9 |
use XML::Simple; |
use XML::Simple; |
10 |
use Template; |
use Template; |
11 |
use Log::Log4perl qw(get_logger :levels); |
use Log::Log4perl qw(get_logger :levels); |
12 |
|
use Time::HiRes qw(time); |
13 |
|
|
14 |
use Data::Dumper; |
use Data::Dumper; |
15 |
|
|
30 |
|
|
31 |
=head2 new |
=head2 new |
32 |
|
|
33 |
This will create new instance of WebPAC using configuration specified by C<config_file>. |
Create new instance of WebPAC using configuration specified by C<config_file>. |
34 |
|
|
35 |
my $webpac = new WebPAC( |
my $webpac = new WebPAC( |
36 |
config_file => 'name.conf', |
config_file => 'name.conf', |
37 |
[code_page => 'ISO-8859-2',] |
[code_page => 'ISO-8859-2',] |
38 |
|
[low_mem => 1,] |
39 |
); |
); |
40 |
|
|
41 |
Default C<code_page> is C<ISO-8859-2>. |
Default C<code_page> is C<ISO-8859-2>. |
42 |
|
|
43 |
It will also read configuration files |
Default is not to use C<low_mem> options (see L<MEMORY USAGE> below). |
44 |
|
|
45 |
|
This method will also read configuration files |
46 |
C<global.conf> (used by indexer and Web font-end) |
C<global.conf> (used by indexer and Web font-end) |
47 |
and configuration file specified by C<config_file> |
and configuration file specified by C<config_file> |
48 |
which describes databases to be indexed. |
which describes databases to be indexed. |
63 |
my $self = {@_}; |
my $self = {@_}; |
64 |
bless($self, $class); |
bless($self, $class); |
65 |
|
|
66 |
|
$self->{'start_t'} = time(); |
67 |
|
|
68 |
my $log_file = $self->{'log'} || "log.conf"; |
my $log_file = $self->{'log'} || "log.conf"; |
69 |
Log::Log4perl->init($log_file); |
Log::Log4perl->init($log_file); |
70 |
|
|
112 |
EVAL_PERL => 1, |
EVAL_PERL => 1, |
113 |
); |
); |
114 |
|
|
115 |
|
# running with low_mem flag? well, use DBM::Deep then. |
116 |
|
if ($self->{'low_mem'}) { |
117 |
|
$log->info("running with low_mem which impacts performance (<32 Mb memory usage)"); |
118 |
|
|
119 |
|
my $db_file = "data.db"; |
120 |
|
|
121 |
|
if (-e $db_file) { |
122 |
|
unlink $db_file or $log->logdie("can't remove '$db_file' from last run"); |
123 |
|
$log->debug("removed '$db_file' from last run"); |
124 |
|
} |
125 |
|
|
126 |
|
use DBM::Deep; |
127 |
|
|
128 |
|
my $db = new DBM::Deep $db_file; |
129 |
|
|
130 |
|
$log->logdie("DBM::Deep error: $!") unless ($db); |
131 |
|
|
132 |
|
if ($db->error()) { |
133 |
|
$log->logdie("can't open '$db_file' under low_mem: ",$db->error()); |
134 |
|
} else { |
135 |
|
$log->debug("using file '$db_file' for DBM::Deep"); |
136 |
|
} |
137 |
|
|
138 |
|
$self->{'db'} = $db; |
139 |
|
} |
140 |
|
|
141 |
return $self; |
return $self; |
142 |
} |
} |
143 |
|
|
148 |
$webpac->open_isis( |
$webpac->open_isis( |
149 |
filename => '/data/ISIS/ISIS', |
filename => '/data/ISIS/ISIS', |
150 |
code_page => '852', |
code_page => '852', |
151 |
limit_mfn => '500', |
limit_mfn => 500, |
152 |
|
start_mfn => 6000, |
153 |
lookup => [ ... ], |
lookup => [ ... ], |
154 |
); |
); |
155 |
|
|
156 |
By default, ISIS code page is assumed to be C<852>. |
By default, ISIS code page is assumed to be C<852>. |
157 |
|
|
158 |
|
If optional parametar C<start_mfn> is set, this will be first MFN to read |
159 |
|
from database (so you can skip beginning of your database if you need to). |
160 |
|
|
161 |
If optional parametar C<limit_mfn> is set, it will read just 500 records |
If optional parametar C<limit_mfn> is set, it will read just 500 records |
162 |
from database in example above. |
from database in example above. |
163 |
|
|
185 |
$log->logcroak("need filename") if (! $arg->{'filename'}); |
$log->logcroak("need filename") if (! $arg->{'filename'}); |
186 |
my $code_page = $arg->{'code_page'} || '852'; |
my $code_page = $arg->{'code_page'} || '852'; |
187 |
|
|
188 |
|
# store data in object |
189 |
|
$self->{'isis_filename'} = $arg->{'filename'}; |
190 |
|
$self->{'isis_code_page'} = $code_page; |
191 |
|
|
192 |
use OpenIsis; |
use OpenIsis; |
193 |
|
|
194 |
#$self->{'isis_code_page'} = $code_page; |
#$self->{'isis_code_page'} = $code_page; |
197 |
my $cp = Text::Iconv->new($code_page,$self->{'code_page'}); |
my $cp = Text::Iconv->new($code_page,$self->{'code_page'}); |
198 |
|
|
199 |
$log->info("reading ISIS database '",$arg->{'filename'},"'"); |
$log->info("reading ISIS database '",$arg->{'filename'},"'"); |
200 |
|
$log->debug("isis code page: $code_page"); |
201 |
|
|
202 |
my $isis_db = OpenIsis::open($arg->{'filename'}); |
my $isis_db = OpenIsis::open($arg->{'filename'}); |
203 |
|
|
204 |
my $maxmfn = OpenIsis::maxRowid( $isis_db ) || 1; |
my $maxmfn = OpenIsis::maxRowid( $isis_db ) || 1; |
205 |
|
my $startmfn = 1; |
206 |
|
|
207 |
|
if (my $s = $self->{'start_mfn'}) { |
208 |
|
$log->info("skipping to MFN $s"); |
209 |
|
$startmfn = $s; |
210 |
|
} |
211 |
|
|
212 |
$maxmfn = $self->{limit_mfn} if ($self->{limit_mfn}); |
$maxmfn = $startmfn + $self->{limit_mfn} if ($self->{limit_mfn}); |
213 |
|
|
214 |
$log->info("processing $maxmfn records..."); |
$log->info("processing ",($maxmfn-$startmfn)." records..."); |
215 |
|
|
216 |
# read database |
# read database |
217 |
for (my $mfn = 1; $mfn <= $maxmfn; $mfn++) { |
for (my $mfn = $startmfn; $mfn <= $maxmfn; $mfn++) { |
218 |
|
|
219 |
|
|
220 |
|
$log->debug("mfn: $mfn\n"); |
221 |
|
|
222 |
|
my $rec; |
223 |
|
|
224 |
# read record |
# read record |
225 |
my $row = OpenIsis::read( $isis_db, $mfn ); |
my $row = OpenIsis::read( $isis_db, $mfn ); |
238 |
$val = $l; |
$val = $l; |
239 |
} |
} |
240 |
|
|
241 |
push @{$self->{'data'}->{$mfn}->{$k}}, $val; |
push @{$rec->{$k}}, $val; |
242 |
} |
} |
243 |
} else { |
} else { |
244 |
push @{$self->{'data'}->{$mfn}->{'000'}}, $mfn; |
push @{$rec->{'000'}}, $mfn; |
245 |
} |
} |
246 |
|
|
247 |
} |
} |
248 |
|
|
249 |
|
$log->confess("record $mfn empty?") unless ($rec); |
250 |
|
|
251 |
|
# store |
252 |
|
if ($self->{'low_mem'}) { |
253 |
|
$self->{'db'}->put($mfn, $rec); |
254 |
|
} else { |
255 |
|
$self->{'data'}->{$mfn} = $rec; |
256 |
|
} |
257 |
|
|
258 |
# create lookup |
# create lookup |
|
my $rec = $self->{'data'}->{$mfn}; |
|
259 |
$self->create_lookup($rec, @{$arg->{'lookup'}}); |
$self->create_lookup($rec, @{$arg->{'lookup'}}); |
260 |
|
|
261 |
|
$self->progress_bar($mfn,$maxmfn); |
262 |
|
|
263 |
} |
} |
264 |
|
|
265 |
$self->{'current_mfn'} = 1; |
$self->{'current_mfn'} = $startmfn; |
266 |
|
$self->{'last_pcnt'} = 0; |
267 |
|
|
268 |
|
$log->debug("max mfn: $maxmfn"); |
269 |
|
|
270 |
# store max mfn and return it. |
# store max mfn and return it. |
271 |
return $self->{'max_mfn'} = $maxmfn; |
return $self->{'max_mfn'} = $maxmfn; |
293 |
return; |
return; |
294 |
} |
} |
295 |
|
|
296 |
return $self->{'data'}->{$mfn}; |
$self->progress_bar($mfn,$self->{'max_mfn'}); |
297 |
|
|
298 |
|
if ($self->{'low_mem'}) { |
299 |
|
return $self->{'db'}->get($mfn); |
300 |
|
} else { |
301 |
|
return $self->{'data'}->{$mfn}; |
302 |
|
} |
303 |
|
} |
304 |
|
|
305 |
|
=head2 progress_bar |
306 |
|
|
307 |
|
Draw progress bar on STDERR. |
308 |
|
|
309 |
|
$webpac->progress_bar($current, $max); |
310 |
|
|
311 |
|
=cut |
312 |
|
|
313 |
|
sub progress_bar { |
314 |
|
my $self = shift; |
315 |
|
|
316 |
|
my ($curr,$max) = @_; |
317 |
|
|
318 |
|
my $log = $self->_get_logger(); |
319 |
|
|
320 |
|
$log->logconfess("no current value!") if (! $curr); |
321 |
|
$log->logconfess("no maximum value!") if (! $max); |
322 |
|
|
323 |
|
if ($curr > $max) { |
324 |
|
$max = $curr; |
325 |
|
$log->debug("overflow to $curr"); |
326 |
|
} |
327 |
|
|
328 |
|
$self->{'last_pcnt'} ||= 1; |
329 |
|
|
330 |
|
my $p = int($curr * 100 / $max); |
331 |
|
|
332 |
|
# reset on re-run |
333 |
|
if ($p < $self->{'last_pcnt'}) { |
334 |
|
$self->{'last_pcnt'} = $p; |
335 |
|
$self->{'last_t'} = time(); |
336 |
|
$self->{'last_curr'} = 1; |
337 |
|
} |
338 |
|
|
339 |
|
if ($p != $self->{'last_pcnt'}) { |
340 |
|
|
341 |
|
my $last_curr = $self->{'last_curr'} || $curr; |
342 |
|
my $t = time(); |
343 |
|
my $rate = ($curr - $last_curr) / (($t - $self->{'last_t'} || 1)); |
344 |
|
my $eta = ($max-$curr) / ($rate || 1); |
345 |
|
printf STDERR ("%5d [%-38s] %-5d %0.1f/s %s\r",$curr,"=" x ($p/3)."$p%>", $max, $rate, $self->fmt_time($eta)); |
346 |
|
$self->{'last_pcnt'} = $p; |
347 |
|
$self->{'last_t'} = time(); |
348 |
|
$self->{'last_curr'} = $curr; |
349 |
|
} |
350 |
|
print STDERR "\n" if ($p == 100); |
351 |
|
} |
352 |
|
|
353 |
|
=head2 fmt_time |
354 |
|
|
355 |
|
Format time (in seconds) for display. |
356 |
|
|
357 |
|
print $webpac->fmt_time(time()); |
358 |
|
|
359 |
|
This method is called by L<progress_bar> to display remaining time. |
360 |
|
|
361 |
|
=cut |
362 |
|
|
363 |
|
sub fmt_time { |
364 |
|
my $self = shift; |
365 |
|
|
366 |
|
my $t = shift || 0; |
367 |
|
my $out = ""; |
368 |
|
|
369 |
|
my ($ss,$mm,$hh) = gmtime($t); |
370 |
|
$out .= "${hh}h" if ($hh); |
371 |
|
$out .= sprintf("%02d:%02d", $mm,$ss); |
372 |
|
$out .= " " if ($hh == 0); |
373 |
|
return $out; |
374 |
} |
} |
375 |
|
|
376 |
=head2 open_import_xml |
=head2 open_import_xml |
432 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
433 |
|
|
434 |
foreach my $i (@_) { |
foreach my $i (@_) { |
435 |
if ($i->{'eval'}) { |
$log->logconfess("need key") unless defined($i->{'key'}); |
436 |
my $eval = $self->fill_in($rec,$i->{'eval'}); |
$log->logconfess("need val") unless defined($i->{'val'}); |
437 |
my $key = $self->fill_in($rec,$i->{'key'}); |
|
438 |
my @val = $self->fill_in($rec,$i->{'val'}); |
if (defined($i->{'eval'})) { |
439 |
if ($key && @val && eval $eval) { |
# eval first, so we can skip fill_in for key and val |
440 |
|
my $eval = $self->fill_in($rec,$i->{'eval'}) || next; |
441 |
|
if ($self->_eval($eval)) { |
442 |
|
my $key = $self->fill_in($rec,$i->{'key'}) || next; |
443 |
|
my @val = $self->fill_in($rec,$i->{'val'}) || next; |
444 |
$log->debug("stored $key = ",sub { join(" | ",@val) }); |
$log->debug("stored $key = ",sub { join(" | ",@val) }); |
445 |
push @{$self->{'lookup'}->{$key}}, @val; |
push @{$self->{'lookup'}->{$key}}, @val; |
446 |
} |
} |
447 |
} else { |
} else { |
448 |
my $key = $self->fill_in($rec,$i->{'key'}); |
my $key = $self->fill_in($rec,$i->{'key'}) || next; |
449 |
my @val = $self->fill_in($rec,$i->{'val'}); |
my @val = $self->fill_in($rec,$i->{'val'}) || next; |
450 |
if ($key && @val) { |
$log->debug("stored $key = ",sub { join(" | ",@val) }); |
451 |
$log->debug("stored $key = ",sub { join(" | ",@val) }); |
push @{$self->{'lookup'}->{$key}}, @val; |
|
push @{$self->{'lookup'}->{$key}}, @val; |
|
|
} |
|
452 |
} |
} |
453 |
} |
} |
454 |
} |
} |
479 |
|
|
480 |
if ($$rec->{$f}) { |
if ($$rec->{$f}) { |
481 |
return '' if (! $$rec->{$f}->[$i]); |
return '' if (! $$rec->{$f}->[$i]); |
482 |
|
no strict 'refs'; |
483 |
if ($sf && $$rec->{$f}->[$i]->{$sf}) { |
if ($sf && $$rec->{$f}->[$i]->{$sf}) { |
484 |
$$found++ if (defined($$found)); |
$$found++ if (defined($$found)); |
485 |
return $$rec->{$f}->[$i]->{$sf}; |
return $$rec->{$f}->[$i]->{$sf}; |
520 |
This function B<does not> perform parsing of format to inteligenty skip |
This function B<does not> perform parsing of format to inteligenty skip |
521 |
delimiters before fields which aren't used. |
delimiters before fields which aren't used. |
522 |
|
|
523 |
|
This method will automatically decode UTF-8 string to local code page |
524 |
|
if needed. |
525 |
|
|
526 |
=cut |
=cut |
527 |
|
|
528 |
sub fill_in { |
sub fill_in { |
538 |
# FIXME remove for speedup? |
# FIXME remove for speedup? |
539 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
540 |
|
|
541 |
|
if (utf8::is_utf8($format)) { |
542 |
|
$format = $self->_x($format); |
543 |
|
} |
544 |
|
|
545 |
my $found = 0; |
my $found = 0; |
546 |
|
|
547 |
my $eval_code; |
my $eval_code; |
761 |
my @ds = $webpac->data_structure($rec); |
my @ds = $webpac->data_structure($rec); |
762 |
|
|
763 |
This method will also set C<$webpac->{'currnet_filename'}> if there is |
This method will also set C<$webpac->{'currnet_filename'}> if there is |
764 |
<filename> tag in C<import_xml>. |
<filename> tag in C<import_xml> and C<$webpac->{'headline'}> if there is |
765 |
|
<headline> tag. |
766 |
|
|
767 |
=cut |
=cut |
768 |
|
|
775 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
776 |
|
|
777 |
undef $self->{'currnet_filename'}; |
undef $self->{'currnet_filename'}; |
778 |
|
undef $self->{'headline'}; |
779 |
|
|
780 |
my @sorted_tags; |
my @sorted_tags; |
781 |
if ($self->{tags_by_order}) { |
if ($self->{tags_by_order}) { |
813 |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
814 |
} |
} |
815 |
|
|
816 |
|
if ($field eq 'filename') { |
817 |
|
$self->{'current_filename'} = join('',@v); |
818 |
|
$log->debug("filename: ",$self->{'current_filename'}); |
819 |
|
} elsif ($field eq 'headline') { |
820 |
|
$self->{'headline'} .= join('',@v); |
821 |
|
$log->debug("headline: ",$self->{'headline'}); |
822 |
|
next; # don't return headline in data_structure! |
823 |
|
} |
824 |
|
|
825 |
# does tag have type? |
# does tag have type? |
826 |
if ($tag->{'type'}) { |
if ($tag->{'type'}) { |
827 |
push @{$row->{$tag->{'type'}}}, @v; |
push @{$row->{$tag->{'type'}}}, @v; |
830 |
push @{$row->{'swish'}}, @v; |
push @{$row->{'swish'}}, @v; |
831 |
} |
} |
832 |
|
|
|
if ($field eq 'filename') { |
|
|
$self->{'current_filename'} = join('',@v); |
|
|
$log->debug("filename: ",$self->{'current_filename'}); |
|
|
} |
|
833 |
|
|
834 |
} |
} |
835 |
|
|
880 |
return $out; |
return $out; |
881 |
} |
} |
882 |
|
|
883 |
|
=head2 output_file |
884 |
|
|
885 |
|
Create output from in-memory data structure using Template Toolkit template |
886 |
|
to a file. |
887 |
|
|
888 |
|
$webpac->output_file( |
889 |
|
file => 'out.txt', |
890 |
|
template => 'text.tt', |
891 |
|
data => @ds |
892 |
|
); |
893 |
|
|
894 |
|
=cut |
895 |
|
|
896 |
|
sub output_file { |
897 |
|
my $self = shift; |
898 |
|
|
899 |
|
my $args = {@_}; |
900 |
|
|
901 |
|
my $log = $self->_get_logger(); |
902 |
|
|
903 |
|
my $file = $args->{'file'} || $log->logconfess("need file name"); |
904 |
|
|
905 |
|
$log->debug("creating file ",$file); |
906 |
|
|
907 |
|
open(my $fh, ">", $file) || $log->logdie("can't open output file '$file': $!"); |
908 |
|
print $fh $self->output( |
909 |
|
template => $args->{'template'}, |
910 |
|
data => $args->{'data'}, |
911 |
|
) || $log->logdie("print: $!"); |
912 |
|
close($fh) || $log->logdie("close: $!"); |
913 |
|
} |
914 |
|
|
915 |
=head2 apply_format |
=head2 apply_format |
916 |
|
|
917 |
Apply format specified in tag with C<format_name="name"> and |
Apply format specified in tag with C<format_name="name"> and |
1055 |
also use method names, and not only classes (which are just few) |
also use method names, and not only classes (which are just few) |
1056 |
to filter logging. |
to filter logging. |
1057 |
|
|
1058 |
|
|
1059 |
|
=head1 MEMORY USAGE |
1060 |
|
|
1061 |
|
C<low_mem> options is double-edged sword. If enabled, WebPAC |
1062 |
|
will run on memory constraint machines (which doesn't have enough |
1063 |
|
physical RAM to create memory structure for whole source database). |
1064 |
|
|
1065 |
|
If your machine has 512Mb or more of RAM and database is around 10000 records, |
1066 |
|
memory shouldn't be an issue. If you don't have enough physical RAM, you |
1067 |
|
might consider using virtual memory (if your operating system is handling it |
1068 |
|
well, like on FreeBSD or Linux) instead of dropping to L<DBD::Deep> to handle |
1069 |
|
parsed structure of ISIS database (this is what C<low_mem> option does). |
1070 |
|
|
1071 |
|
Hitting swap at end of reading source database is probably o.k. However, |
1072 |
|
hitting swap before 90% will dramatically decrease performance and you will |
1073 |
|
be better off with C<low_mem> and using rest of availble memory for |
1074 |
|
operating system disk cache (Linux is particuallary good about this). |
1075 |
|
However, every access to database record will require disk access, so |
1076 |
|
generation phase will be slower 10-100 times. |
1077 |
|
|
1078 |
|
Parsed structures are essential - you just have option to trade RAM memory |
1079 |
|
(which is fast) for disk space (which is slow). Be sure to have planty of |
1080 |
|
disk space if you are using C<low_mem> and thus L<DBD::Deep>. |
1081 |
|
|
1082 |
|
However, when WebPAC is running on desktop machines (or laptops :-), it's |
1083 |
|
highly undesireable for system to start swapping. Using C<low_mem> option can |
1084 |
|
reduce WecPAC memory usage to around 64Mb for same database with lookup |
1085 |
|
fields and sorted indexes which stay in RAM. Performance will suffer, but |
1086 |
|
memory usage will really be minimal. It might be also more confortable to |
1087 |
|
run WebPAC reniced on those machines. |
1088 |
|
|
1089 |
=cut |
=cut |
1090 |
|
|
1091 |
1; |
1; |