
|
should be easy to replace Plucene with this module for increased |
29 |
|
|
performance. However, this module is not plug-in replacement. |
30 |
|
|
|
31 |
dpavlin |
1 |
=head1 METHODS |
32 |
|
|
|
33 |
|
|
=head2 open |
34 |
|
|
|
35 |
|
|
Create new indexing object. |
36 |
|
|
|
37 |
dpavlin |
3 |
my $i = SWISH::PlusPlus->open( |
38 |
|
|
index_dir => '/path/to/index', |
39 |
|
|
index => 'index++', |
40 |
|
|
search => 'search++', |
41 |
dpavlin |
1 |
); |
42 |
|
|
|
43 |
|
|
Options to open are following: |
44 |
|
|
|
45 |
|
|
=over 5 |
46 |
|
|
|
47 |
dpavlin |
3 |
=item C<index_dir> |
48 |
|
|
|
49 |
|
|
Path to directory in which index will be created. |
50 |
|
|
|
51 |
dpavlin |
1 |
=item C<index> |
52 |
|
|
|
53 |
dpavlin |
3 |
Full or partial path to SWISH++ index executable. By default, it's B<index> |
54 |
|
|
for self-compiled version. If you use Debian GNU/Linux package specify |
55 |
|
|
B<index++>. See C<Debian>. |
56 |
dpavlin |
1 |
|
57 |
dpavlin |
3 |
=item C<search> |
58 |
|
|
|
59 |
|
|
Full or partial path to SWISH++ search executable. By default, it's B<search>. |
60 |
|
|
|
61 |
dpavlin |
1 |
=back |
62 |
|
|
|
63 |
|
|
=cut |
64 |
|
|
|
65 |
dpavlin |
3 |
sub open { |
66 |
dpavlin |
1 |
my $class = shift; |
67 |
|
|
my $self = {@_}; |
68 |
|
|
bless($self, $class); |
69 |
|
|
|
70 |
dpavlin |
3 |
foreach (qw(index_dir)) { |
71 |
dpavlin |
1 |
croak "need $_" unless $self->{$_}; |
72 |
|
|
} |
73 |
|
|
|
74 |
dpavlin |
3 |
if (! -e $self->{'index_dir'}) { |
75 |
|
|
mkdir $self->{'index_dir'} || confess "can't create index ",$self->{'index'},": $!"; |
76 |
dpavlin |
1 |
} |
77 |
|
|
|
78 |
dpavlin |
3 |
# default executables |
79 |
|
|
$self->{'index'} ||= 'index'; |
80 |
|
|
$self->{'search'} ||= 'search'; |
81 |
|
|
|
82 |
dpavlin |
1 |
$self ? return $self : return undef; |
83 |
|
|
} |
84 |
|
|
|
85 |
|
|
|
86 |
dpavlin |
3 |
=head2 check_bin |
87 |
|
|
|
88 |
|
|
Check if swish++ binaries specified in L<open> are available and verify |
89 |
|
|
version signature. |
90 |
|
|
|
91 |
|
|
if ($i->check_bin) { |
92 |
|
|
print "swish++ binaries found\n"; |
93 |
|
|
}; |
94 |
|
|
|
95 |
|
|
It will also setup property |
96 |
|
|
|
97 |
|
|
$i->{'version'} |
98 |
|
|
|
99 |
|
|
which you can examine to see version. |
100 |
|
|
|
101 |
|
|
=cut |
102 |
|
|
|
103 |
|
|
sub check_bin { |
104 |
|
|
my $self = shift; |
105 |
|
|
|
106 |
|
|
my $i = `$self->{'index'} -V 2>&1` || confess "can't find '",$self->{'index'},"' binary"; |
107 |
|
|
my $s = `$self->{'search'} -V 2>&1` || confess "can't find '",$self->{'search'},"' binary"; |
108 |
|
|
|
109 |
|
|
chomp $i; |
110 |
|
|
chomp $s; |
111 |
|
|
|
112 |
|
|
confess $self->{'index'}," binary is not SWISH++" unless ($i =~ m/^SWISH\+\+/); |
113 |
|
|
confess $self->{'search'}," binary is not SWISH++" unless ($s =~ m/^SWISH\+\+/); |
114 |
|
|
|
115 |
|
|
if ($i eq $s) { |
116 |
|
|
$self->{'version'} = $i; |
117 |
|
|
return 1; |
118 |
|
|
} else { |
119 |
|
|
carp "version difference: index is $i while search is $s"; |
120 |
|
|
return; |
121 |
|
|
} |
122 |
|
|
|
123 |
|
|
} |
124 |
|
|
|
125 |
dpavlin |
4 |
=head2 index_document |
126 |
|
|
|
127 |
|
|
Quick way to add simple data to index. |
128 |
|
|
|
129 |
|
|
$i->index_document($key, $data); |
130 |
|
|
$i->index_document( 42 => 'meaning of life' ); |
131 |
|
|
|
132 |
|
|
=cut |
133 |
|
|
|
134 |
|
|
sub index_document { |
135 |
|
|
my $self = shift; |
136 |
|
|
|
137 |
|
|
my %doc = @_; |
138 |
|
|
|
139 |
|
|
foreach my $id (keys %doc) { |
140 |
|
|
$self->_create_doc( |
141 |
|
|
path => $id, |
142 |
|
|
body => $doc{$id}, |
143 |
|
|
); |
144 |
|
|
} |
145 |
|
|
|
146 |
|
|
return 1; |
147 |
|
|
} |
148 |
|
|
|
149 |
|
|
=head1 PRIVATE METHODS |
150 |
|
|
|
151 |
|
|
Private methods implement internals for creating temporary file needed for |
152 |
|
|
swish++. You should have no need to call them directly, and they are here |
153 |
|
|
just to have documentation. |
154 |
|
|
|
155 |
|
|
=head2 _init_index |
156 |
|
|
|
157 |
|
|
Create temporary directory in which files for indexing will be created and |
158 |
|
|
start index process. |
159 |
|
|
|
160 |
|
|
my $i->_init_index || die "can't start indexer"; |
161 |
|
|
|
162 |
|
|
=cut |
163 |
|
|
|
164 |
|
|
sub _init_index { |
165 |
|
|
my $self = shift; |
166 |
|
|
|
167 |
|
|
$self->{'tmp_dir'} = tempdir( CLEANUP => 1 ) || confess "can't create temporary directory: $!"; |
168 |
|
|
|
169 |
|
|
my $opt = "-v 4"; |
170 |
|
|
|
171 |
|
|
my $open_cmd = '| index '.$opt.' -e "html:*" -i '.$self->{'index_dir'}.'/index -'; |
172 |
|
|
|
173 |
|
|
chdir $self->{'tmp_dir'} || confess "can't chdir to ".$self->{'tmp_dir'}.": $!"; |
174 |
|
|
|
175 |
|
|
CORE::open($self->{'index_fh'}, $open_cmd) || confess "can't start index with $open_cmd: $!"; |
176 |
|
|
|
177 |
|
|
return $self->{'index_fh'}; |
178 |
|
|
} |
179 |
|
|
|
180 |
|
|
=head2 _create_doc |
181 |
|
|
|
182 |
|
|
Create temporary file and pass it's name to swish++ |
183 |
|
|
|
184 |
|
|
$i->_create_doc( |
185 |
|
|
path => 'path/to/store/in/index', |
186 |
|
|
body => 'data to story in body tag', |
187 |
|
|
meta => { |
188 |
|
|
'meta name' => 'data for this meta', |
189 |
|
|
'another' => 'again more data', |
190 |
|
|
} |
191 |
|
|
); |
192 |
|
|
|
193 |
|
|
=cut |
194 |
|
|
|
195 |
|
|
sub _create_doc { |
196 |
|
|
my $self = shift; |
197 |
|
|
|
198 |
|
|
my $arg = {@_}; |
199 |
|
|
|
200 |
|
|
# open indexer if needed |
201 |
|
|
$self->{'index_fh'} ||= $self->_init_index; |
202 |
|
|
|
203 |
|
|
my $path = $self->{'tmp_dir'} || confess "no tmp_dir?"; |
204 |
|
|
|
205 |
|
|
CORE::open(TMP, '>', $arg->{'path'}) || die "can't create temp file ".$arg->{'path'}.": $!"; |
206 |
|
|
|
207 |
|
|
print TMP '<html>'; |
208 |
|
|
|
209 |
|
|
if ($arg->{'meta'}) { |
210 |
|
|
confess "not yet implemented"; |
211 |
|
|
} |
212 |
|
|
|
213 |
|
|
print TMP '<body>' . ($arg->{'body'} || '') . '</body></html>'; |
214 |
|
|
|
215 |
|
|
close(TMP) || confess "can't close tmp file ".$arg->{'path'}.": $!"; |
216 |
|
|
|
217 |
|
|
print { $self->{'index_fh'} } $arg->{'path'}."\n"; |
218 |
|
|
} |
219 |
|
|
|
220 |
dpavlin |
1 |
1; |
221 |
|
|
__END__ |
222 |
|
|
|
223 |
|
|
=head2 EXPORT |
224 |
|
|
|
225 |
|
|
None by default. |
226 |
|
|
|
227 |
dpavlin |
3 |
=head1 RELATED |
228 |
|
|
|
229 |
|
|
=head2 Debian |
230 |
|
|
|
231 |
|
|
Debian version of swish++ is often old (version 5 at moment of this writing |
232 |
|
|
while version 6 is available in source code), so this module by default |
233 |
|
|
uses executable names B<index> and B<search> for self-compiled version |
234 |
|
|
instead of one from Debian package. See L<open> how to specify Debian |
235 |
|
|
default binaries B<index++> and B<search++>. |
236 |
|
|
|
237 |
dpavlin |
5 |
=head2 SWISH++ |
238 |
dpavlin |
1 |
|
239 |
dpavlin |
5 |
Aside from very good rewrite in C++, SWISH++ is fatster because it has |
240 |
|
|
claver heuristics about which data in input files are words to index and |
241 |
|
|
which are not. It's based on English language and might be best choice if |
242 |
|
|
you plan to install large amount of long text documents. |
243 |
dpavlin |
1 |
|
244 |
dpavlin |
5 |
However, if you plan to index all data from structured storage (e.g. RDBMS) |
245 |
|
|
you might want B<all> words from data to end up in index as opposed to just |
246 |
|
|
those which look like English words. This is especially important if you |
247 |
|
|
don't plan to index English texts with this module. |
248 |
dpavlin |
1 |
|
249 |
dpavlin |
5 |
With distribution build versions of SWISH++ you might have problems with |
250 |
|
|
disepearing words. To overcome this problem, you will have to compile and |
251 |
|
|
configure SWISH++ yourself (because language characteristics are |
252 |
|
|
compilation-time option). |
253 |
dpavlin |
1 |
|
254 |
dpavlin |
5 |
Compilation of SWISH++ is easy process well described on project's web |
255 |
|
|
pages. To see my very relaxed sample configuration take a look at C<swish++> |
256 |
|
|
directory included in distribution. |
257 |
|
|
|
258 |
|
|
=head1 SEE ALSO |
259 |
|
|
|
260 |
|
|
C<swish++> web site L<http://homepage.mac.com/pauljlucas/software/swish/> |
261 |
|
|
|
262 |
dpavlin |
1 |
=head1 AUTHOR |
263 |
|
|
|
264 |
dpavlin |
5 |
Dobrica Pavlinusic, E<lt>dpavlin@rot13.orgE<gt> |
265 |
dpavlin |
1 |
|
266 |
|
|
=head1 COPYRIGHT AND LICENSE |
267 |
|
|
|
268 |
|
|
Copyright (C) 2004 by Dobrica Pavlinusic |
269 |
|
|
|
270 |
|
|
This library is free software; you can redistribute it and/or modify |
271 |
|
|
it under the same terms as Perl itself, either Perl version 5.8.4 or, |
272 |
|
|
at your option, any later version of Perl 5 you may have available. |
273 |
|
|
|
274 |
|
|
|
275 |
|
|
=cut |