4 |
use strict; |
use strict; |
5 |
use warnings; |
use warnings; |
6 |
|
|
7 |
our $VERSION = '0.04'; |
our $VERSION = '0.06'; |
8 |
|
|
9 |
use Carp; |
use Carp; |
10 |
use File::Temp qw/ tempdir /; |
use File::Temp qw/ tempdir /; |
90 |
croak "need $_" unless $self->{$_}; |
croak "need $_" unless $self->{$_}; |
91 |
} |
} |
92 |
|
|
93 |
if (! -e $self->{'index_dir'}) { |
my $index_dir = $self->{'index_dir'}; |
94 |
mkdir $self->{'index_dir'} || confess "can't create index ",$self->{'index'},": $!"; |
|
95 |
|
if ($index_dir !~ m#^/#) { |
96 |
|
chomp(my $cwd = `pwd`); |
97 |
|
$index_dir = "$cwd/$index_dir"; |
98 |
|
print STDERR "## full path to index_dir: $index_dir\n" if ($self->{'debug'}); |
99 |
|
$self->{'index_dir'} = $index_dir; |
100 |
|
} |
101 |
|
|
102 |
|
if (! -e $index_dir) { |
103 |
|
mkdir $index_dir || confess "can't create index ",$self->{'index'},": $!"; |
104 |
} |
} |
105 |
|
|
106 |
# default executables |
# default executables |
107 |
$self->{'index'} ||= 'index'; |
$self->{'index'} ||= 'index'; |
108 |
$self->{'search'} ||= 'search'; |
$self->{'search'} ||= 'search'; |
109 |
|
|
110 |
print STDERR "## new index_dir: ",$self->{'index_dir'}," index: ",$self->{'index'}, " search: ",$self->{'search'},"\n" if ($self->{'debug'}); |
print STDERR "## new index_dir: ",$index_dir," index: ",$self->{'index'}, " search: ",$self->{'search'},"\n" if ($self->{'debug'}); |
111 |
|
|
112 |
$self ? return $self : return undef; |
$self ? return $self : return undef; |
113 |
} |
} |
248 |
return @results; |
return @results; |
249 |
} |
} |
250 |
|
|
251 |
|
=head2 finish_update |
252 |
|
|
253 |
|
This method will close index. |
254 |
|
|
255 |
|
$i->finish_update; |
256 |
|
|
257 |
|
It will be called on DESTROY when $i goes out of scope. |
258 |
|
|
259 |
|
=cut |
260 |
|
|
261 |
|
sub finish_update { |
262 |
|
my $self = shift; |
263 |
|
|
264 |
|
$self->_close_index; |
265 |
|
} |
266 |
|
|
267 |
|
sub DESTROY { |
268 |
|
my $self = shift; |
269 |
|
$self->finish_update; |
270 |
|
} |
271 |
|
|
272 |
=head1 PRIVATE METHODS |
=head1 PRIVATE METHODS |
273 |
|
|
274 |
Private methods implement internals for creating temporary file needed for |
Private methods implement internals for creating temporary file needed for |
293 |
|
|
294 |
chdir $self->{'tmp_dir'} || confess "can't chdir to ".$self->{'tmp_dir'}.": $!"; |
chdir $self->{'tmp_dir'} || confess "can't chdir to ".$self->{'tmp_dir'}.": $!"; |
295 |
|
|
296 |
my $opt = "-v 4"; |
my $opt = "-v " . ($self->{'debug'} || '0'); |
297 |
|
|
298 |
unless ($self->{'use_stopwrods'}) { |
unless ($self->{'use_stopwrods'}) { |
299 |
open(STOP, '>', "_stopwords_") || carp "can't create empty stopword file, skipping\n"; |
open(STOP, '>', "_stopwords_") || carp "can't create empty stopword file, skipping\n"; |
304 |
|
|
305 |
my $open_cmd = '| '.$self->{'index'}.' '.$opt.' -e "html:*" -i '.$self->{'index_dir'}.'/index -'; |
my $open_cmd = '| '.$self->{'index'}.' '.$opt.' -e "html:*" -i '.$self->{'index_dir'}.'/index -'; |
306 |
|
|
307 |
|
print STDERR "## open index $open_cmd\n" if ($self->{'index'}); |
308 |
|
|
309 |
open($self->{'index_fh'}, $open_cmd) || confess "can't start index with $open_cmd: $!"; |
open($self->{'index_fh'}, $open_cmd) || confess "can't start index with $open_cmd: $!"; |
310 |
|
|
347 |
$arg->{'body'} ||= ''; |
$arg->{'body'} ||= ''; |
348 |
|
|
349 |
if ($arg->{'meta'}) { |
if ($arg->{'meta'}) { |
350 |
confess "not yet implemented"; |
foreach my $name (keys %{$arg->{'meta'}}) { |
351 |
|
my $content = $arg->{'meta'}->{$name}; |
352 |
|
print TMP qq{<meta name="$name" content="$content">}; |
353 |
|
$arg->{'body'} .= " $content" if ($self->{'meta_in_body'}); |
354 |
|
} |
355 |
} |
} |
356 |
|
|
357 |
if (defined($arg->{'title'})) { |
if (defined($arg->{'title'})) { |
425 |
pages. To see my very relaxed sample configuration take a look at C<swish++> |
pages. To see my very relaxed sample configuration take a look at C<swish++> |
426 |
directory included in distribution. |
directory included in distribution. |
427 |
|
|
428 |
|
=head2 SWISH++ config |
429 |
|
|
430 |
|
C<config.h> located in C<swish++> directory of this distribution is relaxed |
431 |
|
SWISH++ configuration that will index all words passed to it. This |
432 |
|
configuration is needed for B<date test> because default configuration |
433 |
|
doesn't recognize 2004-12-05 as date. Have in mind that your index size |
434 |
|
might explode. |
435 |
|
|
436 |
=head1 SEE ALSO |
=head1 SEE ALSO |
437 |
|
|
438 |
C<swish++> web site L<http://homepage.mac.com/pauljlucas/software/swish/> |
C<swish++> web site L<http://homepage.mac.com/pauljlucas/software/swish/> |