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 Data::Dump qw/dump/; |
11 |
use Encode qw/from_to/; |
12 |
|
13 |
=head1 NAME |
14 |
|
15 |
WebPAC::Input - read different file formats into WebPAC |
16 |
|
17 |
=head1 VERSION |
18 |
|
19 |
Version 0.18 |
20 |
|
21 |
=cut |
22 |
|
23 |
our $VERSION = '0.18'; |
24 |
|
25 |
=head1 SYNOPSIS |
26 |
|
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>. |
30 |
|
31 |
Simply, something that is indexed by unmber from 1 .. I<size>. |
32 |
|
33 |
Examples of such databases are CDS/ISIS files, MARC files, lines in |
34 |
text file, and so on. |
35 |
|
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. |
39 |
|
40 |
Perhaps a little code snippet. |
41 |
|
42 |
use WebPAC::Input; |
43 |
|
44 |
my $db = WebPAC::Input->new( |
45 |
module => 'WebPAC::Input::ISIS', |
46 |
); |
47 |
|
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 |
52 |
} |
53 |
|
54 |
|
55 |
|
56 |
=head1 FUNCTIONS |
57 |
|
58 |
=head2 new |
59 |
|
60 |
Create new input database object. |
61 |
|
62 |
my $db = new WebPAC::Input( |
63 |
module => 'WebPAC::Input::MARC', |
64 |
encoding => 'ISO-8859-2', |
65 |
recode => 'char pairs', |
66 |
no_progress_bar => 1, |
67 |
input_config => { |
68 |
mapping => [ 'foo', 'bar', 'baz' ], |
69 |
}, |
70 |
); |
71 |
|
72 |
C<module> is low-level file format module. See L<WebPAC::Input::ISIS> and |
73 |
L<WebPAC::Input::MARC>. |
74 |
|
75 |
Optional parametar C<encoding> specify application code page (which will be |
76 |
used internally). This should probably be your terminal encoding, and by |
77 |
default, it C<ISO-8859-2>. |
78 |
|
79 |
C<recode> is optional string constisting of character or words pairs that |
80 |
should be replaced in input stream. |
81 |
|
82 |
C<no_progress_bar> disables progress bar output on C<STDOUT> |
83 |
|
84 |
This function will also call low-level C<init> if it exists with same |
85 |
parametars. |
86 |
|
87 |
=cut |
88 |
|
89 |
sub new { |
90 |
my $class = shift; |
91 |
my $self = {@_}; |
92 |
bless($self, $class); |
93 |
|
94 |
my $log = $self->_get_logger; |
95 |
|
96 |
$log->logconfess("code_page argument is not suppored any more. change it to encoding") if ($self->{lookup}); |
97 |
$log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_ref") if ($self->{lookup}); |
98 |
$log->logconfess("low_mem argument is not suppored any more. rewrite it to load_row and save_row") if ($self->{low_mem}); |
99 |
|
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"); |
105 |
|
106 |
require $module_path; |
107 |
|
108 |
$self->{'encoding'} ||= 'ISO-8859-2'; |
109 |
|
110 |
$self ? return $self : return undef; |
111 |
} |
112 |
|
113 |
=head2 open |
114 |
|
115 |
This function will read whole database in memory and produce lookups. |
116 |
|
117 |
my $store; # simple in-memory hash |
118 |
|
119 |
$input->open( |
120 |
path => '/path/to/database/file', |
121 |
code_page => 'cp852', |
122 |
limit => 500, |
123 |
offset => 6000, |
124 |
stats => 1, |
125 |
lookup_coderef => sub { |
126 |
my $rec = shift; |
127 |
# store lookups |
128 |
}, |
129 |
modify_records => { |
130 |
900 => { '^a' => { ' : ' => '^b' } }, |
131 |
901 => { '*' => { '^b' => ' ; ' } }, |
132 |
}, |
133 |
modify_file => 'conf/modify/mapping.map', |
134 |
save_row => sub { |
135 |
my $a = shift; |
136 |
$store->{ $a->{id} } = $a->{row}; |
137 |
}, |
138 |
load_row => sub { |
139 |
my $a = shift; |
140 |
return defined($store->{ $a->{id} }) && |
141 |
$store->{ $a->{id} }; |
142 |
}, |
143 |
|
144 |
); |
145 |
|
146 |
By default, C<code_page> is assumed to be C<cp852>. |
147 |
|
148 |
C<offset> is optional parametar to position at some offset before reading from database. |
149 |
|
150 |
C<limit> is optional parametar to read just C<limit> records from database |
151 |
|
152 |
C<stats> create optional report about usage of fields and subfields |
153 |
|
154 |
C<lookup_coderef> is closure to called to save data into lookups |
155 |
|
156 |
C<modify_records> specify mapping from subfields to delimiters or from |
157 |
delimiters to subfields, as well as oprations on fields (if subfield is |
158 |
defined as C<*>. |
159 |
|
160 |
C<modify_file> is alternative for C<modify_records> above which preserves order and offers |
161 |
(hopefully) simplier sintax than YAML or perl (see L</modify_file_regex>). This option |
162 |
overrides C<modify_records> if both exists for same input. |
163 |
|
164 |
C<save_row> and C<load_row> are low-level implementation of store engine. Calling convention |
165 |
is documented in example above. |
166 |
|
167 |
Returns size of database, regardless of C<offset> and C<limit> |
168 |
parametars, see also C<size>. |
169 |
|
170 |
=cut |
171 |
|
172 |
sub open { |
173 |
my $self = shift; |
174 |
my $arg = {@_}; |
175 |
|
176 |
my $log = $self->_get_logger(); |
177 |
|
178 |
$log->logconfess("lookup argument is not suppored any more. rewrite call to lookup_coderef") if ($arg->{lookup}); |
179 |
$log->logconfess("lookup_coderef must be CODE, not ",ref($arg->{lookup_coderef})) |
180 |
if ($arg->{lookup_coderef} && ref($arg->{lookup_coderef}) ne 'CODE'); |
181 |
|
182 |
$log->debug( $arg->{lookup_coderef} ? '' : 'not ', "using lookup_coderef"); |
183 |
|
184 |
$log->logcroak("need path") if (! $arg->{'path'}); |
185 |
my $code_page = $arg->{'code_page'} || 'cp852'; |
186 |
|
187 |
# store data in object |
188 |
$self->{'input_code_page'} = $code_page; |
189 |
foreach my $v (qw/path offset limit/) { |
190 |
$self->{$v} = $arg->{$v} if ($arg->{$v}); |
191 |
} |
192 |
|
193 |
if ($arg->{load_row} || $arg->{save_row}) { |
194 |
$log->logconfess("save_row and load_row must be defined in pair and be CODE") unless ( |
195 |
ref($arg->{load_row}) eq 'CODE' && |
196 |
ref($arg->{save_row}) eq 'CODE' |
197 |
); |
198 |
$self->{load_row} = $arg->{load_row}; |
199 |
$self->{save_row} = $arg->{save_row}; |
200 |
$log->debug("using load_row and save_row instead of in-memory hash"); |
201 |
} |
202 |
|
203 |
my $filter_ref; |
204 |
my $recode_regex; |
205 |
my $recode_map; |
206 |
|
207 |
if ($self->{recode}) { |
208 |
my @r = split(/\s/, $self->{recode}); |
209 |
if ($#r % 2 != 1) { |
210 |
$log->logwarn("recode needs even number of elements (some number of valid pairs)"); |
211 |
} else { |
212 |
while (@r) { |
213 |
my $from = shift @r; |
214 |
my $to = shift @r; |
215 |
$recode_map->{$from} = $to; |
216 |
} |
217 |
|
218 |
$recode_regex = join '|' => keys %{ $recode_map }; |
219 |
|
220 |
$log->debug("using recode regex: $recode_regex"); |
221 |
} |
222 |
|
223 |
} |
224 |
|
225 |
my $rec_regex; |
226 |
if (my $p = $arg->{modify_file}) { |
227 |
$log->debug("using modify_file $p"); |
228 |
$rec_regex = $self->modify_file_regexps( $p ); |
229 |
} elsif (my $h = $arg->{modify_records}) { |
230 |
$log->debug("using modify_records ", sub { dump( $h ) }); |
231 |
$rec_regex = $self->modify_record_regexps(%{ $h }); |
232 |
} |
233 |
$log->debug("rec_regex: ", sub { dump($rec_regex) }) if ($rec_regex); |
234 |
|
235 |
my $class = $self->{module} || $log->logconfess("can't get low-level module name!"); |
236 |
|
237 |
my $ll_db = $class->new( |
238 |
path => $arg->{path}, |
239 |
input_config => $arg->{input_config} || $self->{input_config}, |
240 |
# filter => sub { |
241 |
# my ($l,$f_nr) = @_; |
242 |
# return unless defined($l); |
243 |
# from_to($l, $code_page, $self->{'encoding'}); |
244 |
# $l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map); |
245 |
# return $l; |
246 |
# }, |
247 |
%{ $arg }, |
248 |
); |
249 |
|
250 |
unless (defined($ll_db)) { |
251 |
$log->logwarn("can't open database $arg->{path}, skipping..."); |
252 |
return; |
253 |
} |
254 |
|
255 |
my $size = $ll_db->size; |
256 |
|
257 |
unless ($size) { |
258 |
$log->logwarn("no records in database $arg->{path}, skipping..."); |
259 |
return; |
260 |
} |
261 |
|
262 |
my $from_rec = 1; |
263 |
my $to_rec = $size; |
264 |
|
265 |
if (my $s = $self->{offset}) { |
266 |
$log->debug("skipping to MFN $s"); |
267 |
$from_rec = $s; |
268 |
} else { |
269 |
$self->{offset} = $from_rec; |
270 |
} |
271 |
|
272 |
if ($self->{limit}) { |
273 |
$log->debug("limiting to ",$self->{limit}," records"); |
274 |
$to_rec = $from_rec + $self->{limit} - 1; |
275 |
$to_rec = $size if ($to_rec > $size); |
276 |
} |
277 |
|
278 |
# store size for later |
279 |
$self->{size} = ($to_rec - $from_rec) ? ($to_rec - $from_rec + 1) : 0; |
280 |
|
281 |
$log->info("processing $self->{size}/$size records [$from_rec-$to_rec] convert $code_page -> $self->{encoding}", $self->{stats} ? ' [stats]' : ''); |
282 |
|
283 |
# read database |
284 |
for (my $pos = $from_rec; $pos <= $to_rec; $pos++) { |
285 |
|
286 |
$log->debug("position: $pos\n"); |
287 |
|
288 |
my $rec = $ll_db->fetch_rec($pos, sub { |
289 |
my ($l,$f_nr,$debug) = @_; |
290 |
# return unless defined($l); |
291 |
# return $l unless ($rec_regex && $f_nr); |
292 |
|
293 |
return unless ( defined($l) && defined($f_nr) ); |
294 |
|
295 |
warn "-=> $f_nr ## |$l|\n" if ($debug); |
296 |
$log->debug("-=> $f_nr ## $l"); |
297 |
|
298 |
# codepage conversion and recode_regex |
299 |
from_to($l, $code_page, $self->{'encoding'}); |
300 |
$l =~ s/($recode_regex)/$recode_map->{$1}/g if ($recode_regex && $recode_map); |
301 |
|
302 |
# apply regexps |
303 |
if ($rec_regex && defined($rec_regex->{$f_nr})) { |
304 |
$log->logconfess("regexps->{$f_nr} must be ARRAY") if (ref($rec_regex->{$f_nr}) ne 'ARRAY'); |
305 |
my $c = 0; |
306 |
foreach my $r (@{ $rec_regex->{$f_nr} }) { |
307 |
my $old_l = $l; |
308 |
$log->logconfess("expected regex in ", dump( $r )) unless defined($r->{regex}); |
309 |
eval '$l =~ ' . $r->{regex}; |
310 |
if ($old_l ne $l) { |
311 |
my $d = "|$old_l| -> |$l| "; # . $r->{regex}; |
312 |
$d .= ' +' . $r->{line} . ' ' . $r->{file} if defined($r->{line}); |
313 |
$d .= ' ' . $r->{debug} if defined($r->{debug}); |
314 |
$log->debug("MODIFY $d"); |
315 |
warn "*** $d\n" if ($debug); |
316 |
|
317 |
} |
318 |
$log->error("error applying regex: $r") if ($@); |
319 |
} |
320 |
} |
321 |
|
322 |
$log->debug("<=- $f_nr ## |$l|"); |
323 |
warn "<=- $f_nr ## $l\n" if ($debug); |
324 |
return $l; |
325 |
}); |
326 |
|
327 |
$log->debug(sub { dump($rec) }); |
328 |
|
329 |
if (! $rec) { |
330 |
$log->warn("record $pos empty? skipping..."); |
331 |
next; |
332 |
} |
333 |
|
334 |
# store |
335 |
if ($self->{save_row}) { |
336 |
$self->{save_row}->({ |
337 |
id => $pos, |
338 |
row => $rec, |
339 |
}); |
340 |
} else { |
341 |
$self->{data}->{$pos} = $rec; |
342 |
} |
343 |
|
344 |
# create lookup |
345 |
$arg->{'lookup_coderef'}->( $rec ) if ($rec && $arg->{'lookup_coderef'}); |
346 |
|
347 |
# update counters for statistics |
348 |
if ($self->{stats}) { |
349 |
|
350 |
# fetch clean record with regexpes applied for statistics |
351 |
my $rec = $ll_db->fetch_rec($pos); |
352 |
|
353 |
foreach my $fld (keys %{ $rec }) { |
354 |
$self->{_stats}->{fld}->{ $fld }++; |
355 |
|
356 |
$log->logdie("invalid record fild $fld, not ARRAY") |
357 |
unless (ref($rec->{ $fld }) eq 'ARRAY'); |
358 |
|
359 |
foreach my $row (@{ $rec->{$fld} }) { |
360 |
|
361 |
if (ref($row) eq 'HASH') { |
362 |
|
363 |
foreach my $sf (keys %{ $row }) { |
364 |
next if ($sf eq 'subfields'); |
365 |
$self->{_stats}->{sf}->{ $fld }->{ $sf }->{count}++; |
366 |
$self->{_stats}->{sf}->{ $fld }->{ $sf }->{repeatable}++ |
367 |
if (ref($row->{$sf}) eq 'ARRAY'); |
368 |
} |
369 |
|
370 |
} else { |
371 |
$self->{_stats}->{repeatable}->{ $fld }++; |
372 |
} |
373 |
} |
374 |
} |
375 |
} |
376 |
|
377 |
$self->progress_bar($pos,$to_rec) unless ($self->{no_progress_bar}); |
378 |
|
379 |
} |
380 |
|
381 |
$self->{pos} = -1; |
382 |
$self->{last_pcnt} = 0; |
383 |
|
384 |
# store max mfn and return it. |
385 |
$self->{max_pos} = $to_rec; |
386 |
$log->debug("max_pos: $to_rec"); |
387 |
|
388 |
# save for dump |
389 |
$self->{ll_db} = $ll_db; |
390 |
|
391 |
return $size; |
392 |
} |
393 |
|
394 |
=head2 fetch |
395 |
|
396 |
Fetch next record from database. It will also displays progress bar. |
397 |
|
398 |
my $rec = $isis->fetch; |
399 |
|
400 |
Record from this function should probably go to C<data_structure> for |
401 |
normalisation. |
402 |
|
403 |
=cut |
404 |
|
405 |
sub fetch { |
406 |
my $self = shift; |
407 |
|
408 |
my $log = $self->_get_logger(); |
409 |
|
410 |
$log->logconfess("it seems that you didn't load database!") unless ($self->{pos}); |
411 |
|
412 |
if ($self->{pos} == -1) { |
413 |
$self->{pos} = $self->{offset}; |
414 |
} else { |
415 |
$self->{pos}++; |
416 |
} |
417 |
|
418 |
my $mfn = $self->{pos}; |
419 |
|
420 |
if ($mfn > $self->{max_pos}) { |
421 |
$self->{pos} = $self->{max_pos}; |
422 |
$log->debug("at EOF"); |
423 |
return; |
424 |
} |
425 |
|
426 |
$self->progress_bar($mfn,$self->{max_pos}) unless ($self->{no_progress_bar}); |
427 |
|
428 |
my $rec; |
429 |
|
430 |
if ($self->{load_row}) { |
431 |
$rec = $self->{load_row}->({ id => $mfn }); |
432 |
} else { |
433 |
$rec = $self->{data}->{$mfn}; |
434 |
} |
435 |
|
436 |
$rec ||= 0E0; |
437 |
} |
438 |
|
439 |
=head2 pos |
440 |
|
441 |
Returns current record number (MFN). |
442 |
|
443 |
print $isis->pos; |
444 |
|
445 |
First record in database has position 1. |
446 |
|
447 |
=cut |
448 |
|
449 |
sub pos { |
450 |
my $self = shift; |
451 |
return $self->{pos}; |
452 |
} |
453 |
|
454 |
|
455 |
=head2 size |
456 |
|
457 |
Returns number of records in database |
458 |
|
459 |
print $isis->size; |
460 |
|
461 |
Result from this function can be used to loop through all records |
462 |
|
463 |
foreach my $mfn ( 1 ... $isis->size ) { ... } |
464 |
|
465 |
because it takes into account C<offset> and C<limit>. |
466 |
|
467 |
=cut |
468 |
|
469 |
sub size { |
470 |
my $self = shift; |
471 |
return $self->{size}; |
472 |
} |
473 |
|
474 |
=head2 seek |
475 |
|
476 |
Seek to specified MFN in file. |
477 |
|
478 |
$isis->seek(42); |
479 |
|
480 |
First record in database has position 1. |
481 |
|
482 |
=cut |
483 |
|
484 |
sub seek { |
485 |
my $self = shift; |
486 |
my $pos = shift; |
487 |
|
488 |
my $log = $self->_get_logger(); |
489 |
|
490 |
$log->logconfess("called without pos") unless defined($pos); |
491 |
|
492 |
if ($pos < 1) { |
493 |
$log->warn("seek before first record"); |
494 |
$pos = 1; |
495 |
} elsif ($pos > $self->{max_pos}) { |
496 |
$log->warn("seek beyond last record"); |
497 |
$pos = $self->{max_pos}; |
498 |
} |
499 |
|
500 |
return $self->{pos} = (($pos - 1) || -1); |
501 |
} |
502 |
|
503 |
=head2 stats |
504 |
|
505 |
Dump statistics about field and subfield usage |
506 |
|
507 |
print $input->stats; |
508 |
|
509 |
=cut |
510 |
|
511 |
sub stats { |
512 |
my $self = shift; |
513 |
|
514 |
my $log = $self->_get_logger(); |
515 |
|
516 |
my $s = $self->{_stats}; |
517 |
if (! $s) { |
518 |
$log->warn("called stats, but there is no statistics collected"); |
519 |
return; |
520 |
} |
521 |
|
522 |
my $max_fld = 0; |
523 |
|
524 |
my $out = join("\n", |
525 |
map { |
526 |
my $f = $_; |
527 |
die "no field in ", dump( $s->{fld} ) unless defined( $f ); |
528 |
my $v = $s->{fld}->{$f} || die "no s->{fld}->{$f}"; |
529 |
$max_fld = $v if ($v > $max_fld); |
530 |
|
531 |
my $o = sprintf("%4s %d ~", $f, $v); |
532 |
|
533 |
if (defined($s->{sf}->{$f})) { |
534 |
my @subfields = keys %{ $s->{sf}->{$f} }; |
535 |
map { |
536 |
$o .= sprintf(" %s:%d%s", $_, |
537 |
$s->{sf}->{$f}->{$_}->{count}, |
538 |
$s->{sf}->{$f}->{$_}->{repeatable} ? '*' : '', |
539 |
); |
540 |
} ( |
541 |
# first indicators and other special subfields |
542 |
sort( grep { length($_) > 1 } @subfields ), |
543 |
# then subfileds (single char) |
544 |
sort( grep { length($_) == 1 } @subfields ), |
545 |
); |
546 |
} |
547 |
|
548 |
if (my $v_r = $s->{repeatable}->{$f}) { |
549 |
$o .= " ($v_r)" if ($v_r != $v); |
550 |
} |
551 |
|
552 |
$o; |
553 |
} sort { $a <=> $b } keys %{ $s->{fld} } |
554 |
); |
555 |
|
556 |
$log->debug( sub { dump($s) } ); |
557 |
|
558 |
return $out; |
559 |
} |
560 |
|
561 |
=head2 dump_ascii |
562 |
|
563 |
Display humanly readable dump of record |
564 |
|
565 |
=cut |
566 |
|
567 |
sub dump_ascii { |
568 |
my $self = shift; |
569 |
|
570 |
return unless $self->{ll_db}; |
571 |
|
572 |
if ($self->{ll_db}->can('dump_ascii')) { |
573 |
return $self->{ll_db}->dump_ascii( $self->{pos} ); |
574 |
} else { |
575 |
return dump( $self->{ll_db}->fetch_rec( $self->{pos} ) ); |
576 |
} |
577 |
} |
578 |
|
579 |
=head2 _get_regex |
580 |
|
581 |
Helper function called which create regexps to be execute on code. |
582 |
|
583 |
_get_regex( 900, 'regex:[0-9]+' ,'numbers' ); |
584 |
_get_regex( 900, '^b', ' : ^b' ); |
585 |
|
586 |
It supports perl regexps with C<regex:> prefix to from value and has |
587 |
additional logic to skip empty subfields. |
588 |
|
589 |
=cut |
590 |
|
591 |
sub _get_regex { |
592 |
my ($sf,$from,$to) = @_; |
593 |
|
594 |
# protect / |
595 |
$from =~ s!/!\\/!gs; |
596 |
$to =~ s!/!\\/!gs; |
597 |
|
598 |
if ($from =~ m/^regex:(.+)$/) { |
599 |
$from = $1; |
600 |
} else { |
601 |
$from = '\Q' . $from . '\E'; |
602 |
} |
603 |
if ($sf =~ /^\^/) { |
604 |
my $need_subfield_data = '*'; # no |
605 |
# if from is also subfield, require some data in between |
606 |
# to correctly skip empty subfields |
607 |
$need_subfield_data = '+' if ($from =~ m/^\\Q\^/); |
608 |
return |
609 |
's/\Q'. $sf .'\E([^\^]' . $need_subfield_data . '?)'. $from .'([^\^]*?)/'. $sf .'$1'. $to .'$2/'; |
610 |
} else { |
611 |
return |
612 |
's/'. $from .'/'. $to .'/g'; |
613 |
} |
614 |
} |
615 |
|
616 |
|
617 |
=head2 modify_record_regexps |
618 |
|
619 |
Generate hash with regexpes to be applied using L<filter>. |
620 |
|
621 |
my $regexpes = $input->modify_record_regexps( |
622 |
900 => { '^a' => { ' : ' => '^b' } }, |
623 |
901 => { '*' => { '^b' => ' ; ' } }, |
624 |
); |
625 |
|
626 |
=cut |
627 |
|
628 |
sub modify_record_regexps { |
629 |
my $self = shift; |
630 |
my $modify_record = {@_}; |
631 |
|
632 |
my $regexpes; |
633 |
|
634 |
my $log = $self->_get_logger(); |
635 |
|
636 |
foreach my $f (keys %$modify_record) { |
637 |
$log->debug("field: $f"); |
638 |
|
639 |
foreach my $sf (keys %{ $modify_record->{$f} }) { |
640 |
$log->debug("subfield: $sf"); |
641 |
|
642 |
foreach my $from (keys %{ $modify_record->{$f}->{$sf} }) { |
643 |
my $to = $modify_record->{$f}->{$sf}->{$from}; |
644 |
#die "no field?" unless defined($to); |
645 |
my $d = "|$from| -> |$to|"; |
646 |
$log->debug("transform: $d"); |
647 |
|
648 |
my $regex = _get_regex($sf,$from,$to); |
649 |
push @{ $regexpes->{$f} }, { regex => $regex, debug => $d }; |
650 |
$log->debug("regex: $regex"); |
651 |
} |
652 |
} |
653 |
} |
654 |
|
655 |
return $regexpes; |
656 |
} |
657 |
|
658 |
=head2 modify_file_regexps |
659 |
|
660 |
Generate hash with regexpes to be applied using L<filter> from |
661 |
pseudo hash/yaml format for regex mappings. |
662 |
|
663 |
It should be obvious: |
664 |
|
665 |
200 |
666 |
'^a' |
667 |
' : ' => '^e' |
668 |
' = ' => '^d' |
669 |
|
670 |
In field I<200> find C<'^a'> and then C<' : '>, and replace it with C<'^e'>. |
671 |
In field I<200> find C<'^a'> and then C<' = '>, and replace it with C<'^d'>. |
672 |
|
673 |
my $regexpes = $input->modify_file_regexps( 'conf/modify/common.pl' ); |
674 |
|
675 |
On undef path it will just return. |
676 |
|
677 |
=cut |
678 |
|
679 |
sub modify_file_regexps { |
680 |
my $self = shift; |
681 |
|
682 |
my $modify_path = shift || return; |
683 |
|
684 |
my $log = $self->_get_logger(); |
685 |
|
686 |
my $regexpes; |
687 |
|
688 |
CORE::open(my $fh, $modify_path) || $log->logdie("can't open modify file $modify_path: $!"); |
689 |
|
690 |
my ($f,$sf); |
691 |
|
692 |
while(<$fh>) { |
693 |
chomp; |
694 |
next if (/^#/ || /^\s*$/); |
695 |
|
696 |
if (/^\s*(\d+)\s*$/) { |
697 |
$f = $1; |
698 |
$log->debug("field: $f"); |
699 |
next; |
700 |
} elsif (/^\s*'([^']*)'\s*$/) { |
701 |
$sf = $1; |
702 |
$log->die("can't define subfiled before field in: $_") unless ($f); |
703 |
$log->debug("subfield: $sf"); |
704 |
} elsif (/^\s*'([^']*)'\s*=>\s*'([^']*)'\s*$/) { |
705 |
my ($from,$to) = ($1, $2); |
706 |
|
707 |
$log->debug("transform: |$from| -> |$to|"); |
708 |
|
709 |
my $regex = _get_regex($sf,$from,$to); |
710 |
push @{ $regexpes->{$f} }, { |
711 |
regex => $regex, |
712 |
file => $modify_path, |
713 |
line => $., |
714 |
}; |
715 |
$log->debug("regex: $regex"); |
716 |
} |
717 |
} |
718 |
|
719 |
return $regexpes; |
720 |
} |
721 |
|
722 |
=head1 AUTHOR |
723 |
|
724 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
725 |
|
726 |
=head1 COPYRIGHT & LICENSE |
727 |
|
728 |
Copyright 2005-2006 Dobrica Pavlinusic, All Rights Reserved. |
729 |
|
730 |
This program is free software; you can redistribute it and/or modify it |
731 |
under the same terms as Perl itself. |
732 |
|
733 |
=cut |
734 |
|
735 |
1; # End of WebPAC::Input |