1 |
dpavlin |
514 |
package WebPAC::Validate; |
2 |
|
|
|
3 |
|
|
use warnings; |
4 |
|
|
use strict; |
5 |
|
|
|
6 |
|
|
use blib; |
7 |
|
|
|
8 |
|
|
use base 'WebPAC::Common'; |
9 |
|
|
use File::Slurp; |
10 |
dpavlin |
515 |
use List::Util qw/first/; |
11 |
dpavlin |
652 |
use Data::Dump qw/dump/; |
12 |
dpavlin |
643 |
use WebPAC::Normalize qw/_pack_subfields_hash/; |
13 |
|
|
use Storable qw/dclone/; |
14 |
dpavlin |
514 |
|
15 |
|
|
=head1 NAME |
16 |
|
|
|
17 |
|
|
WebPAC::Validate - provide simple validation for records |
18 |
|
|
|
19 |
|
|
=head1 VERSION |
20 |
|
|
|
21 |
dpavlin |
827 |
Version 0.11 |
22 |
dpavlin |
514 |
|
23 |
|
|
=cut |
24 |
|
|
|
25 |
dpavlin |
827 |
our $VERSION = '0.11'; |
26 |
dpavlin |
514 |
|
27 |
|
|
=head1 SYNOPSIS |
28 |
|
|
|
29 |
|
|
This module provide a simple way to validate your file against a simple |
30 |
|
|
configuration file in following format: |
31 |
|
|
|
32 |
|
|
# field 10 doesn't have any subfields |
33 |
|
|
10 |
34 |
|
|
# same with 101 |
35 |
|
|
101 |
36 |
|
|
# field 200 have valid subfields a-g |
37 |
dpavlin |
645 |
# and field e is repeatable |
38 |
|
|
200 a b c d e* f g |
39 |
dpavlin |
514 |
# field 205 can have only subfield a |
40 |
dpavlin |
645 |
# and must exists |
41 |
|
|
205! a |
42 |
dpavlin |
514 |
# while 210 can have a c or d |
43 |
|
|
210 a c d |
44 |
dpavlin |
677 |
# field which is ignored in validation |
45 |
|
|
999- |
46 |
dpavlin |
514 |
|
47 |
|
|
=head1 FUNCTIONS |
48 |
|
|
|
49 |
|
|
=head2 new |
50 |
|
|
|
51 |
|
|
Create new validation object |
52 |
|
|
|
53 |
|
|
my $validate = new WebPAC::Validate( |
54 |
dpavlin |
643 |
path => 'conf/validate/file', |
55 |
dpavlin |
827 |
delimiters => [ ' : ', ' / ', ' ; ', ' , ' ], |
56 |
dpavlin |
514 |
); |
57 |
|
|
|
58 |
dpavlin |
827 |
Optional parametar C<delimiters> will turn on validating of delimiters. Be |
59 |
|
|
careful here, those delimiters are just stuck into regex, so they can |
60 |
|
|
contain L<perlre> regexpes. |
61 |
|
|
|
62 |
dpavlin |
514 |
=cut |
63 |
|
|
|
64 |
|
|
sub new { |
65 |
|
|
my $class = shift; |
66 |
|
|
my $self = {@_}; |
67 |
|
|
bless($self, $class); |
68 |
|
|
|
69 |
|
|
my $log = $self->_get_logger(); |
70 |
|
|
|
71 |
|
|
foreach my $p (qw/path/) { |
72 |
|
|
$log->logconfess("need $p") unless ($self->{$p}); |
73 |
|
|
} |
74 |
|
|
|
75 |
|
|
my $v_file = read_file( $self->{path} ) || |
76 |
|
|
$log->logdie("can't open validate path $self->{path}: $!"); |
77 |
|
|
|
78 |
|
|
my $v; |
79 |
|
|
my $curr_line = 1; |
80 |
|
|
|
81 |
|
|
foreach my $l (split(/[\n\r]+/, $v_file)) { |
82 |
|
|
$curr_line++; |
83 |
|
|
|
84 |
dpavlin |
516 |
# skip comments and whitespaces |
85 |
|
|
next if ($l =~ /^#/ || $l =~ /^\s*$/); |
86 |
|
|
|
87 |
dpavlin |
514 |
$l =~ s/^\s+//; |
88 |
|
|
$l =~ s/\s+$//; |
89 |
|
|
|
90 |
|
|
my @d = split(/\s+/, $l); |
91 |
|
|
|
92 |
dpavlin |
516 |
my $fld = shift @d; |
93 |
dpavlin |
514 |
|
94 |
dpavlin |
645 |
if ($fld =~ s/!$//) { |
95 |
|
|
$self->{must_exist}->{$fld}++; |
96 |
dpavlin |
677 |
} elsif ($fld =~ s/-$//) { |
97 |
|
|
$self->{dont_validate}->{$fld}++; |
98 |
dpavlin |
645 |
} |
99 |
|
|
|
100 |
dpavlin |
516 |
$log->logdie("need field name in line $curr_line: $l") unless (defined($fld)); |
101 |
|
|
|
102 |
dpavlin |
514 |
if (@d) { |
103 |
dpavlin |
646 |
$v->{$fld} = [ map { |
104 |
|
|
my $sf = $_; |
105 |
|
|
if ( $sf =~ s/!(\*)?$/$1/ ) { |
106 |
|
|
$self->{must_exist_sf}->{ $fld }->{ $sf }++; |
107 |
|
|
}; |
108 |
|
|
$sf; |
109 |
|
|
} @d ]; |
110 |
dpavlin |
514 |
} else { |
111 |
dpavlin |
515 |
$v->{$fld} = 1; |
112 |
dpavlin |
514 |
} |
113 |
|
|
|
114 |
|
|
} |
115 |
|
|
|
116 |
dpavlin |
652 |
$log->debug("current validation rules: ", dump($v)); |
117 |
dpavlin |
514 |
|
118 |
|
|
$self->{rules} = $v; |
119 |
|
|
|
120 |
dpavlin |
516 |
$log->info("validation uses rules from $self->{path}"); |
121 |
|
|
|
122 |
dpavlin |
827 |
if ( $self->{delimiters} ) { |
123 |
|
|
$self->{delimiters_regex} = '(\^[a-z0-9]|' . join('|', @{ $self->{delimiters} }) . ')'; |
124 |
|
|
$log->info("validation check delimiters with regex $self->{delimiters_regex}"); |
125 |
|
|
} |
126 |
|
|
|
127 |
dpavlin |
514 |
$self ? return $self : return undef; |
128 |
|
|
} |
129 |
|
|
|
130 |
dpavlin |
768 |
=head2 validate_rec |
131 |
dpavlin |
515 |
|
132 |
|
|
Validate record and return errors |
133 |
|
|
|
134 |
dpavlin |
768 |
my @errors = $validate->validate_rec( $rec, $rec_dump ); |
135 |
dpavlin |
515 |
|
136 |
|
|
=cut |
137 |
|
|
|
138 |
dpavlin |
768 |
sub validate_rec { |
139 |
dpavlin |
515 |
my $self = shift; |
140 |
|
|
|
141 |
|
|
my $log = $self->_get_logger(); |
142 |
|
|
|
143 |
dpavlin |
768 |
my $rec = shift || $log->logdie("validate_rec need record"); |
144 |
dpavlin |
653 |
my $rec_dump = shift; |
145 |
dpavlin |
515 |
|
146 |
|
|
$log->logdie("rec isn't HASH") unless (ref($rec) eq 'HASH'); |
147 |
|
|
$log->logdie("can't find validation rules") unless (my $r = $self->{rules}); |
148 |
|
|
|
149 |
dpavlin |
652 |
my $errors; |
150 |
dpavlin |
515 |
|
151 |
dpavlin |
652 |
$log->debug("rec = ", sub { dump($rec) }, "keys = ", keys %{ $rec }); |
152 |
dpavlin |
515 |
|
153 |
dpavlin |
645 |
my $fields; |
154 |
|
|
|
155 |
dpavlin |
515 |
foreach my $f (keys %{ $rec }) { |
156 |
|
|
|
157 |
dpavlin |
516 |
next if (!defined($f) || $f eq '' || $f eq '000'); |
158 |
dpavlin |
515 |
|
159 |
dpavlin |
827 |
# first check delimiters |
160 |
|
|
if ( my $regex = $self->{delimiters_regex} ) { |
161 |
|
|
|
162 |
|
|
foreach my $v (@{ $rec->{$f} }) { |
163 |
|
|
my $l = _pack_subfields_hash( $v, 1 ); |
164 |
dpavlin |
828 |
my $template = ''; |
165 |
dpavlin |
830 |
$l =~ s/$regex/$template.=$1/eg; |
166 |
|
|
# warn "## template: $template\n"; |
167 |
dpavlin |
828 |
$self->{_delimiters_templates}->{$f}->{$template}++ if ( $template ); |
168 |
dpavlin |
827 |
} |
169 |
|
|
|
170 |
|
|
} |
171 |
|
|
|
172 |
dpavlin |
677 |
next if (defined( $self->{dont_validate}->{$f} )); |
173 |
|
|
|
174 |
|
|
# track field usage |
175 |
dpavlin |
645 |
$fields->{$f}++; |
176 |
|
|
|
177 |
|
|
if ( ! defined($r->{$f}) ) { |
178 |
dpavlin |
666 |
$errors->{ $f }->{unexpected} = "this field is not expected"; |
179 |
dpavlin |
515 |
next; |
180 |
|
|
} |
181 |
|
|
|
182 |
dpavlin |
645 |
|
183 |
dpavlin |
515 |
if (ref($rec->{$f}) ne 'ARRAY') { |
184 |
dpavlin |
666 |
$errors->{ $f }->{not_repeatable} = "probably bug in parsing input data"; |
185 |
dpavlin |
515 |
next; |
186 |
|
|
} |
187 |
|
|
|
188 |
|
|
foreach my $v (@{ $rec->{$f} }) { |
189 |
|
|
# can we have subfields? |
190 |
|
|
if (ref($r->{$f}) eq 'ARRAY') { |
191 |
|
|
# are values hashes? (has subfields) |
192 |
dpavlin |
652 |
if (! defined($v)) { |
193 |
dpavlin |
666 |
# $errors->{$f}->{empty} = undef; |
194 |
dpavlin |
656 |
# $errors->{dump} = $rec_dump if ($rec_dump); |
195 |
dpavlin |
652 |
} elsif (ref($v) ne 'HASH') { |
196 |
dpavlin |
666 |
$errors->{$f}->{missing_subfield} = join(",", @{ $r->{$f} }) . " required"; |
197 |
dpavlin |
515 |
next; |
198 |
|
|
} else { |
199 |
dpavlin |
643 |
|
200 |
|
|
my $h = dclone( $v ); |
201 |
|
|
|
202 |
dpavlin |
645 |
my $sf_repeatable; |
203 |
|
|
|
204 |
dpavlin |
643 |
delete($v->{subfields}) if (defined($v->{subfields})); |
205 |
|
|
|
206 |
dpavlin |
646 |
my $subfields; |
207 |
|
|
|
208 |
dpavlin |
515 |
foreach my $sf (keys %{ $v }) { |
209 |
dpavlin |
643 |
|
210 |
dpavlin |
646 |
$subfields->{ $sf }++; |
211 |
|
|
|
212 |
dpavlin |
645 |
# is non-repeatable but with multiple values? |
213 |
|
|
if ( ! first { $_ eq $sf.'*' } @{$r->{$f}} ) { |
214 |
|
|
if ( ref($v->{$sf}) eq 'ARRAY' ) { |
215 |
|
|
$sf_repeatable->{$sf}++; |
216 |
|
|
}; |
217 |
|
|
if (! first { $_ eq $sf } @{ $r->{$f} }) { |
218 |
dpavlin |
666 |
$errors->{ $f }->{subfield}->{extra}->{$sf}++; |
219 |
dpavlin |
645 |
} |
220 |
dpavlin |
515 |
} |
221 |
dpavlin |
643 |
|
222 |
dpavlin |
515 |
} |
223 |
dpavlin |
645 |
if (my @r_sf = sort keys( %$sf_repeatable )) { |
224 |
|
|
|
225 |
dpavlin |
652 |
foreach my $sf (@r_sf) { |
226 |
dpavlin |
666 |
$errors->{$f}->{subfield}->{extra_repeatable}->{$sf}++; |
227 |
dpavlin |
670 |
$errors->{$f}->{dump} = _pack_subfields_hash( $h, 1 ); |
228 |
dpavlin |
652 |
} |
229 |
|
|
|
230 |
dpavlin |
645 |
} |
231 |
dpavlin |
646 |
|
232 |
|
|
if ( defined( $self->{must_exist_sf}->{$f} ) ) { |
233 |
|
|
foreach my $sf (sort keys %{ $self->{must_exist_sf}->{$f} }) { |
234 |
dpavlin |
647 |
#warn "====> $f $sf must exist\n"; |
235 |
dpavlin |
666 |
$errors->{$f}->{subfield}->{missing}->{$sf}++ |
236 |
dpavlin |
652 |
unless defined( $subfields->{$sf} ); |
237 |
dpavlin |
646 |
} |
238 |
|
|
} |
239 |
|
|
|
240 |
dpavlin |
515 |
} |
241 |
|
|
} elsif (ref($v) eq 'HASH') { |
242 |
dpavlin |
666 |
$errors->{$f}->{unexpected_subfields}++; |
243 |
dpavlin |
670 |
$errors->{$f}->{dump} = _pack_subfields_hash( $v, 1 ); |
244 |
dpavlin |
515 |
} |
245 |
|
|
} |
246 |
|
|
} |
247 |
|
|
|
248 |
dpavlin |
827 |
$log->debug("_delimiters_templates = ", dump( $self->{_delimiters_templates} ) ); |
249 |
|
|
|
250 |
dpavlin |
645 |
foreach my $must (sort keys %{ $self->{must_exist} }) { |
251 |
|
|
next if ($fields->{$must}); |
252 |
dpavlin |
666 |
$errors->{$must}->{missing}++; |
253 |
dpavlin |
653 |
$errors->{dump} = $rec_dump if ($rec_dump); |
254 |
dpavlin |
645 |
} |
255 |
|
|
|
256 |
dpavlin |
652 |
if ($errors) { |
257 |
dpavlin |
671 |
$log->debug("errors: ", $self->report_error( $errors ) ); |
258 |
dpavlin |
652 |
|
259 |
dpavlin |
653 |
my $mfn = $rec->{'000'}->[0] || $log->logconfess("record ", dump( $rec ), " doesn't have MFN"); |
260 |
dpavlin |
652 |
$self->{errors}->{$mfn} = $errors; |
261 |
|
|
} |
262 |
|
|
|
263 |
dpavlin |
515 |
#$log->logcluck("return from this function is ARRAY") unless wantarray; |
264 |
|
|
|
265 |
dpavlin |
652 |
return $errors; |
266 |
dpavlin |
515 |
} |
267 |
|
|
|
268 |
dpavlin |
653 |
=head2 reset_errors |
269 |
|
|
|
270 |
|
|
Clean all accumulated errors for this input |
271 |
|
|
|
272 |
|
|
$validate->reset_errors; |
273 |
|
|
|
274 |
|
|
=cut |
275 |
|
|
|
276 |
|
|
sub reset_errors { |
277 |
|
|
my $self = shift; |
278 |
|
|
delete ($self->{errors}); |
279 |
|
|
} |
280 |
|
|
|
281 |
|
|
=head2 all_errors |
282 |
|
|
|
283 |
|
|
Return hash with all errors |
284 |
|
|
|
285 |
|
|
print dump( $validate->all_errors ); |
286 |
|
|
|
287 |
|
|
=cut |
288 |
|
|
|
289 |
|
|
sub all_errors { |
290 |
|
|
my $self = shift; |
291 |
|
|
return $self->{errors}; |
292 |
|
|
} |
293 |
|
|
|
294 |
dpavlin |
671 |
=head2 report_error |
295 |
dpavlin |
664 |
|
296 |
dpavlin |
671 |
Produce nice humanly readable report of single error |
297 |
dpavlin |
664 |
|
298 |
dpavlin |
671 |
print $validate->report_error( $error_hash ); |
299 |
dpavlin |
664 |
|
300 |
|
|
=cut |
301 |
|
|
|
302 |
dpavlin |
671 |
sub report_error { |
303 |
dpavlin |
664 |
my $self = shift; |
304 |
|
|
|
305 |
dpavlin |
671 |
my $h = shift || die "no hash?"; |
306 |
|
|
|
307 |
|
|
sub _unroll { |
308 |
dpavlin |
666 |
my ($self, $tree, $accumulated) = @_; |
309 |
dpavlin |
664 |
|
310 |
dpavlin |
666 |
my $log = $self->_get_logger(); |
311 |
|
|
|
312 |
dpavlin |
665 |
$log->debug("# ", |
313 |
|
|
( $tree ? "tree: $tree " : '' ), |
314 |
|
|
( $accumulated ? "accumulated: $accumulated " : '' ), |
315 |
|
|
); |
316 |
dpavlin |
664 |
|
317 |
dpavlin |
665 |
my $results; |
318 |
dpavlin |
664 |
|
319 |
dpavlin |
665 |
if (ref($tree) ne 'HASH') { |
320 |
|
|
return ("$accumulated\t($tree)", undef); |
321 |
dpavlin |
664 |
} |
322 |
|
|
|
323 |
dpavlin |
665 |
my $dump; |
324 |
dpavlin |
664 |
|
325 |
dpavlin |
665 |
foreach my $k (sort keys %{ $tree }) { |
326 |
|
|
|
327 |
dpavlin |
664 |
if ($k eq 'dump') { |
328 |
dpavlin |
665 |
$dump = $tree->{dump}; |
329 |
dpavlin |
670 |
# warn "## dump: ",dump($dump),"\n"; |
330 |
dpavlin |
664 |
next; |
331 |
|
|
} |
332 |
|
|
|
333 |
dpavlin |
665 |
$log->debug("current: $k"); |
334 |
dpavlin |
664 |
|
335 |
dpavlin |
671 |
my ($new_results, $new_dump) = $self->_unroll($tree->{$k}, |
336 |
dpavlin |
665 |
$accumulated ? "$accumulated\t$k" : $k |
337 |
|
|
); |
338 |
|
|
|
339 |
|
|
$log->debug( |
340 |
|
|
( $new_results ? "new_results: " . dump($new_results) ." " : '' ), |
341 |
|
|
); |
342 |
|
|
|
343 |
|
|
push @$results, $new_results if ($new_results); |
344 |
|
|
$dump = $new_dump if ($new_dump); |
345 |
|
|
|
346 |
dpavlin |
664 |
} |
347 |
dpavlin |
665 |
|
348 |
|
|
$log->debug( |
349 |
|
|
( $results ? "results: " . dump($results) ." " : '' ), |
350 |
|
|
); |
351 |
|
|
|
352 |
|
|
if ($#$results == 0) { |
353 |
|
|
return ($results->[0], $dump); |
354 |
|
|
} else { |
355 |
|
|
return ($results, $dump); |
356 |
|
|
} |
357 |
dpavlin |
664 |
} |
358 |
|
|
|
359 |
dpavlin |
666 |
|
360 |
dpavlin |
671 |
sub _reformat { |
361 |
dpavlin |
666 |
my $l = shift; |
362 |
|
|
$l =~ s/\t/ /g; |
363 |
|
|
$l =~ s/_/ /; |
364 |
|
|
return $l; |
365 |
|
|
} |
366 |
|
|
|
367 |
dpavlin |
671 |
my $out = ''; |
368 |
dpavlin |
666 |
|
369 |
dpavlin |
671 |
for my $f (sort keys %{ $h }) { |
370 |
|
|
$out .= "$f: "; |
371 |
|
|
|
372 |
|
|
my ($r, $d) = $self->_unroll( $h->{$f} ); |
373 |
|
|
my $e; |
374 |
|
|
if (ref($r) eq 'ARRAY') { |
375 |
|
|
$e .= join(", ", map { _reformat( $_ ) } @$r); |
376 |
|
|
} else { |
377 |
|
|
$e .= _reformat( $r ); |
378 |
dpavlin |
666 |
} |
379 |
dpavlin |
671 |
$e .= "\n\t$d" if ($d); |
380 |
|
|
|
381 |
|
|
$out .= $e . "\n"; |
382 |
dpavlin |
664 |
} |
383 |
dpavlin |
671 |
return $out; |
384 |
|
|
} |
385 |
dpavlin |
664 |
|
386 |
dpavlin |
671 |
|
387 |
|
|
=head2 report |
388 |
|
|
|
389 |
|
|
Produce nice humanly readable report of errors |
390 |
|
|
|
391 |
|
|
print $validate->report; |
392 |
|
|
|
393 |
|
|
=cut |
394 |
|
|
|
395 |
|
|
sub report { |
396 |
|
|
my $self = shift; |
397 |
|
|
my $e = $self->{errors} || return; |
398 |
|
|
|
399 |
|
|
my $out; |
400 |
|
|
foreach my $mfn (sort { $a <=> $b } keys %$e) { |
401 |
|
|
$out .= "MFN $mfn\n" . $self->report_error( $e->{$mfn} ) . "\n"; |
402 |
|
|
} |
403 |
|
|
|
404 |
dpavlin |
664 |
return $out; |
405 |
dpavlin |
671 |
|
406 |
dpavlin |
664 |
} |
407 |
|
|
|
408 |
dpavlin |
832 |
=head2 delimiters_templates |
409 |
|
|
|
410 |
dpavlin |
834 |
Generate report of delimiter tamplates |
411 |
dpavlin |
832 |
|
412 |
dpavlin |
834 |
my $report = $validate->delimiter_teplates( |
413 |
|
|
report => 1, |
414 |
|
|
); |
415 |
|
|
|
416 |
|
|
Options: |
417 |
|
|
|
418 |
|
|
=over 4 |
419 |
|
|
|
420 |
|
|
=item report |
421 |
|
|
|
422 |
|
|
Generate humanly readable report with single fields |
423 |
|
|
|
424 |
|
|
=back |
425 |
|
|
|
426 |
dpavlin |
832 |
=cut |
427 |
|
|
|
428 |
|
|
sub delimiters_templates { |
429 |
|
|
my $self = shift; |
430 |
|
|
|
431 |
dpavlin |
834 |
my $args = {@_}; |
432 |
|
|
|
433 |
dpavlin |
832 |
my $t = $self->{_delimiters_templates}; |
434 |
|
|
|
435 |
|
|
my $log = $self->_get_logger; |
436 |
|
|
|
437 |
|
|
unless ($t) { |
438 |
dpavlin |
834 |
$log->error("called without delimiters"); |
439 |
dpavlin |
832 |
return; |
440 |
|
|
} |
441 |
|
|
|
442 |
|
|
my $out; |
443 |
|
|
|
444 |
|
|
foreach my $f (sort { $a <=> $b } keys %$t) { |
445 |
dpavlin |
834 |
$out .= "$f\n" if ( $args->{report} ); |
446 |
|
|
foreach my $template (sort { $a cmp $b } keys %{ $t->{$f} }) { |
447 |
|
|
my $count = $t->{$f}->{$template}; |
448 |
|
|
$out .= |
449 |
|
|
( $count ? "" : "# " ) . |
450 |
|
|
( $args->{report} ? "" : "$f\t" ) . |
451 |
|
|
"\t$count\t$template\n"; |
452 |
dpavlin |
832 |
} |
453 |
|
|
} |
454 |
|
|
|
455 |
|
|
return $out; |
456 |
|
|
} |
457 |
|
|
|
458 |
dpavlin |
514 |
=head1 AUTHOR |
459 |
|
|
|
460 |
|
|
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
461 |
|
|
|
462 |
|
|
=head1 COPYRIGHT & LICENSE |
463 |
|
|
|
464 |
|
|
Copyright 2006 Dobrica Pavlinusic, All Rights Reserved. |
465 |
|
|
|
466 |
|
|
This program is free software; you can redistribute it and/or modify it |
467 |
|
|
under the same terms as Perl itself. |
468 |
|
|
|
469 |
|
|
=cut |
470 |
|
|
|
471 |
|
|
1; # End of WebPAC::Validate |