1 |
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 |
use List::Util qw/first/; |
11 |
use Data::Dump qw/dump/; |
12 |
use WebPAC::Normalize qw/_pack_subfields_hash/; |
13 |
use Storable qw/dclone/; |
14 |
|
15 |
=head1 NAME |
16 |
|
17 |
WebPAC::Validate - provide simple validation for records |
18 |
|
19 |
=head1 VERSION |
20 |
|
21 |
Version 0.11 |
22 |
|
23 |
=cut |
24 |
|
25 |
our $VERSION = '0.11'; |
26 |
|
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 |
# and field e is repeatable |
38 |
200 a b c d e* f g |
39 |
# field 205 can have only subfield a |
40 |
# and must exists |
41 |
205! a |
42 |
# while 210 can have a c or d |
43 |
210 a c d |
44 |
# field which is ignored in validation |
45 |
999- |
46 |
|
47 |
=head1 FUNCTIONS |
48 |
|
49 |
=head2 new |
50 |
|
51 |
Create new validation object |
52 |
|
53 |
my $validate = new WebPAC::Validate( |
54 |
path => 'conf/validate/file', |
55 |
delimiters => [ ' : ', ' / ', ' ; ', ' , ' ], |
56 |
); |
57 |
|
58 |
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 |
=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 |
$curr_line++; |
83 |
|
84 |
# skip comments and whitespaces |
85 |
next if ($l =~ /^#/ || $l =~ /^\s*$/); |
86 |
|
87 |
$l =~ s/^\s+//; |
88 |
$l =~ s/\s+$//; |
89 |
|
90 |
my @d = split(/\s+/, $l); |
91 |
|
92 |
my $fld = shift @d; |
93 |
|
94 |
if ($fld =~ s/!$//) { |
95 |
$self->{must_exist}->{$fld}++; |
96 |
} elsif ($fld =~ s/-$//) { |
97 |
$self->{dont_validate}->{$fld}++; |
98 |
} |
99 |
|
100 |
$log->logdie("need field name in line $curr_line: $l") unless (defined($fld)); |
101 |
|
102 |
if (@d) { |
103 |
$v->{$fld} = [ map { |
104 |
my $sf = $_; |
105 |
if ( $sf =~ s/!(\*)?$/$1/ ) { |
106 |
$self->{must_exist_sf}->{ $fld }->{ $sf }++; |
107 |
}; |
108 |
$sf; |
109 |
} @d ]; |
110 |
} else { |
111 |
$v->{$fld} = 1; |
112 |
} |
113 |
|
114 |
} |
115 |
|
116 |
$log->debug("current validation rules: ", dump($v)); |
117 |
|
118 |
$self->{rules} = $v; |
119 |
|
120 |
$log->info("validation uses rules from $self->{path}"); |
121 |
|
122 |
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 |
$self ? return $self : return undef; |
128 |
} |
129 |
|
130 |
=head2 validate_rec |
131 |
|
132 |
Validate record and return errors |
133 |
|
134 |
my @errors = $validate->validate_rec( $rec, $rec_dump ); |
135 |
|
136 |
=cut |
137 |
|
138 |
sub validate_rec { |
139 |
my $self = shift; |
140 |
|
141 |
my $log = $self->_get_logger(); |
142 |
|
143 |
my $rec = shift || $log->logdie("validate_rec need record"); |
144 |
my $rec_dump = shift; |
145 |
|
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 |
my $errors; |
150 |
|
151 |
$log->debug("rec = ", sub { dump($rec) }, "keys = ", keys %{ $rec }); |
152 |
|
153 |
my $fields; |
154 |
|
155 |
foreach my $f (keys %{ $rec }) { |
156 |
|
157 |
next if (!defined($f) || $f eq '' || $f eq '000'); |
158 |
|
159 |
# first check delimiters |
160 |
if ( my $regex = $self->{delimiters_regex} ) { |
161 |
|
162 |
my $template = ''; |
163 |
|
164 |
foreach my $v (@{ $rec->{$f} }) { |
165 |
warn "## v = ", dump( $v ); |
166 |
my $l = _pack_subfields_hash( $v, 1 ); |
167 |
warn "## $l [$regex]\n"; |
168 |
$l =~ s/$regex/$template.=$1/eg && warn "## new: $l\n"; |
169 |
warn "## template: $template\n"; |
170 |
$self->{_delimiters_templates}->{$f}->{$template}++; |
171 |
} |
172 |
|
173 |
} |
174 |
|
175 |
next if (defined( $self->{dont_validate}->{$f} )); |
176 |
|
177 |
# track field usage |
178 |
$fields->{$f}++; |
179 |
|
180 |
if ( ! defined($r->{$f}) ) { |
181 |
$errors->{ $f }->{unexpected} = "this field is not expected"; |
182 |
next; |
183 |
} |
184 |
|
185 |
|
186 |
if (ref($rec->{$f}) ne 'ARRAY') { |
187 |
$errors->{ $f }->{not_repeatable} = "probably bug in parsing input data"; |
188 |
next; |
189 |
} |
190 |
|
191 |
foreach my $v (@{ $rec->{$f} }) { |
192 |
# can we have subfields? |
193 |
if (ref($r->{$f}) eq 'ARRAY') { |
194 |
# are values hashes? (has subfields) |
195 |
if (! defined($v)) { |
196 |
# $errors->{$f}->{empty} = undef; |
197 |
# $errors->{dump} = $rec_dump if ($rec_dump); |
198 |
} elsif (ref($v) ne 'HASH') { |
199 |
$errors->{$f}->{missing_subfield} = join(",", @{ $r->{$f} }) . " required"; |
200 |
next; |
201 |
} else { |
202 |
|
203 |
my $h = dclone( $v ); |
204 |
|
205 |
my $sf_repeatable; |
206 |
|
207 |
delete($v->{subfields}) if (defined($v->{subfields})); |
208 |
|
209 |
my $subfields; |
210 |
|
211 |
foreach my $sf (keys %{ $v }) { |
212 |
|
213 |
$subfields->{ $sf }++; |
214 |
|
215 |
# is non-repeatable but with multiple values? |
216 |
if ( ! first { $_ eq $sf.'*' } @{$r->{$f}} ) { |
217 |
if ( ref($v->{$sf}) eq 'ARRAY' ) { |
218 |
$sf_repeatable->{$sf}++; |
219 |
}; |
220 |
if (! first { $_ eq $sf } @{ $r->{$f} }) { |
221 |
$errors->{ $f }->{subfield}->{extra}->{$sf}++; |
222 |
} |
223 |
} |
224 |
|
225 |
} |
226 |
if (my @r_sf = sort keys( %$sf_repeatable )) { |
227 |
|
228 |
foreach my $sf (@r_sf) { |
229 |
$errors->{$f}->{subfield}->{extra_repeatable}->{$sf}++; |
230 |
$errors->{$f}->{dump} = _pack_subfields_hash( $h, 1 ); |
231 |
} |
232 |
|
233 |
} |
234 |
|
235 |
if ( defined( $self->{must_exist_sf}->{$f} ) ) { |
236 |
foreach my $sf (sort keys %{ $self->{must_exist_sf}->{$f} }) { |
237 |
#warn "====> $f $sf must exist\n"; |
238 |
$errors->{$f}->{subfield}->{missing}->{$sf}++ |
239 |
unless defined( $subfields->{$sf} ); |
240 |
} |
241 |
} |
242 |
|
243 |
} |
244 |
} elsif (ref($v) eq 'HASH') { |
245 |
$errors->{$f}->{unexpected_subfields}++; |
246 |
$errors->{$f}->{dump} = _pack_subfields_hash( $v, 1 ); |
247 |
} |
248 |
} |
249 |
} |
250 |
|
251 |
$log->debug("_delimiters_templates = ", dump( $self->{_delimiters_templates} ) ); |
252 |
|
253 |
foreach my $must (sort keys %{ $self->{must_exist} }) { |
254 |
next if ($fields->{$must}); |
255 |
$errors->{$must}->{missing}++; |
256 |
$errors->{dump} = $rec_dump if ($rec_dump); |
257 |
} |
258 |
|
259 |
if ($errors) { |
260 |
$log->debug("errors: ", $self->report_error( $errors ) ); |
261 |
|
262 |
my $mfn = $rec->{'000'}->[0] || $log->logconfess("record ", dump( $rec ), " doesn't have MFN"); |
263 |
$self->{errors}->{$mfn} = $errors; |
264 |
} |
265 |
|
266 |
#$log->logcluck("return from this function is ARRAY") unless wantarray; |
267 |
|
268 |
return $errors; |
269 |
} |
270 |
|
271 |
=head2 reset_errors |
272 |
|
273 |
Clean all accumulated errors for this input |
274 |
|
275 |
$validate->reset_errors; |
276 |
|
277 |
=cut |
278 |
|
279 |
sub reset_errors { |
280 |
my $self = shift; |
281 |
delete ($self->{errors}); |
282 |
} |
283 |
|
284 |
=head2 all_errors |
285 |
|
286 |
Return hash with all errors |
287 |
|
288 |
print dump( $validate->all_errors ); |
289 |
|
290 |
=cut |
291 |
|
292 |
sub all_errors { |
293 |
my $self = shift; |
294 |
return $self->{errors}; |
295 |
} |
296 |
|
297 |
=head2 report_error |
298 |
|
299 |
Produce nice humanly readable report of single error |
300 |
|
301 |
print $validate->report_error( $error_hash ); |
302 |
|
303 |
=cut |
304 |
|
305 |
sub report_error { |
306 |
my $self = shift; |
307 |
|
308 |
my $h = shift || die "no hash?"; |
309 |
|
310 |
sub _unroll { |
311 |
my ($self, $tree, $accumulated) = @_; |
312 |
|
313 |
my $log = $self->_get_logger(); |
314 |
|
315 |
$log->debug("# ", |
316 |
( $tree ? "tree: $tree " : '' ), |
317 |
( $accumulated ? "accumulated: $accumulated " : '' ), |
318 |
); |
319 |
|
320 |
my $results; |
321 |
|
322 |
if (ref($tree) ne 'HASH') { |
323 |
return ("$accumulated\t($tree)", undef); |
324 |
} |
325 |
|
326 |
my $dump; |
327 |
|
328 |
foreach my $k (sort keys %{ $tree }) { |
329 |
|
330 |
if ($k eq 'dump') { |
331 |
$dump = $tree->{dump}; |
332 |
# warn "## dump: ",dump($dump),"\n"; |
333 |
next; |
334 |
} |
335 |
|
336 |
$log->debug("current: $k"); |
337 |
|
338 |
my ($new_results, $new_dump) = $self->_unroll($tree->{$k}, |
339 |
$accumulated ? "$accumulated\t$k" : $k |
340 |
); |
341 |
|
342 |
$log->debug( |
343 |
( $new_results ? "new_results: " . dump($new_results) ." " : '' ), |
344 |
); |
345 |
|
346 |
push @$results, $new_results if ($new_results); |
347 |
$dump = $new_dump if ($new_dump); |
348 |
|
349 |
} |
350 |
|
351 |
$log->debug( |
352 |
( $results ? "results: " . dump($results) ." " : '' ), |
353 |
); |
354 |
|
355 |
if ($#$results == 0) { |
356 |
return ($results->[0], $dump); |
357 |
} else { |
358 |
return ($results, $dump); |
359 |
} |
360 |
} |
361 |
|
362 |
|
363 |
sub _reformat { |
364 |
my $l = shift; |
365 |
$l =~ s/\t/ /g; |
366 |
$l =~ s/_/ /; |
367 |
return $l; |
368 |
} |
369 |
|
370 |
my $out = ''; |
371 |
|
372 |
for my $f (sort keys %{ $h }) { |
373 |
$out .= "$f: "; |
374 |
|
375 |
my ($r, $d) = $self->_unroll( $h->{$f} ); |
376 |
my $e; |
377 |
if (ref($r) eq 'ARRAY') { |
378 |
$e .= join(", ", map { _reformat( $_ ) } @$r); |
379 |
} else { |
380 |
$e .= _reformat( $r ); |
381 |
} |
382 |
$e .= "\n\t$d" if ($d); |
383 |
|
384 |
$out .= $e . "\n"; |
385 |
} |
386 |
return $out; |
387 |
} |
388 |
|
389 |
|
390 |
=head2 report |
391 |
|
392 |
Produce nice humanly readable report of errors |
393 |
|
394 |
print $validate->report; |
395 |
|
396 |
=cut |
397 |
|
398 |
sub report { |
399 |
my $self = shift; |
400 |
my $e = $self->{errors} || return; |
401 |
|
402 |
my $out; |
403 |
foreach my $mfn (sort { $a <=> $b } keys %$e) { |
404 |
$out .= "MFN $mfn\n" . $self->report_error( $e->{$mfn} ) . "\n"; |
405 |
} |
406 |
|
407 |
return $out; |
408 |
|
409 |
} |
410 |
|
411 |
=head1 AUTHOR |
412 |
|
413 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
414 |
|
415 |
=head1 COPYRIGHT & LICENSE |
416 |
|
417 |
Copyright 2006 Dobrica Pavlinusic, All Rights Reserved. |
418 |
|
419 |
This program is free software; you can redistribute it and/or modify it |
420 |
under the same terms as Perl itself. |
421 |
|
422 |
=cut |
423 |
|
424 |
1; # End of WebPAC::Validate |