| 1 |
package WebPAC::Input; |
| 2 |
|
| 3 |
use warnings; |
| 4 |
use strict; |
| 5 |
|
| 6 |
use blib; |
| 7 |
|
| 8 |
use WebPAC::Common; |
| 9 |
use base qw/WebPAC::Common/; |
| 10 |
use Text::Iconv; |
| 11 |
|
| 12 |
=head1 NAME |
| 13 |
|
| 14 |
WebPAC::Input - core module for input file format |
| 15 |
|
| 16 |
=head1 VERSION |
| 17 |
|
| 18 |
Version 0.02 |
| 19 |
|
| 20 |
=cut |
| 21 |
|
| 22 |
our $VERSION = '0.02'; |
| 23 |
|
| 24 |
=head1 SYNOPSIS |
| 25 |
|
| 26 |
This module is used as base class for all database specific modules |
| 27 |
(basically, files which have one handle, fixed size while indexing and some |
| 28 |
kind of numeric idefinirier which goes from 1 to filesize). |
| 29 |
|
| 30 |
Perhaps a little code snippet. |
| 31 |
|
| 32 |
use WebPAC::Input; |
| 33 |
|
| 34 |
my $db = WebPAC::Input->new( |
| 35 |
format => 'NULL', |
| 36 |
config => $config, |
| 37 |
lookup => $lookup_obj, |
| 38 |
low_mem => 1, |
| 39 |
); |
| 40 |
|
| 41 |
$db->open('/path/to/database'); |
| 42 |
print "database size: ",$db->size,"\n"; |
| 43 |
while (my $row = $db->fetch) { |
| 44 |
... |
| 45 |
} |
| 46 |
|
| 47 |
=head1 FUNCTIONS |
| 48 |
|
| 49 |
=head2 new |
| 50 |
|
| 51 |
Create new input database object. |
| 52 |
|
| 53 |
my $db = new WebPAC::Input( |
| 54 |
format => 'NULL' |
| 55 |
code_page => 'ISO-8859-2', |
| 56 |
low_mem => 1, |
| 57 |
); |
| 58 |
|
| 59 |
Optional parametar C<code_page> specify application code page (which will be |
| 60 |
used internally). This should probably be your terminal encoding, and by |
| 61 |
default, it C<ISO-8859-2>. |
| 62 |
|
| 63 |
Default is not to use C<low_mem> options (see L<MEMORY USAGE> below). |
| 64 |
|
| 65 |
This function will also call low-level C<init> if it exists with same |
| 66 |
parametars. |
| 67 |
|
| 68 |
=cut |
| 69 |
|
| 70 |
sub new { |
| 71 |
my $class = shift; |
| 72 |
my $self = {@_}; |
| 73 |
bless($self, $class); |
| 74 |
|
| 75 |
my $log = $self->_get_logger; |
| 76 |
|
| 77 |
# check if required subclasses are implemented |
| 78 |
foreach my $subclass (qw/open_db fetch_rec/) { |
| 79 |
$log->logdie("missing implementation of $subclass") unless ($self->SUPER::can($subclass)); |
| 80 |
} |
| 81 |
|
| 82 |
if ($self->can('init')) { |
| 83 |
$log->debug("calling init"); |
| 84 |
$self->init(@_); |
| 85 |
} |
| 86 |
|
| 87 |
$self->{'code_page'} ||= 'ISO-8859-2'; |
| 88 |
|
| 89 |
# running with low_mem flag? well, use DBM::Deep then. |
| 90 |
if ($self->{'low_mem'}) { |
| 91 |
$log->info("running with low_mem which impacts performance (<32 Mb memory usage)"); |
| 92 |
|
| 93 |
my $db_file = "data.db"; |
| 94 |
|
| 95 |
if (-e $db_file) { |
| 96 |
unlink $db_file or $log->logdie("can't remove '$db_file' from last run"); |
| 97 |
$log->debug("removed '$db_file' from last run"); |
| 98 |
} |
| 99 |
|
| 100 |
require DBM::Deep; |
| 101 |
|
| 102 |
my $db = new DBM::Deep $db_file; |
| 103 |
|
| 104 |
$log->logdie("DBM::Deep error: $!") unless ($db); |
| 105 |
|
| 106 |
if ($db->error()) { |
| 107 |
$log->logdie("can't open '$db_file' under low_mem: ",$db->error()); |
| 108 |
} else { |
| 109 |
$log->debug("using file '$db_file' for DBM::Deep"); |
| 110 |
} |
| 111 |
|
| 112 |
$self->{'db'} = $db; |
| 113 |
} |
| 114 |
|
| 115 |
$self ? return $self : return undef; |
| 116 |
} |
| 117 |
|
| 118 |
=head2 open |
| 119 |
|
| 120 |
This function will read whole database in memory and produce lookups. |
| 121 |
|
| 122 |
$isis->open( |
| 123 |
path => '/path/to/database/file', |
| 124 |
code_page => '852', |
| 125 |
limit_mfn => 500, |
| 126 |
start_mfn => 6000, |
| 127 |
lookup => $lookup_obj, |
| 128 |
); |
| 129 |
|
| 130 |
By default, C<code_page> is assumed to be C<852>. |
| 131 |
|
| 132 |
If optional parametar C<start_mfn> is set, this will be first MFN to read |
| 133 |
from database (so you can skip beginning of your database if you need to). |
| 134 |
|
| 135 |
If optional parametar C<limit_mfn> is set, it will read just 500 records |
| 136 |
from database in example above. |
| 137 |
|
| 138 |
Returns size of database, regardless of C<start_mfn> and C<limit_mfn> |
| 139 |
parametars, see also C<$isis->size>. |
| 140 |
|
| 141 |
=cut |
| 142 |
|
| 143 |
sub open { |
| 144 |
my $self = shift; |
| 145 |
my $arg = {@_}; |
| 146 |
|
| 147 |
my $log = $self->_get_logger(); |
| 148 |
|
| 149 |
$log->logcroak("need path") if (! $arg->{'path'}); |
| 150 |
my $code_page = $arg->{'code_page'} || '852'; |
| 151 |
|
| 152 |
# store data in object |
| 153 |
$self->{'code_page'} = $code_page; |
| 154 |
foreach my $v (qw/path start_mfn limit_mfn/) { |
| 155 |
$self->{$v} = $arg->{$v} if ($arg->{$v}); |
| 156 |
} |
| 157 |
|
| 158 |
# create Text::Iconv object |
| 159 |
$self->{iconv} = Text::Iconv->new($code_page,$self->{'code_page'}); |
| 160 |
|
| 161 |
my ($db, $size) = $self->open_db( |
| 162 |
path => $arg->{path}, |
| 163 |
); |
| 164 |
|
| 165 |
unless ($db) { |
| 166 |
$log->logwarn("can't open database $arg->{path}, skipping..."); |
| 167 |
return; |
| 168 |
} |
| 169 |
|
| 170 |
unless ($size) { |
| 171 |
$log->logwarn("no records in database $arg->{path}, skipping..."); |
| 172 |
return; |
| 173 |
} |
| 174 |
|
| 175 |
my $startmfn = 1; |
| 176 |
my $maxmfn = $size; |
| 177 |
|
| 178 |
if (my $s = $self->{start_mfn}) { |
| 179 |
$log->info("skipping to MFN $s"); |
| 180 |
$startmfn = $s; |
| 181 |
} else { |
| 182 |
$self->{start_mfn} = $startmfn; |
| 183 |
} |
| 184 |
|
| 185 |
if ($self->{limit_mfn}) { |
| 186 |
$log->info("limiting to ",$self->{limit_mfn}," records"); |
| 187 |
$maxmfn = $startmfn + $self->{limit_mfn} - 1; |
| 188 |
$maxmfn = $size if ($maxmfn > $size); |
| 189 |
} |
| 190 |
|
| 191 |
# store size for later |
| 192 |
$self->{size} = ($maxmfn - $startmfn) ? ($maxmfn - $startmfn + 1) : 0; |
| 193 |
|
| 194 |
$log->info("processing $self->{size} records in $code_page, convert to $self->{code_page}"); |
| 195 |
|
| 196 |
# read database |
| 197 |
for (my $mfn = $startmfn; $mfn <= $maxmfn; $mfn++) { |
| 198 |
|
| 199 |
$log->debug("mfn: $mfn\n"); |
| 200 |
|
| 201 |
my $rec = $self->fetch_rec( $db, $mfn ); |
| 202 |
|
| 203 |
if (! $rec) { |
| 204 |
$log->warn("record $mfn empty? skipping..."); |
| 205 |
next; |
| 206 |
} |
| 207 |
|
| 208 |
# store |
| 209 |
if ($self->{'low_mem'}) { |
| 210 |
$self->{'db'}->put($mfn, $rec); |
| 211 |
} else { |
| 212 |
$self->{'data'}->{$mfn} = $rec; |
| 213 |
} |
| 214 |
|
| 215 |
# create lookup |
| 216 |
$self->{'lookup'}->add( $rec ) if ($rec && $self->{'lookup'}); |
| 217 |
|
| 218 |
$self->progress_bar($mfn,$maxmfn); |
| 219 |
|
| 220 |
} |
| 221 |
|
| 222 |
$self->{'current_mfn'} = -1; |
| 223 |
$self->{'last_pcnt'} = 0; |
| 224 |
|
| 225 |
$log->debug("max mfn: $maxmfn"); |
| 226 |
|
| 227 |
# store max mfn and return it. |
| 228 |
$self->{'max_mfn'} = $maxmfn; |
| 229 |
|
| 230 |
return $size; |
| 231 |
} |
| 232 |
|
| 233 |
=head2 fetch |
| 234 |
|
| 235 |
Fetch next record from database. It will also displays progress bar. |
| 236 |
|
| 237 |
my $rec = $isis->fetch; |
| 238 |
|
| 239 |
Record from this function should probably go to C<data_structure> for |
| 240 |
normalisation. |
| 241 |
|
| 242 |
=cut |
| 243 |
|
| 244 |
sub fetch { |
| 245 |
my $self = shift; |
| 246 |
|
| 247 |
my $log = $self->_get_logger(); |
| 248 |
|
| 249 |
$log->logconfess("it seems that you didn't load database!") unless ($self->{'current_mfn'}); |
| 250 |
|
| 251 |
if ($self->{'current_mfn'} == -1) { |
| 252 |
$self->{'current_mfn'} = $self->{'start_mfn'}; |
| 253 |
} else { |
| 254 |
$self->{'current_mfn'}++; |
| 255 |
} |
| 256 |
|
| 257 |
my $mfn = $self->{'current_mfn'}; |
| 258 |
|
| 259 |
if ($mfn > $self->{'max_mfn'}) { |
| 260 |
$self->{'current_mfn'} = $self->{'max_mfn'}; |
| 261 |
$log->debug("at EOF"); |
| 262 |
return; |
| 263 |
} |
| 264 |
|
| 265 |
$self->progress_bar($mfn,$self->{'max_mfn'}); |
| 266 |
|
| 267 |
my $rec; |
| 268 |
|
| 269 |
if ($self->{'low_mem'}) { |
| 270 |
$rec = $self->{'db'}->get($mfn); |
| 271 |
} else { |
| 272 |
$rec = $self->{'data'}->{$mfn}; |
| 273 |
} |
| 274 |
|
| 275 |
$rec ||= 0E0; |
| 276 |
} |
| 277 |
|
| 278 |
=head2 pos |
| 279 |
|
| 280 |
Returns current record number (MFN). |
| 281 |
|
| 282 |
print $isis->pos; |
| 283 |
|
| 284 |
First record in database has position 1. |
| 285 |
|
| 286 |
=cut |
| 287 |
|
| 288 |
sub pos { |
| 289 |
my $self = shift; |
| 290 |
return $self->{'current_mfn'}; |
| 291 |
} |
| 292 |
|
| 293 |
|
| 294 |
=head2 size |
| 295 |
|
| 296 |
Returns number of records in database |
| 297 |
|
| 298 |
print $isis->size; |
| 299 |
|
| 300 |
Result from this function can be used to loop through all records |
| 301 |
|
| 302 |
foreach my $mfn ( 1 ... $isis->size ) { ... } |
| 303 |
|
| 304 |
because it takes into account C<start_mfn> and C<limit_mfn>. |
| 305 |
|
| 306 |
=cut |
| 307 |
|
| 308 |
sub size { |
| 309 |
my $self = shift; |
| 310 |
return $self->{'size'}; |
| 311 |
} |
| 312 |
|
| 313 |
=head2 seek |
| 314 |
|
| 315 |
Seek to specified MFN in file. |
| 316 |
|
| 317 |
$isis->seek(42); |
| 318 |
|
| 319 |
First record in database has position 1. |
| 320 |
|
| 321 |
=cut |
| 322 |
|
| 323 |
sub seek { |
| 324 |
my $self = shift; |
| 325 |
my $pos = shift || return; |
| 326 |
|
| 327 |
my $log = $self->_get_logger(); |
| 328 |
|
| 329 |
if ($pos < 1) { |
| 330 |
$log->warn("seek before first record"); |
| 331 |
$pos = 1; |
| 332 |
} elsif ($pos > $self->{'max_mfn'}) { |
| 333 |
$log->warn("seek beyond last record"); |
| 334 |
$pos = $self->{'max_mfn'}; |
| 335 |
} |
| 336 |
|
| 337 |
return $self->{'current_mfn'} = (($pos - 1) || -1); |
| 338 |
} |
| 339 |
|
| 340 |
|
| 341 |
=head1 MEMORY USAGE |
| 342 |
|
| 343 |
C<low_mem> options is double-edged sword. If enabled, WebPAC |
| 344 |
will run on memory constraint machines (which doesn't have enough |
| 345 |
physical RAM to create memory structure for whole source database). |
| 346 |
|
| 347 |
If your machine has 512Mb or more of RAM and database is around 10000 records, |
| 348 |
memory shouldn't be an issue. If you don't have enough physical RAM, you |
| 349 |
might consider using virtual memory (if your operating system is handling it |
| 350 |
well, like on FreeBSD or Linux) instead of dropping to L<DBM::Deep> to handle |
| 351 |
parsed structure of ISIS database (this is what C<low_mem> option does). |
| 352 |
|
| 353 |
Hitting swap at end of reading source database is probably o.k. However, |
| 354 |
hitting swap before 90% will dramatically decrease performance and you will |
| 355 |
be better off with C<low_mem> and using rest of availble memory for |
| 356 |
operating system disk cache (Linux is particuallary good about this). |
| 357 |
However, every access to database record will require disk access, so |
| 358 |
generation phase will be slower 10-100 times. |
| 359 |
|
| 360 |
Parsed structures are essential - you just have option to trade RAM memory |
| 361 |
(which is fast) for disk space (which is slow). Be sure to have planty of |
| 362 |
disk space if you are using C<low_mem> and thus L<DBM::Deep>. |
| 363 |
|
| 364 |
However, when WebPAC is running on desktop machines (or laptops :-), it's |
| 365 |
highly undesireable for system to start swapping. Using C<low_mem> option can |
| 366 |
reduce WecPAC memory usage to around 64Mb for same database with lookup |
| 367 |
fields and sorted indexes which stay in RAM. Performance will suffer, but |
| 368 |
memory usage will really be minimal. It might be also more confortable to |
| 369 |
run WebPAC reniced on those machines. |
| 370 |
|
| 371 |
|
| 372 |
=head1 AUTHOR |
| 373 |
|
| 374 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
| 375 |
|
| 376 |
=head1 COPYRIGHT & LICENSE |
| 377 |
|
| 378 |
Copyright 2005 Dobrica Pavlinusic, All Rights Reserved. |
| 379 |
|
| 380 |
This program is free software; you can redistribute it and/or modify it |
| 381 |
under the same terms as Perl itself. |
| 382 |
|
| 383 |
=cut |
| 384 |
|
| 385 |
1; # End of WebPAC::Input |
Parent Directory
| 
Revision Log