| 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 0.12 |
| 22 |
|
| 23 |
=cut |
| 24 |
|
| 25 |
our $VERSION = '0.12'; |
| 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 |
delimiters_path => 'conf/validate/delimiters/file', |
| 57 |
); |
| 58 |
|
| 59 |
Optional parametar C<delimiters> will turn on validating of delimiters. Be |
| 60 |
careful here, those delimiters are just stuck into regex, so they can |
| 61 |
contain L<perlre> regexpes. |
| 62 |
|
| 63 |
C<path> and C<delimiters_path> can be specified by L<read_validate_file> and |
| 64 |
L<read_validate_delimiters> calls. |
| 65 |
|
| 66 |
=cut |
| 67 |
|
| 68 |
sub new { |
| 69 |
my $class = shift; |
| 70 |
my $self = {@_}; |
| 71 |
bless($self, $class); |
| 72 |
|
| 73 |
my $log = $self->_get_logger(); |
| 74 |
|
| 75 |
$self->read_validate_file( $self->{path} ) if ( $self->{path} ); |
| 76 |
|
| 77 |
if ( $self->{delimiters} ) { |
| 78 |
$self->{delimiters_regex} = '(\^[a-z0-9]|' . join('|', @{ $self->{delimiters} }) . ')'; |
| 79 |
$log->info("validation check delimiters with regex $self->{delimiters_regex}"); |
| 80 |
} |
| 81 |
|
| 82 |
$self->read_validate_delimiters_file( $self->{delimiters_path} ) if ( $self->{delimiters_path} ); |
| 83 |
|
| 84 |
return $self; |
| 85 |
} |
| 86 |
|
| 87 |
|
| 88 |
=head2 read_validate_file |
| 89 |
|
| 90 |
Specify validate rules file |
| 91 |
|
| 92 |
$validate->read_validate_file( 'conf/validate/file' ); |
| 93 |
|
| 94 |
Returns number of lines in file |
| 95 |
|
| 96 |
=cut |
| 97 |
|
| 98 |
sub read_validate_file { |
| 99 |
my $self = shift; |
| 100 |
|
| 101 |
my $path = shift || die "no path?"; |
| 102 |
|
| 103 |
my $log = $self->_get_logger(); |
| 104 |
|
| 105 |
my $v_file = read_file( $path ) || |
| 106 |
$log->logdie("can't open validate path $path: $!"); |
| 107 |
|
| 108 |
my $v; |
| 109 |
my $curr_line = 1; |
| 110 |
|
| 111 |
foreach my $l (split(/[\n\r]+/, $v_file)) { |
| 112 |
$curr_line++; |
| 113 |
|
| 114 |
# skip comments and whitespaces |
| 115 |
next if ($l =~ /^#/ || $l =~ /^\s*$/); |
| 116 |
|
| 117 |
$l =~ s/^\s+//; |
| 118 |
$l =~ s/\s+$//; |
| 119 |
|
| 120 |
my @d = split(/\s+/, $l); |
| 121 |
|
| 122 |
my $fld = shift @d; |
| 123 |
|
| 124 |
if ($fld =~ s/!$//) { |
| 125 |
$self->{must_exist}->{$fld}++; |
| 126 |
} elsif ($fld =~ s/-$//) { |
| 127 |
$self->{dont_validate}->{$fld}++; |
| 128 |
} |
| 129 |
|
| 130 |
$log->logdie("need field name in line $curr_line: $l") unless (defined($fld)); |
| 131 |
|
| 132 |
if (@d) { |
| 133 |
$v->{$fld} = [ map { |
| 134 |
my $sf = $_; |
| 135 |
if ( $sf =~ s/!(\*)?$/$1/ ) { |
| 136 |
$self->{must_exist_sf}->{ $fld }->{ $sf }++; |
| 137 |
}; |
| 138 |
$sf; |
| 139 |
} @d ]; |
| 140 |
} else { |
| 141 |
$v->{$fld} = 1; |
| 142 |
} |
| 143 |
|
| 144 |
} |
| 145 |
|
| 146 |
$log->debug("current validation rules: ", dump($v)); |
| 147 |
|
| 148 |
$self->{rules} = $v; |
| 149 |
|
| 150 |
$log->info("validation uses rules from $path"); |
| 151 |
|
| 152 |
return $curr_line; |
| 153 |
} |
| 154 |
|
| 155 |
=head2 read_validate_delimiters_file |
| 156 |
|
| 157 |
$validate->read_validate_delimiters_file( 'conf/validate/delimiters/file' ); |
| 158 |
|
| 159 |
=cut |
| 160 |
|
| 161 |
sub read_validate_delimiters_file { |
| 162 |
my $self = shift; |
| 163 |
|
| 164 |
my $path = shift || die "no path?"; |
| 165 |
|
| 166 |
my $log = $self->_get_logger(); |
| 167 |
|
| 168 |
if ( -e $path ) { |
| 169 |
$log->info("using delimiter validation rules from $path"); |
| 170 |
open(my $d, $path) || $log->fatal("can't open $path: $!"); |
| 171 |
while(<$d>) { |
| 172 |
chomp($d); |
| 173 |
if (/^\s*(#*)\s*(\d+)\t+(\d+)\t+(.*)$/) { |
| 174 |
my ($comment,$field,$count,$template) = ($1,$2,$3,$4); |
| 175 |
$self->{_validate_delimiters_templates}->{$field}->{$template} = $count unless ($comment); |
| 176 |
} else { |
| 177 |
warn "## ignored $d\n"; |
| 178 |
} |
| 179 |
} |
| 180 |
close($d); |
| 181 |
#warn "_validate_delimiters_templates = ",dump( $self->{_validate_delimiters_templates} ); |
| 182 |
} else { |
| 183 |
$log->warn("delimiters path $path doesn't exist, it will be created after this run"); |
| 184 |
} |
| 185 |
} |
| 186 |
|
| 187 |
=head2 validate_rec |
| 188 |
|
| 189 |
Validate record and return errors |
| 190 |
|
| 191 |
my @errors = $validate->validate_rec( $rec, $rec_dump ); |
| 192 |
|
| 193 |
=cut |
| 194 |
|
| 195 |
sub validate_rec { |
| 196 |
my $self = shift; |
| 197 |
|
| 198 |
my $log = $self->_get_logger(); |
| 199 |
|
| 200 |
my $rec = shift || $log->logdie("validate_rec need record"); |
| 201 |
my $rec_dump = shift; |
| 202 |
|
| 203 |
$log->logdie("rec isn't HASH") unless (ref($rec) eq 'HASH'); |
| 204 |
# $log->logdie("can't find validation rules") unless (my $r = $self->{rules}); |
| 205 |
my $r = $self->{rules}; |
| 206 |
|
| 207 |
my $errors; |
| 208 |
|
| 209 |
$log->debug("rec = ", sub { dump($rec) }, "keys = ", keys %{ $rec }); |
| 210 |
|
| 211 |
my $fields; |
| 212 |
|
| 213 |
foreach my $f (keys %{ $rec }) { |
| 214 |
|
| 215 |
next if (!defined($f) || $f eq '' || $f eq '000'); |
| 216 |
|
| 217 |
# first check delimiters |
| 218 |
if ( my $regex = $self->{delimiters_regex} ) { |
| 219 |
|
| 220 |
foreach my $v (@{ $rec->{$f} }) { |
| 221 |
my $l = _pack_subfields_hash( $v, 1 ); |
| 222 |
my $subfield_dump = $l; |
| 223 |
my $template = ''; |
| 224 |
$l =~ s/$regex/$template.=$1/eg; |
| 225 |
#warn "## template: $template\n"; |
| 226 |
|
| 227 |
if ( $template ) { |
| 228 |
$self->{_delimiters_templates}->{$f}->{$template}++; |
| 229 |
|
| 230 |
if ( my $v = $self->{_validate_delimiters_templates} ) { |
| 231 |
if ( ! defined( $v->{$f}->{$template} ) ) { |
| 232 |
$errors->{$f}->{potentially_invalid_combination} = $template; |
| 233 |
$errors->{$f}->{dump} = $subfield_dump; |
| 234 |
#} else { |
| 235 |
# warn "## $f $template ok\n"; |
| 236 |
} |
| 237 |
} |
| 238 |
} |
| 239 |
} |
| 240 |
} |
| 241 |
|
| 242 |
next unless ( $r ); # skip validation of no rules are specified |
| 243 |
|
| 244 |
next if (defined( $self->{dont_validate}->{$f} )); |
| 245 |
|
| 246 |
# track field usage |
| 247 |
$fields->{$f}++; |
| 248 |
|
| 249 |
if ( ! defined($r->{$f}) ) { |
| 250 |
$errors->{ $f }->{unexpected} = "this field is not expected"; |
| 251 |
next; |
| 252 |
} |
| 253 |
|
| 254 |
|
| 255 |
if (ref($rec->{$f}) ne 'ARRAY') { |
| 256 |
$errors->{ $f }->{not_repeatable} = "probably bug in parsing input data"; |
| 257 |
next; |
| 258 |
} |
| 259 |
|
| 260 |
foreach my $v (@{ $rec->{$f} }) { |
| 261 |
# can we have subfields? |
| 262 |
if (ref($r->{$f}) eq 'ARRAY') { |
| 263 |
# are values hashes? (has subfields) |
| 264 |
if (! defined($v)) { |
| 265 |
# $errors->{$f}->{empty} = undef; |
| 266 |
# $errors->{dump} = $rec_dump if ($rec_dump); |
| 267 |
} elsif (ref($v) ne 'HASH') { |
| 268 |
$errors->{$f}->{missing_subfield} = join(",", @{ $r->{$f} }) . " required"; |
| 269 |
next; |
| 270 |
} else { |
| 271 |
|
| 272 |
my $h = dclone( $v ); |
| 273 |
|
| 274 |
my $sf_repeatable; |
| 275 |
|
| 276 |
delete($v->{subfields}) if (defined($v->{subfields})); |
| 277 |
|
| 278 |
my $subfields; |
| 279 |
|
| 280 |
foreach my $sf (keys %{ $v }) { |
| 281 |
|
| 282 |
$subfields->{ $sf }++; |
| 283 |
|
| 284 |
# is non-repeatable but with multiple values? |
| 285 |
if ( ! first { $_ eq $sf.'*' } @{$r->{$f}} ) { |
| 286 |
if ( ref($v->{$sf}) eq 'ARRAY' ) { |
| 287 |
$sf_repeatable->{$sf}++; |
| 288 |
}; |
| 289 |
if (! first { $_ eq $sf } @{ $r->{$f} }) { |
| 290 |
$errors->{ $f }->{subfield}->{extra}->{$sf}++; |
| 291 |
} |
| 292 |
} |
| 293 |
|
| 294 |
} |
| 295 |
if (my @r_sf = sort keys( %$sf_repeatable )) { |
| 296 |
|
| 297 |
foreach my $sf (@r_sf) { |
| 298 |
$errors->{$f}->{subfield}->{extra_repeatable}->{$sf}++; |
| 299 |
$errors->{$f}->{dump} = _pack_subfields_hash( $h, 1 ); |
| 300 |
} |
| 301 |
|
| 302 |
} |
| 303 |
|
| 304 |
if ( defined( $self->{must_exist_sf}->{$f} ) ) { |
| 305 |
foreach my $sf (sort keys %{ $self->{must_exist_sf}->{$f} }) { |
| 306 |
#warn "====> $f $sf must exist\n"; |
| 307 |
$errors->{$f}->{subfield}->{missing}->{$sf}++ |
| 308 |
unless defined( $subfields->{$sf} ); |
| 309 |
} |
| 310 |
} |
| 311 |
|
| 312 |
} |
| 313 |
} elsif (ref($v) eq 'HASH') { |
| 314 |
$errors->{$f}->{unexpected_subfields}++; |
| 315 |
$errors->{$f}->{dump} = _pack_subfields_hash( $v, 1 ); |
| 316 |
} |
| 317 |
} |
| 318 |
} |
| 319 |
|
| 320 |
$log->debug("_delimiters_templates = ", sub { dump( $self->{_delimiters_templates} ) } ); |
| 321 |
|
| 322 |
foreach my $must (sort keys %{ $self->{must_exist} }) { |
| 323 |
next if ($fields->{$must}); |
| 324 |
$errors->{$must}->{missing}++; |
| 325 |
$errors->{dump} = $rec_dump if ($rec_dump); |
| 326 |
} |
| 327 |
|
| 328 |
if ($errors) { |
| 329 |
$log->debug("errors: ", $self->report_error( $errors ) ); |
| 330 |
|
| 331 |
my $mfn = $rec->{'000'}->[0] || $log->logconfess("record ", sub { dump( $rec ) }, " doesn't have MFN"); |
| 332 |
$self->{errors}->{$mfn} = $errors; |
| 333 |
} |
| 334 |
|
| 335 |
#$log->logcluck("return from this function is ARRAY") unless wantarray; |
| 336 |
|
| 337 |
return $errors; |
| 338 |
} |
| 339 |
|
| 340 |
=head2 reset |
| 341 |
|
| 342 |
Clean all accumulated errors for this input and remember delimiter templates |
| 343 |
for L<save_delimiters_templates> |
| 344 |
|
| 345 |
$validate->reset; |
| 346 |
|
| 347 |
This function B<must> be called after each input to provide accurate statistics. |
| 348 |
|
| 349 |
=cut |
| 350 |
|
| 351 |
sub reset { |
| 352 |
my $self = shift; |
| 353 |
|
| 354 |
my $log = $self->_get_logger; |
| 355 |
|
| 356 |
delete ($self->{errors}); |
| 357 |
|
| 358 |
if ( ! $self->{_delimiters_templates} ) { |
| 359 |
$log->debug("called without _delimiters_templates?"); |
| 360 |
return; |
| 361 |
} |
| 362 |
|
| 363 |
foreach my $f ( keys %{ $self->{_delimiters_templates} } ) { |
| 364 |
foreach my $t ( keys %{ $self->{_delimiters_templates}->{$f} } ) { |
| 365 |
$self->{_accumulated_delimiters_templates}->{$f}->{$t} += |
| 366 |
$self->{_delimiters_templates}->{$f}->{$t}; |
| 367 |
} |
| 368 |
} |
| 369 |
$log->debug("_accumulated_delimiters_templates = ", sub { dump( $self->{_accumulated_delimiter_templates} ) } ); |
| 370 |
delete ($self->{_delimiters_templates}); |
| 371 |
} |
| 372 |
|
| 373 |
=head2 all_errors |
| 374 |
|
| 375 |
Return hash with all errors |
| 376 |
|
| 377 |
print dump( $validate->all_errors ); |
| 378 |
|
| 379 |
=cut |
| 380 |
|
| 381 |
sub all_errors { |
| 382 |
my $self = shift; |
| 383 |
return $self->{errors}; |
| 384 |
} |
| 385 |
|
| 386 |
=head2 report_error |
| 387 |
|
| 388 |
Produce nice humanly readable report of single error |
| 389 |
|
| 390 |
print $validate->report_error( $error_hash ); |
| 391 |
|
| 392 |
=cut |
| 393 |
|
| 394 |
sub report_error { |
| 395 |
my $self = shift; |
| 396 |
|
| 397 |
my $h = shift || die "no hash?"; |
| 398 |
|
| 399 |
sub _unroll { |
| 400 |
my ($self, $tree, $accumulated) = @_; |
| 401 |
|
| 402 |
my $log = $self->_get_logger(); |
| 403 |
|
| 404 |
$log->debug("# ", |
| 405 |
( $tree ? "tree: $tree " : '' ), |
| 406 |
( $accumulated ? "accumulated: $accumulated " : '' ), |
| 407 |
); |
| 408 |
|
| 409 |
my $results; |
| 410 |
|
| 411 |
if (ref($tree) ne 'HASH') { |
| 412 |
return ("$accumulated\t($tree)", undef); |
| 413 |
} |
| 414 |
|
| 415 |
my $dump; |
| 416 |
|
| 417 |
foreach my $k (sort keys %{ $tree }) { |
| 418 |
|
| 419 |
if ($k eq 'dump') { |
| 420 |
$dump = $tree->{dump}; |
| 421 |
#warn "## dump ",dump($dump),"\n"; |
| 422 |
next; |
| 423 |
} |
| 424 |
|
| 425 |
$log->debug("current: $k"); |
| 426 |
|
| 427 |
my ($new_results, $new_dump) = $self->_unroll($tree->{$k}, |
| 428 |
$accumulated ? "$accumulated\t$k" : $k |
| 429 |
); |
| 430 |
|
| 431 |
$log->debug( "new_results: ", sub { dump($new_results) } ) if ( $new_results ); |
| 432 |
|
| 433 |
push @$results, $new_results if ($new_results); |
| 434 |
$dump = $new_dump if ($new_dump); |
| 435 |
|
| 436 |
} |
| 437 |
|
| 438 |
$log->debug( "results: ", sub { dump($results) } ) if ( $results ); |
| 439 |
|
| 440 |
if ($#$results == 0) { |
| 441 |
return ($results->[0], $dump); |
| 442 |
} else { |
| 443 |
return ($results, $dump); |
| 444 |
} |
| 445 |
} |
| 446 |
|
| 447 |
|
| 448 |
sub _reformat { |
| 449 |
my $l = shift; |
| 450 |
$l =~ s/\t/ /g; |
| 451 |
$l =~ s/_/ /g; |
| 452 |
return $l; |
| 453 |
} |
| 454 |
|
| 455 |
my $out = ''; |
| 456 |
|
| 457 |
for my $f (sort keys %{ $h }) { |
| 458 |
$out .= "$f: "; |
| 459 |
|
| 460 |
my ($r, $d) = $self->_unroll( $h->{$f} ); |
| 461 |
my $e; |
| 462 |
if (ref($r) eq 'ARRAY') { |
| 463 |
$e .= join(", ", map { _reformat( $_ ) } @$r); |
| 464 |
} else { |
| 465 |
$e .= _reformat( $r ); |
| 466 |
} |
| 467 |
$e .= "\n\t$d" if ($d); |
| 468 |
|
| 469 |
$out .= $e . "\n"; |
| 470 |
} |
| 471 |
return $out; |
| 472 |
} |
| 473 |
|
| 474 |
|
| 475 |
=head2 report |
| 476 |
|
| 477 |
Produce nice humanly readable report of errors |
| 478 |
|
| 479 |
print $validate->report; |
| 480 |
|
| 481 |
=cut |
| 482 |
|
| 483 |
sub report { |
| 484 |
my $self = shift; |
| 485 |
my $e = $self->{errors} || return; |
| 486 |
|
| 487 |
my $out; |
| 488 |
foreach my $mfn (sort { $a <=> $b } keys %$e) { |
| 489 |
$out .= "MFN $mfn\n" . $self->report_error( $e->{$mfn} ) . "\n"; |
| 490 |
} |
| 491 |
|
| 492 |
return $out; |
| 493 |
|
| 494 |
} |
| 495 |
|
| 496 |
=head2 delimiters_templates |
| 497 |
|
| 498 |
Generate report of delimiter tamplates |
| 499 |
|
| 500 |
my $report = $validate->delimiter_teplates( |
| 501 |
report => 1, |
| 502 |
current_input => 1, |
| 503 |
); |
| 504 |
|
| 505 |
Options: |
| 506 |
|
| 507 |
=over 4 |
| 508 |
|
| 509 |
=item report |
| 510 |
|
| 511 |
Generate humanly readable report with single fields |
| 512 |
|
| 513 |
=item current_input |
| 514 |
|
| 515 |
Report just current_input and not accumulated data |
| 516 |
|
| 517 |
=back |
| 518 |
|
| 519 |
=cut |
| 520 |
|
| 521 |
sub delimiters_templates { |
| 522 |
my $self = shift; |
| 523 |
|
| 524 |
my $args = {@_}; |
| 525 |
|
| 526 |
my $t = $self->{_accumulated_delimiters_templates}; |
| 527 |
$t = $self->{_delimiters_templates} if ( $args->{current_input} ); |
| 528 |
|
| 529 |
my $log = $self->_get_logger; |
| 530 |
|
| 531 |
unless ($t) { |
| 532 |
$log->error("called without delimiters"); |
| 533 |
return; |
| 534 |
} |
| 535 |
|
| 536 |
my $out; |
| 537 |
|
| 538 |
foreach my $f (sort { $a <=> $b } keys %$t) { |
| 539 |
$out .= "$f\n" if ( $args->{report} ); |
| 540 |
foreach my $template (sort { $a cmp $b } keys %{ $t->{$f} }) { |
| 541 |
my $count = $t->{$f}->{$template}; |
| 542 |
$out .= |
| 543 |
( $count ? "" : "# " ) . |
| 544 |
( $args->{report} ? "" : "$f" ) . |
| 545 |
"\t$count\t$template\n"; |
| 546 |
} |
| 547 |
} |
| 548 |
|
| 549 |
return $out; |
| 550 |
} |
| 551 |
|
| 552 |
=head2 save_delimiters_templates |
| 553 |
|
| 554 |
Save accumulated delimiter templates |
| 555 |
|
| 556 |
$validator->save_delimiters_template( '/path/to/validate/delimiters' ); |
| 557 |
|
| 558 |
=cut |
| 559 |
|
| 560 |
sub save_delimiters_templates { |
| 561 |
my $self = shift; |
| 562 |
|
| 563 |
my $path = $self->{delimiters_path}; |
| 564 |
|
| 565 |
return unless ( $path ); |
| 566 |
|
| 567 |
my $log = $self->_get_logger; |
| 568 |
|
| 569 |
if ( ! $self->{_accumulated_delimiters_templates} ) { |
| 570 |
$log->error('no _accumulated_delimiters_templates found, reset'); |
| 571 |
$self->reset; |
| 572 |
} |
| 573 |
|
| 574 |
if ( ! $self->{_delimiters_templates} ) { |
| 575 |
$log->error('found _delimiters_templates, calling reset'); |
| 576 |
$self->reset; |
| 577 |
} |
| 578 |
|
| 579 |
$path .= '.new' if ( -e $path ); |
| 580 |
|
| 581 |
open(my $d, '>', $path) || $log->fatal("can't open $path: $!"); |
| 582 |
print $d $self->delimiters_templates; |
| 583 |
close($d); |
| 584 |
|
| 585 |
$log->info("new delimiters templates saved to $path"); |
| 586 |
} |
| 587 |
|
| 588 |
=head1 AUTHOR |
| 589 |
|
| 590 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
| 591 |
|
| 592 |
=head1 COPYRIGHT & LICENSE |
| 593 |
|
| 594 |
Copyright 2006 Dobrica Pavlinusic, All Rights Reserved. |
| 595 |
|
| 596 |
This program is free software; you can redistribute it and/or modify it |
| 597 |
under the same terms as Perl itself. |
| 598 |
|
| 599 |
=cut |
| 600 |
|
| 601 |
1; # End of WebPAC::Validate |
Parent Directory
| 
Revision Log