/[webpac]/trunk/all2xml.pl
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Contents of /trunk/all2xml.pl

Parent Directory Parent Directory | Revision Log Revision Log


Revision 775 - (show annotations)
Sun Aug 3 06:31:00 2008 UTC (11 years, 3 months ago) by dpavlin
File MIME type: text/plain
File size: 30252 byte(s)
switch back to GDBM_File which is included in recent
Debian, and TDB_File won't compile against tdb-dev

1 #!/usr/bin/perl -w
2
3 use strict;
4 use Biblio::Isis 0.23;
5 use Getopt::Std;
6 use Data::Dumper;
7 use XML::Simple;
8 use Text::Iconv;
9 use Config::IniFiles;
10 use Encode;
11 use GDBM_File;
12 use Fcntl; # for O_RDWR
13 #use TDB_File;
14 use Carp;
15
16 $|=1;
17
18 my $config_file = $0;
19 $config_file =~ s/\.pl$/.conf/;
20 $config_file = $ARGV[0] if ($ARGV[0] && -f $ARGV[0]);
21 die "FATAL: can't find configuration file '$config_file'" if (! -e $config_file);
22
23 my $config;
24
25 #use index_DBI; # default DBI module for index
26 #use index_DBI_cache; # faster DBI module using memory cache
27 use index_DBI_filter; # filter support for indexes
28 my $index;
29
30 my %opts;
31
32 # usage:
33 # -d directory name
34 # -m multiple directories
35 # -q quiet
36 # -s run swish
37
38 getopts('d:m:qs', \%opts);
39
40 my $path; # this is name of database
41
42 Text::Iconv->raise_error(0); # Conversion errors don't raise exceptions
43
44 # this is encoding of all files on disk, including import_xml/*.xml file and
45 # filter/*.pm files! It will be used to store strings in perl internally!
46 my $codepage = 'ISO-8859-2';
47
48 my $utf2cp = Text::Iconv->new('UTF-8',$codepage);
49 # this function will convert data from XML files to local encoding
50 sub x {
51 return $utf2cp->convert($_[0]);
52 }
53
54 # decode isis/excel or other import codepage
55 my $import2cp;
56
57 # outgoing xml must be in UTF-8
58 my $cp2utf = Text::Iconv->new($codepage,'UTF-8');
59
60 # mapping between data type and tag which specify
61 # format in XML file
62 my %type2tag = (
63 'isis' => 'isis',
64 'excel' => 'column',
65 'marc' => 'marc',
66 'feed' => 'feed',
67 );
68
69 my $cache; # for cacheing
70
71 # lookup hash (tied to file)
72 my %lhash;
73 # this option will cache all lookup entries in memory.
74 # if you are tight on memory, turn this off
75 my $use_lhash_cache = 1;
76
77 my $last_field_name; # cache to prevent repeated fields
78
79 my $broken_cdata = XMLin('<foo><![CDATA[<bar>]]></foo>') eq '<bar>>';
80 warn "XML::Simple on this system seems broken with <![CDATA[..]]>.\n" if ($broken_cdata);
81
82 sub data2xml {
83
84 use xmlify;
85
86 my $type = shift @_;
87 my $row = shift @_;
88 my $add_xml = shift @_;
89 # needed to read values from configuration file
90 my $cfg = shift @_;
91 my $database = shift @_;
92
93 my $xml;
94
95 use parse_format;
96
97 my $html = ""; # html formatted display output
98
99 my %field_usage; # counter for usage of each field
100
101 # sort subrouting using order="" attribute
102 sub by_order {
103 my $va = $config->{indexer}->{$a}->{order} ||
104 $config->{indexer}->{$a};
105 my $vb = $config->{indexer}->{$b}->{order} ||
106 $config->{indexer}->{$b};
107
108 return $va <=> $vb;
109 }
110
111 my @sorted_tags;
112 if ($cache->{tags_by_order}) {
113 @sorted_tags = @{$cache->{tags_by_order}};
114 } else {
115 @sorted_tags = sort by_order keys %{$config->{indexer}};
116 $cache->{tags_by_order} = \@sorted_tags;
117 }
118
119 if (! @sorted_tags) {
120 print STDERR "WARNING: no tags for this type found in import_xml file!\n";
121 }
122
123 # lookup key
124 my $lookup_key;
125
126 # cache for field in pages
127 delete $cache->{display_data};
128 delete $cache->{swish_data};
129 delete $cache->{swish_exact_data};
130 delete $cache->{index_data};
131 delete $cache->{index_delimiter};
132 delete $cache->{distinct};
133 my @page_fields; # names of fields
134
135
136 # subs used to produce output
137
138 sub get_field_name($$$) {
139 my ($config,$field,$field_usage) = @_;
140
141 # find field name (signular, plural)
142 my $field_name = "";
143 if ($config->{indexer}->{$field}->{name_singular} && $field_usage == 1) {
144 $field_name = $config->{indexer}->{$field}->{name_singular};
145 } elsif ($config->{indexer}->{$field}->{name_plural}) {
146 $field_name = $config->{indexer}->{$field}->{name_plural};
147 } elsif ($config->{indexer}->{$field}->{name}) {
148 $field_name = $config->{indexer}->{$field}->{name};
149 } else {
150 print STDERR "WARNING: field '$field' doesn't have 'name' attribute!";
151 }
152
153 if ($field_name) {
154 $field_name = x($field_name);
155 if (! $last_field_name) {
156 $last_field_name = $field_name;
157 return $last_field_name;
158 } elsif ($field_name ne $last_field_name) {
159 $last_field_name = $field_name;
160 return $last_field_name;
161 }
162 }
163 }
164
165
166 # init variables for different types
167 sub init_visible_type($) {
168 my $type = shift;
169
170 # swish, swish_exact, display, index, index_lookup
171 # swish and display defaults
172 my ($s,$se,$d,$i,$il) = (1,0,1,0,0);
173 if (lc($type) eq "display") {
174 $s = 0;
175 } elsif (lc($type) eq "swish") {
176 $d = 0;
177 } elsif (lc($type) eq "index") {
178 ($s,$se,$d,$i) = (0,1,0,1);
179 } elsif (lc($type) eq "swish_exact") {
180 ($s,$se,$d,$i) = (0,1,0,0);
181 } elsif (lc($type) =~ /^lookup/) {
182 ($s,$se,$d,$i,$il) = (0,1,0,0,1);
183 } elsif ($type) {
184 print STDERR "WARNING: unknown type: $type\n";
185 }
186 return ($s,$se,$d,$i,$il);
187 }
188
189
190 # convert
191 #
192 # <tag>
193 # <delimiter>, </delimiter>
194 # <value>200a</value>
195 # </tag>
196 #
197 # to
198 #
199 # <tag delimiter=", ">200a</tag>
200 #
201 # but without loosing spaces in delimiter (becasue
202 # new XML::Simple strips spaces in attribute values
203 # as defined in XML specification)
204 #
205 sub unroll_x($) {
206 my $x = shift;
207
208 if (defined $x->{value}) {
209 my ($v,$d) = ($x->{value}->{content}, $x->{delimiter}->{content});
210 delete $x->{value};
211 delete $x->{delimiter};
212 $x->{content} = $v;
213 $d =~ s#>$## if ($d && $broken_cdata);
214 $x->{delimiter} = $d;
215 }
216 return $x;
217 }
218
219 # begin real work: go field by field
220 foreach my $field (@sorted_tags) {
221
222 $field=x($field);
223 $field_usage{$field}++;
224
225 my $swish_data = "";
226 my $swish_exact_data = "";
227 my $display_data = "";
228 my @index_data;
229 my $line_delimiter;
230
231 my ($swish,$display);
232
233 my $tag = $cfg->val($database, 'import_xml_tag') || $type2tag{$type} || die "can't find which tag to use for type $type";
234
235 # is this field page-by-page?
236 my $iterate_by_page = $config->{indexer}->{$field}->{iterate_by_page};
237 push @page_fields,$field if ($iterate_by_page);
238 my %page_max = ();
239 # default line_delimiter if using
240 my $page_line_delimiter = $config->{indexer}->{$field}->{page_line_delimiter} || '<br/>';
241 $cache->{index_delimiter}->{$field} = $config->{indexer}->{$field}->{index_delimiter};
242 my $distinct = $config->{indexer}->{$field}->{distinct};
243 if ($distinct && !$iterate_by_page) {
244 warn "WARNING: distinct is currently not supported without iterate_by_page!\n";
245 $distinct = 0;
246 }
247
248 my $format_name = $config->{indexer}->{$field}->{format_name};
249 my $format_delimiter = $config->{indexer}->{$field}->{format_delimiter};
250 if ($format_name && $format_delimiter) {
251 $cache->{format}->{$field}->{format_name} = $format_name;
252 $cache->{format}->{$field}->{format_delimiter} = $format_delimiter;
253 }
254
255 foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) {
256
257 $x = unroll_x($x);
258
259 my $format = x($x->{content});
260 my $delimiter = x($x->{delimiter}) || ' ';
261
262 my $repeat_off = 0; # init repeatable offset
263
264 my ($s,$se,$d,$i,$il) = init_visible_type($x->{type});
265
266 # what will separate last line from this one?
267 if ($display_data && $x->{append}) {
268 $line_delimiter = $delimiter;
269 } elsif ($display_data) {
270 $line_delimiter = '<br/>';
271 }
272
273 # init vars so that we go into while...
274 ($swish,$display) = (1,1);
275
276 sub mkformat($$) {
277 my $x = shift || die "mkformat needs tag reference";
278 my $data = shift || return;
279 my $format_name = x($x->{format_name}) || return $data;
280 my $fmt = x($config->{format}->{$format_name}->{content}) || die "<format name=\"$format_name\"> is not defined!";
281 my $format_delimiter = x($x->{format_delimiter});
282 my @data;
283 if ($format_delimiter) {
284 @data = split(/$format_delimiter/,$data);
285 } else {
286 push @data,$data;
287 }
288
289 if ($fmt) {
290 my $nr = scalar $fmt =~ s/%s/%s/g;
291 if (($#data+1) == $nr) {
292 return sprintf($fmt,@data);
293 } else {
294 #print STDERR "mkformat: [$data] can't be split on [$format_delimiter] to $nr fields!\n";
295 return $data;
296 }
297 } else {
298 print STDERR "usage of link '$format_name' without defined format (<link> tag)\n";
299 }
300 }
301
302 # while because of repeatable fields
303 while ($swish || $display) {
304 my $page = $repeat_off;
305 $page_max{$field} = $page if ($iterate_by_page && $page > ($page_max{$field} || 0));
306 ($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp);
307 if ($repeat_off > 1000) {
308 print STDERR "loop (more than 1000 repeatable fields) deteced in $row, $format\n";
309 last;
310 }
311
312 # is this field is lookup?
313 if ($display && $x->{lookup}) {
314 my $null = "<!-- null -->";
315 if ($use_lhash_cache) {
316 if (!defined($cache->{lhash}->{$display})) {
317 my $new_display = $lhash{$display};
318 if (defined($new_display)) {
319 #print STDERR "lookup cache store '$display' = '$new_display'\n";
320 $display = $new_display;
321 $cache->{lhash}->{$display} = $new_display;
322 } else {
323 # print STDERR "WARNING: lookup for '$display' didn't find anything.\n";
324 $display = "";
325 $cache->{lhash}->{$display} = $null;
326 }
327 } else {
328 $display = $cache->{lhash}->{$display};
329 }
330 } else {
331 $display = $lhash{$display} || $null;
332 }
333 }
334
335 # filter="name" ; filter this field through
336 # filter/[name].pm
337 my $filter = $x->{filter};
338 if ($filter && !$cache->{filter_loaded}->{$filter}) {
339 require "filter/".$filter.".pm";
340 $cache->{filter_loaded}->{$filter}++;
341 }
342 # type="swish" ; field for swish
343 if ($swish) {
344 my $tmp = $swish;
345 if ($filter && ($s || $se)) {
346 no strict 'refs';
347 $tmp = join(" ",&$filter($tmp)) if ($s || $se);
348 }
349
350 $swish_data .= $tmp if ($s && $tmp);
351 $swish_exact_data .= "xxbxx $tmp xxexx " if ($tmp && $tmp ne "" && $se);
352 }
353
354 # type="display" ; field for display
355 if ($d && $display) {
356 my $ldel = $delimiter;
357 if ($line_delimiter && $display_data) {
358 $ldel = $line_delimiter;
359 }
360 if ($filter) {
361 no strict 'refs';
362 my @arr;
363 foreach my $tmp (&$filter($display)) {
364 my $tmp2 = mkformat($x,$tmp);
365 push @arr,$tmp2 if ($tmp2);
366 }
367 $display_data .= $ldel if ($display_data && @arr);
368 $display_data .= join($delimiter,@arr);
369 } else {
370 $display_data .= $ldel if ($display_data);
371 my $tmp = mkformat($x,$display);
372 $display_data .= $tmp if ($tmp);
373 }
374 }
375
376 # type="index" ; insert into index
377 my $idisplay;
378 if ($i && $display) {
379 $idisplay = $display;
380 if ($filter) {
381 no strict 'refs';
382 $idisplay = &$filter($idisplay);
383 }
384 push @index_data, $idisplay if ($idisplay && !$iterate_by_page);
385 }
386
387 # store fields in lookup
388 if ($il && $display) {
389 if (lc($x->{type}) eq "lookup_key") {
390 if ($lookup_key) {
391 print STDERR "WARNING: try to redefine lookup_key (keys shouldn't be repeatable fields!)";
392 } else {
393 if ($filter) {
394 no strict 'refs';
395 $lookup_key = &$filter($display);
396 } else {
397 $lookup_key = $display;
398 }
399 }
400 } elsif (lc($x->{type}) eq "lookup_val") {
401 if ($lookup_key) {
402 if ($filter) {
403 no strict 'refs';
404 $lhash{$lookup_key} = &$filter($display);
405 } else {
406 $lhash{$lookup_key} = $display;
407 }
408 } else {
409 print STDERR "WARNING: no lookup_key defined for '$display'?";
410 }
411 }
412
413 }
414
415 # store data for page-by-page repeatable fields
416 if ($iterate_by_page) {
417 sub iterate_fld($$$$$$) {
418 my ($cache,$what,$field,$page,$data,$append) = @_;
419 return if (!$data);
420
421 my $ldel = $page_line_delimiter;
422 $ldel = " " if ($append);
423 #print STDERR "line delimiter: ",Dumper($ldel) if ($ldel);
424 if (! $cache->{$what}->{$field}->[$page]) {
425 push @{$cache->{$what}->{$field}->[$page]}, {
426 data => $data,
427 delimiter => $ldel,
428 };
429 }
430 }
431
432 if ($display_data) {
433 iterate_fld($cache,'display_data',$field,$page,$display_data,$x->{append});
434 }
435 $display_data = "";
436 if ($swish_data) {
437 iterate_fld($cache,'swish_data',$field,$page,$swish_data,$x->{append});
438 $swish_data = "";
439 }
440 if ($swish_exact_data) {
441 iterate_fld($cache,'swish_exact_data',$field,$page,$swish_exact_data,$x->{append});
442 $swish_exact_data = "";
443 }
444
445 if ($idisplay) {
446 my $ldel=$page_line_delimiter;
447 my @index_data;
448 if ($cache->{index_data}->{$field}->[$page]) {
449
450 @index_data = @{$cache->{index_data}->{$field}->[$page]};
451 }
452 if ($x->{append}) {
453 if (@index_data) {
454 $index_data[$#index_data] .= $idisplay;
455 } else {
456 push @index_data, $idisplay;
457 }
458 } else {
459 push @index_data, $idisplay;
460 }
461 $idisplay = "";
462 @{$cache->{index_data}->{$field}->[$page]} = @index_data;
463 }
464 }
465 }
466
467 if (! $iterate_by_page) {
468 my $idel = $x->{index_delimiter};
469 # fill data in index
470 foreach my $tmp (@index_data) {
471 my $i = $d = $tmp;
472 if ($idel && $tmp =~ m/$idel/) {
473 ($i,$d) = split(/$idel/,$tmp);
474 }
475 $index->insert($field, $i, $d, $path);
476 }
477 @index_data = ();
478 }
479 }
480
481 # now try to parse variables from configuration file
482 foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) {
483
484 $x = unroll_x($x);
485
486 my $delimiter = x($x->{delimiter}) || ' ';
487 my $val = $cfg->val($database, x($x->{content}));
488
489 # FIXME index_lookup is not supported!
490 my ($s,$se,$d,$i,$il) = init_visible_type($x->{type});
491
492 if ($val) {
493 $display_data .= $delimiter.$val if ($d);
494 $swish_data .= " ".$val if ($s);
495 $index->insert($field, $val, $val, $path) if ($i);
496 }
497
498 if ($iterate_by_page) {
499 # FIXME data from config tag will appear just
500 # on first page!!!
501 my $page = 0;
502 if ($display_data) {
503 push @{$cache->{display_data}->{$field}->[$page]}, { data => $display_data };
504 $display_data = "";
505 }
506 if ($swish_data) {
507 push @{$cache->{swish_data}->{$field}->[$page]}, { data => $swish_data };
508 $swish_data = "";
509 }
510 if ($swish_exact_data) {
511 push @{$cache->{swish_exact_data}->{$field}->[$page]}, { data => $swish_exact_data };
512 $swish_exact_data = "";
513 }
514 }
515 }
516
517 # save data page-by-page
518 foreach my $field (@page_fields) {
519 my $nr_pages = $page_max{$field} || next;
520 #print STDERR "field '$field' iterate over ",($nr_pages || 0)," pages...\n";
521 #print STDERR Dumper($cache->{display_data});
522 my $seen; # used for distinct
523 for (my $page=0; $page <= $nr_pages; $page++) {
524 my $display_data;
525 my $delimiter = '';
526 foreach my $element (@{ $cache->{display_data}->{$field}->[$page] }) {
527 my $data = $element->{data};
528 die "BUG! no data in element?" unless ($data);
529
530 if ($distinct) {
531 next if ($cache->{distinct}->{$field}->{ $data });
532 $cache->{distinct}->{$field}->{ $data } = 1;
533 }
534
535 if ($cache->{format}->{$field}) {
536 my $tmp = mkformat($cache->{format}->{$field},$data);
537 $display_data .= $delimiter . $tmp if ($tmp);
538 } else {
539 $display_data .= $delimiter . $data;
540 }
541 $delimiter = $element->{delimiter} if ($element->{delimiter});
542 }
543
544 if ($display_data) { # default
545 if ($field eq "headline") {
546 $xml .= xmlify("headline", $display_data);
547 } else {
548
549 # fallback to empty field name if needed
550 $html .= get_field_name($config,$field,$field_usage{$field}) || '';
551 $html .= "#-#".$display_data."###\n";
552 }
553 }
554
555 my $swish_data = join(" ",map { $_->{data} } @{ $cache->{swish_data}->{$field}->[$page] });
556 if ($swish_data) {
557 # remove extra spaces
558 $swish_data =~ s/ +/ /g;
559 $swish_data =~ s/ +$//g;
560
561 $xml .= xmlify($field."_swish", my_unac_string($codepage,$swish_data));
562 }
563
564 my $swish_exact_data = join(" ", map { $_->{data} } @{ $cache->{swish_exact_data}->{$field}->[$page] });
565 if ($swish_exact_data) {
566 $swish_exact_data =~ s/ +/ /g;
567 $swish_exact_data =~ s/ +$//g;
568
569 # add delimiters before and after word.
570 # That is required to produce exact match
571 $xml .= xmlify($field."_swish_exact", my_unac_string($codepage,$swish_exact_data));
572 }
573
574 my $idel = $cache->{index_delimiter}->{$field};
575 foreach my $tmp (@{$cache->{index_data}->{$field}->[$page]}) {
576 my $i = $tmp;
577 my $d = $tmp;
578 if ($idel && $tmp =~ m/$idel/) {
579 ($i,$d) = split(/$idel/,$tmp);
580 }
581 $index->insert($field, $i, $d, $path);
582 #print STDERR "index [$idel] $field: $i --> $d [$path]\n";
583 }
584 }
585
586 }
587
588 if (! $iterate_by_page) {
589 if ($display_data) {
590 if ($field eq "headline") {
591 $xml .= xmlify("headline", $display_data);
592 } else {
593
594 # fallback to empty field name if needed
595 $html .= get_field_name($config,$field,$field_usage{$field}) || '';
596 $html .= "#-#".$display_data."###\n";
597 }
598 }
599 if ($swish_data) {
600 # remove extra spaces
601 $swish_data =~ s/ +/ /g;
602 $swish_data =~ s/ +$//g;
603
604 $xml .= xmlify($field."_swish", my_unac_string($codepage,$swish_data));
605 }
606
607 if ($swish_exact_data) {
608 $swish_exact_data =~ s/ +/ /g;
609 $swish_exact_data =~ s/ +$//g;
610
611 # add delimiters before and after word.
612 # That is required to produce exact match
613 $xml .= xmlify($field."_swish_exact", my_unac_string($codepage,$swish_exact_data));
614 }
615 }
616 }
617
618 # dump formatted output in <html>
619 if ($html) {
620 #$xml .= xmlify("html",$html);
621 $xml .= "<html><![CDATA[ $html ]]></html>";
622 }
623
624 if ($xml) {
625 $xml .= $add_xml if ($add_xml);
626 return "<xml>\n$xml</xml>\n";
627 } else {
628 return;
629 }
630 }
631
632 ##########################################################################
633
634 # read configuration for this script
635 my $cfg = new Config::IniFiles( -file => $config_file );
636
637 # read global.conf configuration
638 my $cfg_global = new Config::IniFiles( -file => 'global.conf' );
639
640 # open index
641 $index = new index_DBI(
642 $cfg_global->val('global', 'dbi_dbd'),
643 $cfg_global->val('global', 'dbi_dsn'),
644 $cfg_global->val('global', 'dbi_user'),
645 $cfg_global->val('global', 'dbi_passwd') || '',
646 );
647
648 my $show_progress = $cfg_global->val('global', 'show_progress');
649
650 my $my_unac_filter = $cfg_global->val('global', 'my_unac_filter');
651 if ($my_unac_filter) {
652 print STDERR "using $my_unac_filter to filter characters for search\n";
653 require $my_unac_filter;
654 } else {
655 print STDERR "### fallback to default my_unac_string!\n";
656 eval q{
657 sub main::my_unac_string($$) {
658 my ($charset, $string) = (@_);
659 return $string;
660 }
661 };
662 }
663
664 foreach my $database ($cfg->Sections) {
665
666 # save database name in global variable path for later
667 # (need for index filter creation)
668 $path = $database;
669
670 my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined";
671 my $add_xml = $cfg -> val($database, 'xml'); # optional
672
673 # create new lookup file
674 my $lookup_file = $cfg -> val($database, 'lookup_newfile'); # optional
675 if ($lookup_file) {
676 if (! -e $lookup_file) {
677 open(LOOKUP, "> $lookup_file") || die "can't create $lookup_file': $!";
678 close(LOOKUP);
679 }
680 tie %lhash, 'GDBM_File', $lookup_file, &GDBM_NEWDB, 0644;
681 #tie %lhash, 'TDB_File', $lookup_file, TDB_CLEAR_IF_FIRST, O_RDWR, 0644;
682 print STDERR "creating lookup file '$lookup_file'\n";
683 # delete memory cache for lookup file
684 delete $cache->{lhash};
685 }
686
687 # open existing lookup file
688 $lookup_file = $cfg -> val($database, 'lookup_open'); # optional
689 if ($lookup_file) {
690 tie %lhash, 'GDBM_File', $lookup_file, &GDBM_READER, 0644;
691 #tie %lhash, 'TDB_File', $lookup_file, TDB_DEFAULT, O_RDWR, 0644;
692 print STDERR "opening lookup file '$lookup_file'\n";
693 }
694
695 my $import_xml_type = $cfg->val($database, 'import_xml_file') || $type;
696 my $import_xml_file = "./import_xml/$import_xml_type.xml";
697
698 if (! -r $import_xml_file) {
699 print STDERR "ERROR: file $import_xml_file not readable skipping!\n";
700 next;
701 }
702
703 print STDERR "reading $import_xml_file\n";
704
705 # extract just type basic
706 my $type_base = $type;
707 $type_base =~ s/_.+$//g;
708
709 my $tag = $cfg->val($database, 'import_xml_tag') || $type2tag{$type_base} || die "can't find which tag to use for type $type";
710 $config=XMLin($import_xml_file, ForceArray => [ $tag, 'config', 'format' ], ForceContent => 1 );
711
712 # check for broken XML::Simple
713 if ( $broken_cdata ) {
714 map {
715 $config->{format}->{$_}->{content} =~ s#>$##;
716 } keys %{ $config->{format} };
717 }
718
719 # helper for progress bar
720 sub fmt_time {
721 my $t = shift || 0;
722 my $out = "";
723
724 my ($ss,$mm,$hh) = gmtime($t);
725 $out .= "${hh}h" if ($hh);
726 $out .= sprintf("%02d:%02d", $mm,$ss);
727 $out .= " " if ($hh == 0);
728 return $out;
729 }
730
731 # output current progress indicator
732 my $last_p = 0;
733 my $start_t = time();
734 sub progress {
735 return if (! $show_progress);
736 my $current = shift;
737 my $total = shift || 1;
738 my $p = int($current * 100 / $total);
739 if ($p < $last_p || $current == 1) {
740 $start_t = time();
741 $last_p = 0;
742 } elsif ($p != $last_p) {
743 my $rate = ($current / (time() - $start_t || 1));
744 my $eta = ($total-$current) / ($rate || 1);
745 printf STDERR ("%5d [%-38s] %-5d %0.1f/s %s\r",$current,"=" x ($p/3)."$p%>", $total, $rate, fmt_time($eta));
746 $last_p = $p;
747 }
748 }
749
750 my $fake_dir = 1;
751 my $fake_pos = 0;
752 my $last_fake_t = time();
753 sub fakeprogress {
754 return if (! $show_progress);
755 my $current = shift @_;
756
757 my @ind = ('-','\\','|','/','-','\\','|','/');
758
759 if ($current < $fake_pos) {
760 $start_t = time();
761 $last_fake_t = 0;
762 $fake_dir = 1;
763 $fake_pos = 0;
764 }
765
766 if (time()-$last_fake_t >= 1) {
767 $last_fake_t = time();
768 $fake_pos += $fake_dir;
769 $fake_dir = -$fake_dir if ($fake_pos > 38);
770 }
771
772 if ($current % 10 == 0) {
773 my $rate = ($current / (time() - $start_t || 1));
774 printf STDERR ("%5d [%-38s] %0.1f/s\r",$current, " " x $fake_pos .$ind[($current / 10) % 8], $rate);
775 }
776 }
777
778 # now read database
779
780 # erase cache for tags by order in this database
781 delete $cache->{tags_by_order};
782
783 if ($type_base eq "isis") {
784
785 my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!";
786
787 $import2cp = Text::Iconv->new($config->{isis_codepage},$codepage);
788 my $db = new Biblio::Isis(
789 isisdb => $isis_db,
790 join_subfields_with => ' ; ',
791 );
792
793 if (! $db) {
794 print STDERR "FATAL: can't read ISIS database: $isis_db, skipping...\n";
795 next;
796 }
797
798 my $max_rowid = $db->count if ($db);
799
800 print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
801
802 for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) {
803 my $row = $db->to_hash( $row_id );
804 if ($row) {
805
806 $row->{mfn} = $row_id;
807 $row->{record} = $db->{record};
808
809 progress($row->{mfn}, $max_rowid);
810
811 my $swishpath = $path."#".int($row->{mfn});
812
813 if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
814 $xml = $cp2utf->convert($xml);
815 use bytes; # as opposed to chars
816 print "Path-Name: $swishpath\n";
817 print "Content-Length: ".(length($xml)+1)."\n";
818 print "Document-Type: XML\n\n$xml\n";
819 }
820 }
821 }
822 print STDERR "\n";
823
824 } elsif ($type_base eq "excel") {
825 require Spreadsheet::ParseExcel;
826 require Spreadsheet::ParseExcel::Utility;
827 import Spreadsheet::ParseExcel::Utility qw(int2col);
828
829 $import2cp = Text::Iconv->new($config->{excel_codepage},$codepage);
830 my $excel_file = $cfg -> val($database, 'excel_file') || die "$database doesn't have 'excel_file' defined!";
831
832 my $sheet = x($config->{sheet}) || die "no sheet in $type.xml";
833 my $start_row = x($config->{start_row}) || die "no start_row in $type.xml";
834 $start_row--;
835
836 my $oBook = Spreadsheet::ParseExcel::Workbook->Parse($excel_file) || die "can't open Excel file '$excel_file'";
837
838 my $sheet_nr = 0;
839 foreach my $oWks (@{$oBook->{Worksheet}}) {
840 #print STDERR "-- SHEET $sheet_nr:", $oWks->{Name}, "\n";
841 last if ($oWks->{Name} eq $sheet);
842 $sheet_nr++;
843 }
844
845 my $oWorksheet = $oBook->{Worksheet}[$sheet_nr];
846
847 print STDERR "using sheet: ",$oWorksheet->{Name},"\n";
848 defined ($oWorksheet) || die "can't find sheet '$sheet' in $excel_file";
849 my $end_row = x($config->{end_row}) || $oWorksheet->{MaxRow};
850
851 for(my $iR = $start_row ; defined $end_row && $iR <= $end_row ; $iR++) {
852 my $row;
853 for(my $iC = $oWorksheet->{MinCol} ; defined $oWorksheet->{MaxCol} && $iC <= $oWorksheet->{MaxCol} ; $iC++) {
854 my $cell = $oWorksheet->{Cells}[$iR][$iC];
855 if ($cell) {
856 # this conversion is a cludge.
857 # Files from Excell could have
858 # characters which don't fit into
859 # destination encoding.
860 $row->{int2col($iC)} = $utf2cp->convert($cell->Value) || $cell->Value;
861 }
862 }
863
864 progress($iR, $end_row);
865
866 # print "row[$iR/$end_row] ";
867 # foreach (keys %{$row}) {
868 # print "$_: ",$row->{$_},"\t";
869 # }
870 # print "\n";
871
872 my $swishpath = $database."#".$iR;
873
874 next if (! $row);
875
876 if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
877 $xml = $cp2utf->convert($xml);
878 use bytes; # as opposed to chars
879 print "Path-Name: $swishpath\n";
880 print "Content-Length: ".(length($xml)+1)."\n";
881 print "Document-Type: XML\n\n$xml\n";
882 }
883 }
884
885 print STDERR "\n";
886
887 } elsif ($type_base eq "marc") {
888
889 require MARC::File::USMARC;
890
891 $import2cp = Text::Iconv->new($config->{marc_codepage},$codepage);
892 my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!";
893
894 # optional argument is format
895 warn "marc_format is no longer used!" if ($config->{marc_format});
896 print STDERR "Reading MARC file '$marc_file'\n";
897
898 my $marc = MARC::File::USMARC->in( $marc_file );
899
900 if (! $marc) {
901 print STDERR "FATAL: can't read MARC file: $marc_file, skipping...\n";
902 next;
903 }
904
905 # count records in MARC file
906 sub marc_count {
907 my $filename = shift || die;
908 my $file = MARC::File::USMARC->in($filename) || return;
909 my $count = 0;
910 while ($file->skip()) {
911 $count++;
912 }
913 return $count;
914 }
915
916 my $count = marc_count($marc_file) || warn "no records in '$marc_file'?";
917
918 my $i = 1;
919
920 while( my $rec = $marc->next() ) {
921
922 progress($i,$count);
923
924 my $swishpath = $database."#".$i;
925
926 if (my $xml = data2xml($type_base,$rec,$add_xml,$cfg,$database)) {
927 $xml = $cp2utf->convert($xml);
928 use bytes; # as opposed to chars
929 print "Path-Name: $swishpath\n";
930 print "Content-Length: ".(length($xml)+1)."\n";
931 print "Document-Type: XML\n\n$xml\n";
932 }
933
934 $i++;
935 }
936
937 print STDERR "\n";
938
939 } elsif ($type_base eq "feed") {
940
941 $import2cp = Text::Iconv->new($config->{feed_codepage},$codepage);
942 my $prog = x($config->{prog}) || die "$database doesn't have 'prog' defined!";
943
944 print STDERR "Reading feed from program '$prog'\n";
945
946 open(FEED,"feeds/$prog |") || die "can't start $prog: $!";
947
948 my $i=1; # record nr.
949
950 my $data;
951 my $line=1;
952
953 while (<FEED>) {
954 chomp;
955
956 if (/^$/) {
957 my $swishpath = $database."#".$i++;
958
959 if (my $xml = data2xml($type_base,$data,$add_xml,$cfg,$database)) {
960 $xml = $cp2utf->convert($xml);
961 use bytes; # as opposed to chars
962 print "Path-Name: $swishpath\n";
963 print "Content-Length: ".(length($xml)+1)."\n";
964 print "Document-Type: XML\n\n$xml\n";
965 }
966 $line = 1;
967 $data = {};
968 next;
969 }
970
971 $line = $1 if (s/^(\d+):\s*//);
972 $data->{$line++} = $_;
973
974 fakeprogress($i);
975
976 }
977 # close lookup
978 untie %lhash if (%lhash);
979
980 } elsif ($type_base eq "dbf") {
981
982 my $dbf_file = $cfg -> val($database, 'dbf_file') || die "$database doesn't have 'dbf_file' defined!";
983 my $dbf_codepage = $cfg -> val($database, 'dbf_codepage') || die "$database doesn't have 'dbf_codepage' defined!";
984 my $dbf_mapping = $cfg -> val($database, 'dbf_mapping') || die "$database doesn't have 'dbf_mapping' defined!";
985
986 $import2cp = Text::Iconv->new($dbf_codepage,$codepage);
987 require XBase;
988 my $db = new XBase $dbf_file;
989
990 if (! $db) {
991 print STDERR "ERROR: can't read DBF database: $dbf_file, skipping...\n";
992 next;
993 }
994
995 my $max_rowid = $db->last_record;
996
997 print STDERR "Reading database: $dbf_file [$max_rowid rows]\n";
998
999 my %dbf2iso;
1000 foreach my $m (split(/[\n\r]+/,$dbf_mapping)) {
1001 my ($col,$fld) = split(/\s+/,$m,2);
1002 $dbf2iso{$col} = $fld;
1003 }
1004
1005 #print STDERR "## dbf2iso: ",Dumper(\%dbf2iso),"\n## /dbf2iso\n";
1006
1007 # bad, bad...
1008 require "to_hash.pm";
1009
1010 foreach my $row_id (0 .. $max_rowid) {
1011 my $dbf_row = $db->get_record_as_hash($row_id);
1012 if ($dbf_row) {
1013
1014 #print STDERR "## dbf_row: ",Dumper($dbf_row),"\n## /dbf_row\n";
1015 # apply mapping from config file
1016 # all unspecified records will get _ in
1017 # front of them - _DELETE will be __DELETE
1018 my $rec;
1019 map {
1020 my $new_fld = $dbf2iso{$_} || '_'.$_;
1021 my $data = $dbf_row->{$_};
1022 push @{ $rec->{$new_fld} }, $data if ($data && $data !~ /^(?:\s+|\$a\.|)$/);
1023 } keys %{$dbf_row};
1024 #print STDERR "## rec: ",Dumper($rec),"\n## /rec\n";
1025 my $row = to_hash($row_id+1, $rec);
1026
1027 $row->{mfn} = $row_id+1;
1028 $row->{record} = $rec;
1029
1030 #print STDERR "## row: ",Dumper($row),"\n## /row\n";
1031 progress($row->{mfn}, $max_rowid);
1032
1033 my $swishpath = $path."#".int($row->{mfn});
1034
1035 if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
1036 $xml = $cp2utf->convert($xml);
1037 use bytes; # as opposed to chars
1038 print "Path-Name: $swishpath\n";
1039 print "Content-Length: ".(length($xml)+1)."\n";
1040 print "Document-Type: XML\n\n$xml\n";
1041 }
1042 }
1043 }
1044 print STDERR "\n";
1045 }
1046 }
1047
1048 # call this to commit index
1049 $index->close;
1050
1051 1;
1052 __END__
1053 ##########################################################################
1054
1055 =head1 NAME
1056
1057 all2xml.pl - read various file formats and dump XML for SWISH-E
1058
1059 =head1 SYNOPSYS
1060
1061 $ all2xml.pl [test.conf]
1062
1063 =head1 DESCRIPTION
1064
1065 This command will read ISIS data file using Biblio::Isis perl module, MARC
1066 records using MARC::File module and optionally Micro$oft Excel files to
1067 create one XML file for usage with I<SWISH-E> indexer. Dispite it's name,
1068 this script B<isn't general xml generator> from isis files (isis allready
1069 has something like that). Output of this script is tailor-made for SWISH-E.
1070
1071 If no configuration file is specified, it will use default one called
1072 C<all2xml.conf>.
1073
1074 =head1 BUGS
1075
1076 Documentation is really lacking. However, in true Open Source spirit, source
1077 is best documentation. I even made considerable effort to comment parts
1078 which are not intuitively clear, so...
1079
1080 =head1 AUTHOR
1081
1082 Dobrica Pavlinusic <dpavlin@rot13.org>
1083
1084 =head1 COPYRIGHT
1085
1086 GNU Public License (GPL) v2 or later
1087
1088 =head1 SEE ALSO
1089
1090 SWISH-E web site at http://www.swish-e.org
1091
1092 =cut

Properties

Name Value
cvs2svn:cvs-rev 1.64
svn:executable *

  ViewVC Help
Powered by ViewVC 1.1.26