9 use base qw/WebPAC::Common/;
11 use Encode qw/from_to/;
15 WebPAC::Input - read different file formats into WebPAC
23 our $VERSION = '0.13';
27 This module implements input as database which have fixed and known
28 I<size> while indexing and single unique numeric identifier for database
29 position ranging from 1 to I<size>.
31 Simply, something that is indexed by unmber from 1 .. I<size>.
33 Examples of such databases are CDS/ISIS files, MARC files, lines in
36 Specific file formats are implemented using low-level interface modules,
37 located in C<WebPAC::Input::*> namespace which export C<open_db>,
38 C<fetch_rec> and optional C<init> functions.
40 Perhaps a little code snippet.
44 my $db = WebPAC::Input->new(
45 module => 'WebPAC::Input::ISIS',
49 $db->open( path => '/path/to/database' );
50 print "database size: ",$db->size,"\n";
51 while (my $rec = $db->fetch) {
52 # do something with $rec
61 Create new input database object.
63 my $db = new WebPAC::Input(
64 module => 'WebPAC::Input::MARC',
65 encoding => 'ISO-8859-2',
67 recode => 'char pairs',
71 C<module> is low-level file format module. See L<WebPAC::Input::ISIS> and
72 L<WebPAC::Input::MARC>.
74 Optional parametar C<encoding> specify application code page (which will be
75 used internally). This should probably be your terminal encoding, and by
76 default, it C<ISO-8859-2>.
78 Default is not to use C<low_mem> options (see L<MEMORY USAGE> below).
80 C<recode> is optional string constisting of character or words pairs that
81 should be replaced in input stream.
83 C<no_progress_bar> disables progress bar output on C<STDOUT>
85 This function will also call low-level C<init> if it exists with same
95 my $log = $self->_get_logger;
97 $log->logconfess("code_page argument is not suppored any more. change it to encoding") if ($self->{lookup});
98 $log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_ref") if ($self->{lookup});
100 $log->logconfess("specify low-level file format module") unless ($self->{module});
101 my $module_path = $self->{module};
102 $module_path =~ s#::#/#g;
103 $module_path .= '.pm';
104 $log->debug("require low-level module $self->{module} from $module_path");
106 require $module_path;
108 # check if required subclasses are implemented
109 foreach my $subclass (qw/open_db fetch_rec init dump_rec/) {
113 $self->{'encoding'} ||= 'ISO-8859-2';
115 # running with low_mem flag? well, use DBM::Deep then.
116 if ($self->{'low_mem'}) {
117 $log->info("running with low_mem which impacts performance (<32 Mb memory usage)");
119 my $db_file = "data.db";
122 unlink $db_file or $log->logdie("can't remove '$db_file' from last run");
123 $log->debug("removed '$db_file' from last run");
128 my $db = new DBM::Deep $db_file;
130 $log->logdie("DBM::Deep error: $!") unless ($db);
133 $log->logdie("can't open '$db_file' under low_mem: ",$db->error());
135 $log->debug("using file '$db_file' for DBM::Deep");
141 $self ? return $self : return undef;
146 This function will read whole database in memory and produce lookups.
149 path => '/path/to/database/file',
150 code_page => 'cp852',
154 lookup_coderef => sub {
159 900 => { '^a' => { ' : ' => '^b' } },
160 901 => { '*' => { '^b' => ' ; ' } },
162 modify_file => 'conf/modify/mapping.map',
165 By default, C<code_page> is assumed to be C<cp852>.
167 C<offset> is optional parametar to position at some offset before reading from database.
169 C<limit> is optional parametar to read just C<limit> records from database
171 C<stats> create optional report about usage of fields and subfields
173 C<lookup_coderef> is closure to called to save data into lookups
175 C<modify_records> specify mapping from subfields to delimiters or from
176 delimiters to subfields, as well as oprations on fields (if subfield is
179 C<modify_file> is alternative for C<modify_records> above which preserves order and offers
180 (hopefully) simplier sintax than YAML or perl (see L</modify_file_regex>). This option
181 overrides C<modify_records> if both exists for same input.
183 Returns size of database, regardless of C<offset> and C<limit>
184 parametars, see also C<size>.
192 my $log = $self->_get_logger();
194 $log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_coderef") if ($arg->{lookup});
195 $log->logconfess("lookup_coderef must be CODE, not ",ref($arg->{lookup_coderef}))
196 if ($arg->{lookup_coderef} && ref($arg->{lookup_coderef}) ne 'CODE');
198 $log->debug( $arg->{lookup_coderef} ? '' : 'not ', "using lookup_coderef");
200 $log->logcroak("need path") if (! $arg->{'path'});
201 my $code_page = $arg->{'code_page'} || 'cp852';
203 # store data in object
204 $self->{'input_code_page'} = $code_page;
205 foreach my $v (qw/path offset limit/) {
206 $self->{$v} = $arg->{$v} if ($arg->{$v});
213 if ($self->{recode}) {
214 my @r = split(/\s/, $self->{recode});
216 $log->logwarn("recode needs even number of elements (some number of valid pairs)");
221 $recode_map->{$from} = $to;
224 $recode_regex = join '|' => keys %{ $recode_map };
226 $log->debug("using recode regex: $recode_regex");
232 if (my $p = $arg->{modify_file}) {
233 $log->debug("using modify_file $p");
234 $rec_regex = $self->modify_file_regexps( $p );
235 } elsif (my $h = $arg->{modify_records}) {
236 $log->debug("using modify_records ", Dumper( $h ));
237 $rec_regex = $self->modify_record_regexps(%{ $h });
239 $log->debug("rec_regex: ", Dumper($rec_regex)) if ($rec_regex);
241 my $class = $self->{module} || $log->logconfess("can't get low-level module name!");
243 my $ll_db = $class->new(
244 path => $arg->{path},
246 # my ($l,$f_nr) = @_;
247 # return unless defined($l);
248 # from_to($l, $code_page, $self->{'encoding'});
249 # $l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map);
255 unless (defined($ll_db)) {
256 $log->logwarn("can't open database $arg->{path}, skipping...");
260 my $size = $ll_db->size;
263 $log->logwarn("no records in database $arg->{path}, skipping...");
270 if (my $s = $self->{offset}) {
271 $log->debug("skipping to MFN $s");
274 $self->{offset} = $from_rec;
277 if ($self->{limit}) {
278 $log->debug("limiting to ",$self->{limit}," records");
279 $to_rec = $from_rec + $self->{limit} - 1;
280 $to_rec = $size if ($to_rec > $size);
283 # store size for later
284 $self->{size} = ($to_rec - $from_rec) ? ($to_rec - $from_rec + 1) : 0;
286 $log->info("processing $self->{size}/$size records [$from_rec-$to_rec] convert $code_page -> $self->{encoding}", $self->{stats} ? ' [stats]' : '');
289 for (my $pos = $from_rec; $pos <= $to_rec; $pos++) {
291 $log->debug("position: $pos\n");
293 my $rec = $ll_db->fetch_rec($pos, sub {
295 # return unless defined($l);
296 # return $l unless ($rec_regex && $f_nr);
298 $log->debug("-=> $f_nr ## $l");
300 # codepage conversion and recode_regex
301 from_to($l, $code_page, $self->{'encoding'});
302 $l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map);
305 if ($rec_regex && defined($rec_regex->{$f_nr})) {
306 $log->logconfess("regexps->{$f_nr} must be ARRAY") if (ref($rec_regex->{$f_nr}) ne 'ARRAY');
308 foreach my $r (@{ $rec_regex->{$f_nr} }) {
312 $log->debug("REGEX on $f_nr eval \$l =~ $r\n## old l: [$old_l]\n## new l: [$l]");
314 $log->error("error applying regex: $r") if ($@);
318 $log->debug("<=- $f_nr ## $l");
322 $log->debug(sub { Dumper($rec) });
325 $log->warn("record $pos empty? skipping...");
330 if ($self->{low_mem}) {
331 $self->{db}->put($pos, $rec);
333 $self->{data}->{$pos} = $rec;
337 $arg->{'lookup_coderef'}->( $rec ) if ($rec && $arg->{'lookup_coderef'});
339 # update counters for statistics
340 if ($self->{stats}) {
342 # fetch clean record with regexpes applied for statistics
343 my $rec = $ll_db->fetch_rec($pos);
345 foreach my $fld (keys %{ $rec }) {
346 $self->{_stats}->{fld}->{ $fld }++;
348 $log->logdie("invalid record fild $fld, not ARRAY")
349 unless (ref($rec->{ $fld }) eq 'ARRAY');
351 foreach my $row (@{ $rec->{$fld} }) {
353 if (ref($row) eq 'HASH') {
355 foreach my $sf (keys %{ $row }) {
356 next if ($sf eq 'subfields');
357 $self->{_stats}->{sf}->{ $fld }->{ $sf }->{count}++;
358 $self->{_stats}->{sf}->{ $fld }->{ $sf }->{repeatable}++
359 if (ref($row->{$sf}) eq 'ARRAY');
363 $self->{_stats}->{repeatable}->{ $fld }++;
369 $self->progress_bar($pos,$to_rec) unless ($self->{no_progress_bar});
374 $self->{last_pcnt} = 0;
376 # store max mfn and return it.
377 $self->{max_pos} = $to_rec;
378 $log->debug("max_pos: $to_rec");
385 Fetch next record from database. It will also displays progress bar.
387 my $rec = $isis->fetch;
389 Record from this function should probably go to C<data_structure> for
397 my $log = $self->_get_logger();
399 $log->logconfess("it seems that you didn't load database!") unless ($self->{pos});
401 if ($self->{pos} == -1) {
402 $self->{pos} = $self->{offset};
407 my $mfn = $self->{pos};
409 if ($mfn > $self->{max_pos}) {
410 $self->{pos} = $self->{max_pos};
411 $log->debug("at EOF");
415 $self->progress_bar($mfn,$self->{max_pos}) unless ($self->{no_progress_bar});
419 if ($self->{low_mem}) {
420 $rec = $self->{db}->get($mfn);
422 $rec = $self->{data}->{$mfn};
430 Returns current record number (MFN).
434 First record in database has position 1.
446 Returns number of records in database
450 Result from this function can be used to loop through all records
452 foreach my $mfn ( 1 ... $isis->size ) { ... }
454 because it takes into account C<offset> and C<limit>.
460 return $self->{size};
465 Seek to specified MFN in file.
469 First record in database has position 1.
475 my $pos = shift || return;
477 my $log = $self->_get_logger();
480 $log->warn("seek before first record");
482 } elsif ($pos > $self->{max_pos}) {
483 $log->warn("seek beyond last record");
484 $pos = $self->{max_pos};
487 return $self->{pos} = (($pos - 1) || -1);
492 Dump statistics about field and subfield usage
501 my $log = $self->_get_logger();
503 my $s = $self->{_stats};
505 $log->warn("called stats, but there is no statistics collected");
513 my $f = $_ || die "no field";
514 my $v = $s->{fld}->{$f} || die "no s->{fld}->{$f}";
515 $max_fld = $v if ($v > $max_fld);
517 my $o = sprintf("%4s %d ~", $f, $v);
519 if (defined($s->{sf}->{$f})) {
521 $o .= sprintf(" %s:%d%s", $_,
522 $s->{sf}->{$f}->{$_}->{count},
523 $s->{sf}->{$f}->{$_}->{repeatable} ? '*' : '',
525 } sort keys %{ $s->{sf}->{$f} };
528 if (my $v_r = $s->{repeatable}->{$f}) {
529 $o .= " ($v_r)" if ($v_r != $v);
533 } sort { $a cmp $b } keys %{ $s->{fld} }
536 $log->debug( sub { Dumper($s) } );
543 Display humanly readable dump of record
550 return $self->{dump_rec}->($self, $self->{pos});
554 =head2 modify_record_regexps
556 Generate hash with regexpes to be applied using l<filter>.
558 my $regexpes = $input->modify_record_regexps(
559 900 => { '^a' => { ' : ' => '^b' } },
560 901 => { '*' => { '^b' => ' ; ' } },
566 my ($sf,$from,$to) = @_;
569 's/\Q'. $sf .'\E([^\^]*?)\Q'. $from .'\E([^\^]*?)/'. $sf .'$1'. $to .'$2/';
572 's/\Q'. $from .'\E/'. $to .'/g';
576 sub modify_record_regexps {
578 my $modify_record = {@_};
582 my $log = $self->_get_logger();
584 foreach my $f (keys %$modify_record) {
585 $log->debug("field: $f");
587 foreach my $sf (keys %{ $modify_record->{$f} }) {
588 $log->debug("subfield: $sf");
590 foreach my $from (keys %{ $modify_record->{$f}->{$sf} }) {
591 my $to = $modify_record->{$f}->{$sf}->{$from};
592 #die "no field?" unless defined($to);
593 $log->debug("transform: |$from| -> |$to|");
595 my $regex = _get_regex($sf,$from,$to);
596 push @{ $regexpes->{$f} }, $regex;
597 $log->debug("regex: $regex");
605 =head2 modify_file_regexps
607 Generate hash with regexpes to be applied using l<filter> from
608 pseudo hash/yaml format for regex mappings.
610 It should be obvious:
617 In field I<200> find C<'^a'> and then C<' : '>, and replace it with C<'^e'>.
618 In field I<200> find C<'^a'> and then C<' = '>, and replace it with C<'^d'>.
620 my $regexpes = $input->modify_file_regexps( 'conf/modify/common.pl' );
622 On undef path it will just return.
626 sub modify_file_regexps {
629 my $modify_path = shift || return;
631 my $log = $self->_get_logger();
635 CORE::open(my $fh, $modify_path) || $log->logdie("can't open modify file $modify_path: $!");
641 next if (/^#/ || /^\s*$/);
643 if (/^\s*(\d+)\s*$/) {
645 $log->debug("field: $f");
647 } elsif (/^\s*'([^']*)'\s*$/) {
649 $log->die("can't define subfiled before field in: $_") unless ($f);
650 $log->debug("subfield: $sf");
651 } elsif (/^\s*'([^']*)'\s*=>\s*'([^']*)'\s*$/) {
652 my ($from,$to) = ($1, $2);
654 $log->debug("transform: |$from| -> |$to|");
656 my $regex = _get_regex($sf,$from,$to);
657 push @{ $regexpes->{$f} }, $regex;
658 $log->debug("regex: $regex");
667 C<low_mem> options is double-edged sword. If enabled, WebPAC
668 will run on memory constraint machines (which doesn't have enough
669 physical RAM to create memory structure for whole source database).
671 If your machine has 512Mb or more of RAM and database is around 10000 records,
672 memory shouldn't be an issue. If you don't have enough physical RAM, you
673 might consider using virtual memory (if your operating system is handling it
674 well, like on FreeBSD or Linux) instead of dropping to L<DBM::Deep> to handle
675 parsed structure of ISIS database (this is what C<low_mem> option does).
677 Hitting swap at end of reading source database is probably o.k. However,
678 hitting swap before 90% will dramatically decrease performance and you will
679 be better off with C<low_mem> and using rest of availble memory for
680 operating system disk cache (Linux is particuallary good about this).
681 However, every access to database record will require disk access, so
682 generation phase will be slower 10-100 times.
684 Parsed structures are essential - you just have option to trade RAM memory
685 (which is fast) for disk space (which is slow). Be sure to have planty of
686 disk space if you are using C<low_mem> and thus L<DBM::Deep>.
688 However, when WebPAC is running on desktop machines (or laptops :-), it's
689 highly undesireable for system to start swapping. Using C<low_mem> option can
690 reduce WecPAC memory usage to around 64Mb for same database with lookup
691 fields and sorted indexes which stay in RAM. Performance will suffer, but
692 memory usage will really be minimal. It might be also more confortable to
693 run WebPAC reniced on those machines.
698 Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
700 =head1 COPYRIGHT & LICENSE
702 Copyright 2005-2006 Dobrica Pavlinusic, All Rights Reserved.
704 This program is free software; you can redistribute it and/or modify it
705 under the same terms as Perl itself.
709 1; # End of WebPAC::Input