9 use base qw/WebPAC::Common/;
10 use Data::Dump qw/dump/;
11 use Encode qw/from_to/;
15 WebPAC::Input - read different file formats into WebPAC
23 our $VERSION = '0.16';
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',
48 $db->open( path => '/path/to/database' );
49 print "database size: ",$db->size,"\n";
50 while (my $rec = $db->fetch) {
51 # do something with $rec
60 Create new input database object.
62 my $db = new WebPAC::Input(
63 module => 'WebPAC::Input::MARC',
64 encoding => 'ISO-8859-2',
65 recode => 'char pairs',
69 C<module> is low-level file format module. See L<WebPAC::Input::ISIS> and
70 L<WebPAC::Input::MARC>.
72 Optional parametar C<encoding> specify application code page (which will be
73 used internally). This should probably be your terminal encoding, and by
74 default, it C<ISO-8859-2>.
76 C<recode> is optional string constisting of character or words pairs that
77 should be replaced in input stream.
79 C<no_progress_bar> disables progress bar output on C<STDOUT>
81 This function will also call low-level C<init> if it exists with same
91 my $log = $self->_get_logger;
93 $log->logconfess("code_page argument is not suppored any more. change it to encoding") if ($self->{lookup});
94 $log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_ref") if ($self->{lookup});
95 $log->logconfess("low_mem argument is not suppored any more. rewrite it to load_row and save_row") if ($self->{low_mem});
97 $log->logconfess("specify low-level file format module") unless ($self->{module});
98 my $module_path = $self->{module};
99 $module_path =~ s#::#/#g;
100 $module_path .= '.pm';
101 $log->debug("require low-level module $self->{module} from $module_path");
103 require $module_path;
105 $self->{'encoding'} ||= 'ISO-8859-2';
107 $self ? return $self : return undef;
112 This function will read whole database in memory and produce lookups.
114 my $store; # simple in-memory hash
117 path => '/path/to/database/file',
118 code_page => 'cp852',
122 lookup_coderef => sub {
127 900 => { '^a' => { ' : ' => '^b' } },
128 901 => { '*' => { '^b' => ' ; ' } },
130 modify_file => 'conf/modify/mapping.map',
133 $store->{ $a->{id} } = $a->{row};
137 return defined($store->{ $a->{id} }) &&
138 $store->{ $a->{id} };
143 By default, C<code_page> is assumed to be C<cp852>.
145 C<offset> is optional parametar to position at some offset before reading from database.
147 C<limit> is optional parametar to read just C<limit> records from database
149 C<stats> create optional report about usage of fields and subfields
151 C<lookup_coderef> is closure to called to save data into lookups
153 C<modify_records> specify mapping from subfields to delimiters or from
154 delimiters to subfields, as well as oprations on fields (if subfield is
157 C<modify_file> is alternative for C<modify_records> above which preserves order and offers
158 (hopefully) simplier sintax than YAML or perl (see L</modify_file_regex>). This option
159 overrides C<modify_records> if both exists for same input.
161 C<save_row> and C<load_row> are low-level implementation of store engine. Calling convention
162 is documented in example above.
164 Returns size of database, regardless of C<offset> and C<limit>
165 parametars, see also C<size>.
173 my $log = $self->_get_logger();
175 $log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_coderef") if ($arg->{lookup});
176 $log->logconfess("lookup_coderef must be CODE, not ",ref($arg->{lookup_coderef}))
177 if ($arg->{lookup_coderef} && ref($arg->{lookup_coderef}) ne 'CODE');
179 $log->debug( $arg->{lookup_coderef} ? '' : 'not ', "using lookup_coderef");
181 $log->logcroak("need path") if (! $arg->{'path'});
182 my $code_page = $arg->{'code_page'} || 'cp852';
184 # store data in object
185 $self->{'input_code_page'} = $code_page;
186 foreach my $v (qw/path offset limit/) {
187 $self->{$v} = $arg->{$v} if ($arg->{$v});
190 if ($arg->{load_row} || $arg->{save_row}) {
191 $log->logconfess("save_row and load_row must be defined in pair and be CODE") unless (
192 ref($arg->{load_row}) eq 'CODE' &&
193 ref($arg->{save_row}) eq 'CODE'
195 $self->{load_row} = $arg->{load_row};
196 $self->{save_row} = $arg->{save_row};
197 $log->debug("using load_row and save_row instead of in-memory hash");
204 if ($self->{recode}) {
205 my @r = split(/\s/, $self->{recode});
207 $log->logwarn("recode needs even number of elements (some number of valid pairs)");
212 $recode_map->{$from} = $to;
215 $recode_regex = join '|' => keys %{ $recode_map };
217 $log->debug("using recode regex: $recode_regex");
223 if (my $p = $arg->{modify_file}) {
224 $log->debug("using modify_file $p");
225 $rec_regex = $self->modify_file_regexps( $p );
226 } elsif (my $h = $arg->{modify_records}) {
227 $log->debug("using modify_records ", sub { dump( $h ) });
228 $rec_regex = $self->modify_record_regexps(%{ $h });
230 $log->debug("rec_regex: ", sub { dump($rec_regex) }) if ($rec_regex);
232 my $class = $self->{module} || $log->logconfess("can't get low-level module name!");
234 my $ll_db = $class->new(
235 path => $arg->{path},
237 # my ($l,$f_nr) = @_;
238 # return unless defined($l);
239 # from_to($l, $code_page, $self->{'encoding'});
240 # $l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map);
246 unless (defined($ll_db)) {
247 $log->logwarn("can't open database $arg->{path}, skipping...");
251 my $size = $ll_db->size;
254 $log->logwarn("no records in database $arg->{path}, skipping...");
261 if (my $s = $self->{offset}) {
262 $log->debug("skipping to MFN $s");
265 $self->{offset} = $from_rec;
268 if ($self->{limit}) {
269 $log->debug("limiting to ",$self->{limit}," records");
270 $to_rec = $from_rec + $self->{limit} - 1;
271 $to_rec = $size if ($to_rec > $size);
274 # store size for later
275 $self->{size} = ($to_rec - $from_rec) ? ($to_rec - $from_rec + 1) : 0;
277 $log->info("processing $self->{size}/$size records [$from_rec-$to_rec] convert $code_page -> $self->{encoding}", $self->{stats} ? ' [stats]' : '');
280 for (my $pos = $from_rec; $pos <= $to_rec; $pos++) {
282 $log->debug("position: $pos\n");
284 my $rec = $ll_db->fetch_rec($pos, sub {
286 # return unless defined($l);
287 # return $l unless ($rec_regex && $f_nr);
289 $log->debug("-=> $f_nr ## $l");
291 # codepage conversion and recode_regex
292 from_to($l, $code_page, $self->{'encoding'});
293 $l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map);
296 if ($rec_regex && defined($rec_regex->{$f_nr})) {
297 $log->logconfess("regexps->{$f_nr} must be ARRAY") if (ref($rec_regex->{$f_nr}) ne 'ARRAY');
299 foreach my $r (@{ $rec_regex->{$f_nr} }) {
303 $log->debug("REGEX on $f_nr eval \$l =~ $r\n## old l: [$old_l]\n## new l: [$l]");
305 $log->error("error applying regex: $r") if ($@);
309 $log->debug("<=- $f_nr ## $l");
313 $log->debug(sub { dump($rec) });
316 $log->warn("record $pos empty? skipping...");
321 if ($self->{save_row}) {
322 $self->{save_row}->({
327 $self->{data}->{$pos} = $rec;
331 $arg->{'lookup_coderef'}->( $rec ) if ($rec && $arg->{'lookup_coderef'});
333 # update counters for statistics
334 if ($self->{stats}) {
336 # fetch clean record with regexpes applied for statistics
337 my $rec = $ll_db->fetch_rec($pos);
339 foreach my $fld (keys %{ $rec }) {
340 $self->{_stats}->{fld}->{ $fld }++;
342 $log->logdie("invalid record fild $fld, not ARRAY")
343 unless (ref($rec->{ $fld }) eq 'ARRAY');
345 foreach my $row (@{ $rec->{$fld} }) {
347 if (ref($row) eq 'HASH') {
349 foreach my $sf (keys %{ $row }) {
350 next if ($sf eq 'subfields');
351 $self->{_stats}->{sf}->{ $fld }->{ $sf }->{count}++;
352 $self->{_stats}->{sf}->{ $fld }->{ $sf }->{repeatable}++
353 if (ref($row->{$sf}) eq 'ARRAY');
357 $self->{_stats}->{repeatable}->{ $fld }++;
363 $self->progress_bar($pos,$to_rec) unless ($self->{no_progress_bar});
368 $self->{last_pcnt} = 0;
370 # store max mfn and return it.
371 $self->{max_pos} = $to_rec;
372 $log->debug("max_pos: $to_rec");
375 $self->{ll_db} = $ll_db;
382 Fetch next record from database. It will also displays progress bar.
384 my $rec = $isis->fetch;
386 Record from this function should probably go to C<data_structure> for
394 my $log = $self->_get_logger();
396 $log->logconfess("it seems that you didn't load database!") unless ($self->{pos});
398 if ($self->{pos} == -1) {
399 $self->{pos} = $self->{offset};
404 my $mfn = $self->{pos};
406 if ($mfn > $self->{max_pos}) {
407 $self->{pos} = $self->{max_pos};
408 $log->debug("at EOF");
412 $self->progress_bar($mfn,$self->{max_pos}) unless ($self->{no_progress_bar});
416 if ($self->{load_row}) {
417 $rec = $self->{load_row}->({ id => $mfn });
419 $rec = $self->{data}->{$mfn};
427 Returns current record number (MFN).
431 First record in database has position 1.
443 Returns number of records in database
447 Result from this function can be used to loop through all records
449 foreach my $mfn ( 1 ... $isis->size ) { ... }
451 because it takes into account C<offset> and C<limit>.
457 return $self->{size};
462 Seek to specified MFN in file.
466 First record in database has position 1.
472 my $pos = shift || return;
474 my $log = $self->_get_logger();
477 $log->warn("seek before first record");
479 } elsif ($pos > $self->{max_pos}) {
480 $log->warn("seek beyond last record");
481 $pos = $self->{max_pos};
484 return $self->{pos} = (($pos - 1) || -1);
489 Dump statistics about field and subfield usage
498 my $log = $self->_get_logger();
500 my $s = $self->{_stats};
502 $log->warn("called stats, but there is no statistics collected");
510 my $f = $_ || die "no field";
511 my $v = $s->{fld}->{$f} || die "no s->{fld}->{$f}";
512 $max_fld = $v if ($v > $max_fld);
514 my $o = sprintf("%4s %d ~", $f, $v);
516 if (defined($s->{sf}->{$f})) {
518 $o .= sprintf(" %s:%d%s", $_,
519 $s->{sf}->{$f}->{$_}->{count},
520 $s->{sf}->{$f}->{$_}->{repeatable} ? '*' : '',
522 } sort keys %{ $s->{sf}->{$f} };
525 if (my $v_r = $s->{repeatable}->{$f}) {
526 $o .= " ($v_r)" if ($v_r != $v);
530 } sort { $a cmp $b } keys %{ $s->{fld} }
533 $log->debug( sub { dump($s) } );
540 Display humanly readable dump of record
547 return unless $self->{ll_db};
549 if ($self->{ll_db}->can('dump_rec')) {
550 return $self->{ll_db}->dump_ascii( $self->{pos} );
552 return dump( $self->{ll_db}->fetch_rec( $self->{pos} ) );
556 =head2 modify_record_regexps
558 Generate hash with regexpes to be applied using l<filter>.
560 my $regexpes = $input->modify_record_regexps(
561 900 => { '^a' => { ' : ' => '^b' } },
562 901 => { '*' => { '^b' => ' ; ' } },
568 my ($sf,$from,$to) = @_;
569 if ($from =~ m/^regex:(.+)$/) {
572 $from = '\Q' . $from . '\E';
576 's/\Q'. $sf .'\E([^\^]*?)'. $from .'([^\^]*?)/'. $sf .'$1'. $to .'$2/';
579 's/'. $from .'/'. $to .'/g';
583 sub modify_record_regexps {
585 my $modify_record = {@_};
589 my $log = $self->_get_logger();
591 foreach my $f (keys %$modify_record) {
592 $log->debug("field: $f");
594 foreach my $sf (keys %{ $modify_record->{$f} }) {
595 $log->debug("subfield: $sf");
597 foreach my $from (keys %{ $modify_record->{$f}->{$sf} }) {
598 my $to = $modify_record->{$f}->{$sf}->{$from};
599 #die "no field?" unless defined($to);
600 $log->debug("transform: |$from| -> |$to|");
602 my $regex = _get_regex($sf,$from,$to);
603 push @{ $regexpes->{$f} }, $regex;
604 $log->debug("regex: $regex");
612 =head2 modify_file_regexps
614 Generate hash with regexpes to be applied using l<filter> from
615 pseudo hash/yaml format for regex mappings.
617 It should be obvious:
624 In field I<200> find C<'^a'> and then C<' : '>, and replace it with C<'^e'>.
625 In field I<200> find C<'^a'> and then C<' = '>, and replace it with C<'^d'>.
627 my $regexpes = $input->modify_file_regexps( 'conf/modify/common.pl' );
629 On undef path it will just return.
633 sub modify_file_regexps {
636 my $modify_path = shift || return;
638 my $log = $self->_get_logger();
642 CORE::open(my $fh, $modify_path) || $log->logdie("can't open modify file $modify_path: $!");
648 next if (/^#/ || /^\s*$/);
650 if (/^\s*(\d+)\s*$/) {
652 $log->debug("field: $f");
654 } elsif (/^\s*'([^']*)'\s*$/) {
656 $log->die("can't define subfiled before field in: $_") unless ($f);
657 $log->debug("subfield: $sf");
658 } elsif (/^\s*'([^']*)'\s*=>\s*'([^']*)'\s*$/) {
659 my ($from,$to) = ($1, $2);
661 $log->debug("transform: |$from| -> |$to|");
663 my $regex = _get_regex($sf,$from,$to);
664 push @{ $regexpes->{$f} }, $regex;
665 $log->debug("regex: $regex");
674 Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
676 =head1 COPYRIGHT & LICENSE
678 Copyright 2005-2006 Dobrica Pavlinusic, All Rights Reserved.
680 This program is free software; you can redistribute it and/or modify it
681 under the same terms as Perl itself.
685 1; # End of WebPAC::Input