1 |
#!/usr/bin/perl -w |
2 |
|
3 |
use strict; |
4 |
use OpenIsis; |
5 |
use Getopt::Std; |
6 |
use Data::Dumper; |
7 |
use XML::Simple; |
8 |
use Text::Unaccent 1.02; # 1.01 won't compile on my platform, |
9 |
use Text::Iconv; |
10 |
use Config::IniFiles; |
11 |
use Encode; |
12 |
#use GDBM_File; |
13 |
use Fcntl; # for O_RDWR |
14 |
use TDB_File; |
15 |
|
16 |
$|=1; |
17 |
|
18 |
my $config_file = $0; |
19 |
$config_file =~ s/\.pl$/.conf/; |
20 |
die "FATAL: can't find configuration file '$config_file'" if (! -e $config_file); |
21 |
|
22 |
my $config; |
23 |
|
24 |
#use index_DBI; # default DBI module for index |
25 |
use index_DBI_cache; # faster DBI module using memory cache |
26 |
my $index; |
27 |
|
28 |
my %opts; |
29 |
|
30 |
# usage: |
31 |
# -d directory name |
32 |
# -m multiple directories |
33 |
# -q quiet |
34 |
# -s run swish |
35 |
|
36 |
getopts('d:m:qs', \%opts); |
37 |
|
38 |
my $path; # this is name of database |
39 |
|
40 |
Text::Iconv->raise_error(0); # Conversion errors don't raise exceptions |
41 |
|
42 |
# this is encoding of all files on disk, including import_xml/*.xml file and |
43 |
# filter/*.pm files! It will be used to store strings in perl internally! |
44 |
my $codepage = 'ISO-8859-2'; |
45 |
|
46 |
my $utf2cp = Text::Iconv->new('UTF-8',$codepage); |
47 |
# this function will convert data from XML files to local encoding |
48 |
sub x { |
49 |
return $utf2cp->convert($_[0]); |
50 |
} |
51 |
|
52 |
# decode isis/excel or other import codepage |
53 |
my $import2cp; |
54 |
|
55 |
# outgoing xml must be in UTF-8 |
56 |
my $cp2utf = Text::Iconv->new($codepage,'UTF-8'); |
57 |
|
58 |
# mapping between data type and tag which specify |
59 |
# format in XML file |
60 |
my %type2tag = ( |
61 |
'isis' => 'isis', |
62 |
'excel' => 'column', |
63 |
'marc' => 'marc', |
64 |
'feed' => 'feed' |
65 |
); |
66 |
|
67 |
my $cache; # for cacheing |
68 |
|
69 |
# lookup hash (tied to file) |
70 |
my %lhash; |
71 |
# this option will cache all lookup entries in memory. |
72 |
# if you are tight on memory, turn this off |
73 |
my $use_lhash_cache = 1; |
74 |
|
75 |
my $last_field_name; # cache to prevent repeated fields |
76 |
|
77 |
sub data2xml { |
78 |
|
79 |
use xmlify; |
80 |
|
81 |
my $type = shift @_; |
82 |
my $row = shift @_; |
83 |
my $add_xml = shift @_; |
84 |
# needed to read values from configuration file |
85 |
my $cfg = shift @_; |
86 |
my $database = shift @_; |
87 |
|
88 |
my $xml; |
89 |
|
90 |
use parse_format; |
91 |
|
92 |
my $html = ""; # html formatted display output |
93 |
|
94 |
my %field_usage; # counter for usage of each field |
95 |
|
96 |
# sort subrouting using order="" attribute |
97 |
sub by_order { |
98 |
my $va = $config->{indexer}->{$a}->{order} || |
99 |
$config->{indexer}->{$a}; |
100 |
my $vb = $config->{indexer}->{$b}->{order} || |
101 |
$config->{indexer}->{$b}; |
102 |
|
103 |
return $va <=> $vb; |
104 |
} |
105 |
|
106 |
my @sorted_tags; |
107 |
if ($cache->{tags_by_order}) { |
108 |
@sorted_tags = @{$cache->{tags_by_order}}; |
109 |
} else { |
110 |
@sorted_tags = sort by_order keys %{$config->{indexer}}; |
111 |
$cache->{tags_by_order} = \@sorted_tags; |
112 |
} |
113 |
|
114 |
# lookup key |
115 |
my $lookup_key; |
116 |
|
117 |
# cache for field in pages |
118 |
delete $cache->{display_data}; |
119 |
delete $cache->{swish_data}; |
120 |
delete $cache->{swish_exact_data}; |
121 |
delete $cache->{index_data}; |
122 |
delete $cache->{index_delimiter}; |
123 |
my @page_fields; # names of fields |
124 |
|
125 |
|
126 |
# subs used to produce output |
127 |
|
128 |
sub get_field_name($$$) { |
129 |
my ($config,$field,$field_usage) = @_; |
130 |
|
131 |
# find field name (signular, plural) |
132 |
my $field_name = ""; |
133 |
if ($config->{indexer}->{$field}->{name_singular} && $field_usage == 1) { |
134 |
$field_name = $config->{indexer}->{$field}->{name_singular}; |
135 |
} elsif ($config->{indexer}->{$field}->{name_plural}) { |
136 |
$field_name = $config->{indexer}->{$field}->{name_plural}; |
137 |
} elsif ($config->{indexer}->{$field}->{name}) { |
138 |
$field_name = $config->{indexer}->{$field}->{name}; |
139 |
} else { |
140 |
print STDERR "WARNING: field '$field' doesn't have 'name' attribute!"; |
141 |
} |
142 |
if ($field_name) { |
143 |
if (! $last_field_name) { |
144 |
$last_field_name = x($field_name); |
145 |
return $last_field_name; |
146 |
} elsif ($field_name ne $last_field_name) { |
147 |
$last_field_name = x($field_name); |
148 |
return $last_field_name; |
149 |
} |
150 |
} |
151 |
} |
152 |
|
153 |
|
154 |
# begin real work: go field by field |
155 |
foreach my $field (@sorted_tags) { |
156 |
|
157 |
$field=x($field); |
158 |
$field_usage{$field}++; |
159 |
|
160 |
my $swish_data = ""; |
161 |
my $swish_exact_data = ""; |
162 |
my $display_data = ""; |
163 |
my @index_data; |
164 |
my $line_delimiter; |
165 |
|
166 |
my ($swish,$display); |
167 |
|
168 |
my $tag = $type2tag{$type} || die "can't find which tag to use for type $type"; |
169 |
|
170 |
# is this field page-by-page? |
171 |
my $iterate_by_page = $config->{indexer}->{$field}->{iterate_by_page}; |
172 |
push @page_fields,$field if ($iterate_by_page); |
173 |
my %page_max = (); |
174 |
# default line_delimiter if using |
175 |
my $page_line_delimiter = $config->{indexer}->{$field}->{page_line_delimiter} || '<br/>'; |
176 |
$cache->{index_delimiter}->{$field} = $config->{indexer}->{$field}->{index_delimiter}; |
177 |
|
178 |
my $format_name = $config->{indexer}->{$field}->{format_name}; |
179 |
my $format_delimiter = $config->{indexer}->{$field}->{format_delimiter}; |
180 |
if ($format_name && $format_delimiter) { |
181 |
$cache->{format}->{$field}->{format_name} = $format_name; |
182 |
$cache->{format}->{$field}->{format_delimiter} = $format_delimiter; |
183 |
} |
184 |
|
185 |
foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) { |
186 |
|
187 |
my $format = x($x->{content}); |
188 |
my $delimiter = x($x->{delimiter}) || ' '; |
189 |
|
190 |
my $repeat_off = 0; # init repeatable offset |
191 |
|
192 |
# swish, swish_exact, display, index, index_lookup |
193 |
# swish and display defaults |
194 |
my ($s,$se,$d,$i,$il) = (1,0,1,0,0); |
195 |
$s = 0 if (lc($x->{type}) eq "display"); |
196 |
$d = 0 if (lc($x->{type}) eq "swish"); |
197 |
($s,$se,$d,$i) = (0,1,0,1) if (lc($x->{type}) eq "index"); |
198 |
($s,$se,$d,$i) = (0,1,0,0) if (lc($x->{type}) eq "swish_exact"); |
199 |
($s,$se,$d,$i,$il) = (0,1,0,0,1) if (lc($x->{type}) =~ /^lookup/); |
200 |
|
201 |
# what will separate last line from this one? |
202 |
if ($display_data && $x->{append}) { |
203 |
$line_delimiter = ' '; |
204 |
} elsif ($display_data) { |
205 |
$line_delimiter = '<br/>'; |
206 |
} |
207 |
|
208 |
# init vars so that we go into while... |
209 |
($swish,$display) = (1,1); |
210 |
|
211 |
# placeholder for all repeatable entries for index |
212 |
|
213 |
sub mkformat($$) { |
214 |
my $x = shift || die "mkformat needs tag reference"; |
215 |
my $data = shift || return; |
216 |
my $format_name = x($x->{format_name}) || return $data; |
217 |
my $fmt = x($config->{format}->{$format_name}->{content}) || die "<format name=\"$format_name\"> is not defined!"; |
218 |
my $format_delimiter = x($x->{format_delimiter}); |
219 |
my @data; |
220 |
if ($format_delimiter) { |
221 |
@data = split(/$format_delimiter/,$data); |
222 |
} else { |
223 |
push @data,$data; |
224 |
} |
225 |
|
226 |
if ($fmt) { |
227 |
my $nr = scalar $fmt =~ s/%s/%s/g; |
228 |
if (($#data+1) == $nr) { |
229 |
return sprintf($fmt,@data); |
230 |
} else { |
231 |
#print STDERR "mkformat: [$data] can't be split on [$format_delimiter] to $nr fields!\n"; |
232 |
return $data; |
233 |
} |
234 |
} else { |
235 |
print STDERR "usage of link '$format_name' without defined format (<link> tag)\n"; |
236 |
} |
237 |
} |
238 |
|
239 |
# while because of repeatable fields |
240 |
while ($swish || $display) { |
241 |
my $page = $repeat_off; |
242 |
$page_max{$field} = $page if ($iterate_by_page && $page > ($page_max{$field} || 0)); |
243 |
($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp); |
244 |
if ($repeat_off > 1000) { |
245 |
print STDERR "loop (more than 1000 repeatable fields) deteced in $row, $format\n"; |
246 |
last; |
247 |
} |
248 |
|
249 |
# is this field is lookup? |
250 |
if ($display && $x->{lookup}) { |
251 |
my $null = "<!-- null -->"; |
252 |
if ($use_lhash_cache) { |
253 |
if (!defined($cache->{lhash}->{$display})) { |
254 |
my $new_display = $lhash{$display}; |
255 |
if (defined($new_display)) { |
256 |
#print STDERR "lookup cache store '$display' = '$new_display'\n"; |
257 |
$display = $new_display; |
258 |
$cache->{lhash}->{$display} = $new_display; |
259 |
} else { |
260 |
# print STDERR "WARNING: lookup for '$display' didn't find anything.\n"; |
261 |
$display = ""; |
262 |
$cache->{lhash}->{$display} = $null; |
263 |
} |
264 |
} else { |
265 |
$display = $cache->{lhash}->{$display}; |
266 |
} |
267 |
} else { |
268 |
$display = $lhash{$display} || $null; |
269 |
} |
270 |
} |
271 |
|
272 |
# filter="name" ; filter this field through |
273 |
# filter/[name].pm |
274 |
my $filter = $x->{filter}; |
275 |
if ($filter && !$cache->{filter_loaded}->{$filter}) { |
276 |
require "filter/".$filter.".pm"; |
277 |
$cache->{filter_loaded}->{$filter}++; |
278 |
} |
279 |
# type="swish" ; field for swish |
280 |
if ($swish) { |
281 |
my $tmp = $swish; |
282 |
if ($filter && ($s || $se)) { |
283 |
no strict 'refs'; |
284 |
$tmp = join(" ",&$filter($tmp)) if ($s || $se); |
285 |
} |
286 |
|
287 |
$swish_data .= $tmp if ($s && $tmp); |
288 |
$swish_exact_data .= "xxbxx $tmp xxexx " if ($tmp && $tmp ne "" && $se); |
289 |
} |
290 |
|
291 |
# type="display" ; field for display |
292 |
if ($d && $display) { |
293 |
my $ldel = $delimiter; |
294 |
if ($line_delimiter && $display_data) { |
295 |
$ldel = $line_delimiter; |
296 |
} |
297 |
if ($filter) { |
298 |
no strict 'refs'; |
299 |
my @arr; |
300 |
foreach my $tmp (&$filter($display)) { |
301 |
my $tmp2 = mkformat($x,$tmp); |
302 |
push @arr,$tmp2 if ($tmp2); |
303 |
} |
304 |
$display_data .= $ldel if ($display_data && @arr); |
305 |
$display_data .= join($delimiter,@arr); |
306 |
} else { |
307 |
$display_data .= $ldel if ($display_data); |
308 |
my $tmp = mkformat($x,$display); |
309 |
$display_data .= $tmp if ($tmp); |
310 |
} |
311 |
} |
312 |
|
313 |
# type="index" ; insert into index |
314 |
my $idisplay; |
315 |
if ($i && $display) { |
316 |
$idisplay = $display; |
317 |
if ($filter) { |
318 |
no strict 'refs'; |
319 |
$idisplay = &$filter($idisplay); |
320 |
} |
321 |
push @index_data, $idisplay if ($idisplay && !$iterate_by_page); |
322 |
} |
323 |
|
324 |
# store fields in lookup |
325 |
if ($il && $display) { |
326 |
if (lc($x->{type}) eq "lookup_key") { |
327 |
if ($lookup_key) { |
328 |
print STDERR "WARNING: try to redefine lookup_key (keys shouldn't be repeatable fields!)"; |
329 |
} else { |
330 |
if ($filter) { |
331 |
no strict 'refs'; |
332 |
$lookup_key = &$filter($display); |
333 |
} else { |
334 |
$lookup_key = $display; |
335 |
} |
336 |
} |
337 |
} elsif (lc($x->{type}) eq "lookup_val") { |
338 |
if ($lookup_key) { |
339 |
if ($filter) { |
340 |
no strict 'refs'; |
341 |
$lhash{$lookup_key} = &$filter($display); |
342 |
} else { |
343 |
$lhash{$lookup_key} = $display; |
344 |
} |
345 |
} else { |
346 |
print STDERR "WARNING: no lookup_key defined for '$display'?"; |
347 |
} |
348 |
} |
349 |
|
350 |
} |
351 |
|
352 |
# store data for page-by-page repeatable fields |
353 |
if ($iterate_by_page) { |
354 |
sub iterate_fld($$$$$$) { |
355 |
my ($cache,$what,$field,$page,$data,$append) = @_; |
356 |
return if (!$data); |
357 |
|
358 |
my $ldel = $page_line_delimiter; |
359 |
$ldel = " " if ($append); |
360 |
#print STDERR "line delimiter: ",Dumper($ldel) if ($ldel); |
361 |
if (! $cache->{$what}->{$field}->[$page]) { |
362 |
$cache->{$what}->{$field}->[$page] = $data; |
363 |
} else { |
364 |
$cache->{$what}->{$field}->[$page] .= $ldel.$data; |
365 |
} |
366 |
} |
367 |
|
368 |
if ($display_data) { |
369 |
iterate_fld($cache,'display_data',$field,$page,$display_data,$x->{append}); |
370 |
} |
371 |
$display_data = ""; |
372 |
if ($swish_data) { |
373 |
iterate_fld($cache,'swish_data',$field,$page,$swish_data,$x->{append}); |
374 |
$swish_data = ""; |
375 |
} |
376 |
if ($swish_exact_data) { |
377 |
iterate_fld($cache,'swish_exact_data',$field,$page,$swish_exact_data,$x->{append}); |
378 |
$swish_exact_data = ""; |
379 |
} |
380 |
|
381 |
if ($idisplay) { |
382 |
my $ldel=$page_line_delimiter; |
383 |
my @index_data; |
384 |
if ($cache->{index_data}->{$field}->[$page]) { |
385 |
|
386 |
@index_data = @{$cache->{index_data}->{$field}->[$page]}; |
387 |
} |
388 |
if ($x->{append}) { |
389 |
if (@index_data) { |
390 |
$index_data[$#index_data] .= $idisplay; |
391 |
} else { |
392 |
push @index_data, $idisplay; |
393 |
} |
394 |
} else { |
395 |
push @index_data, $idisplay; |
396 |
} |
397 |
$idisplay = ""; |
398 |
@{$cache->{index_data}->{$field}->[$page]} = @index_data; |
399 |
} |
400 |
} |
401 |
} |
402 |
|
403 |
if (! $iterate_by_page) { |
404 |
my $idel = $x->{index_delimiter}; |
405 |
# fill data in index |
406 |
foreach my $tmp (@index_data) { |
407 |
my $i = $d = $tmp; |
408 |
if ($idel && $tmp =~ m/$idel/) { |
409 |
($i,$d) = split(/$idel/,$tmp); |
410 |
} |
411 |
$index->insert($field, $i, $d, $path); |
412 |
} |
413 |
@index_data = (); |
414 |
} |
415 |
} |
416 |
|
417 |
# now try to parse variables from configuration file |
418 |
foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) { |
419 |
|
420 |
my $delimiter = x($x->{delimiter}) || ' '; |
421 |
my $val = $cfg->val($database, x($x->{content})); |
422 |
|
423 |
my ($s,$d,$i) = (1,1,0); # swish, display default |
424 |
$s = 0 if (lc($x->{type}) eq "display"); |
425 |
$d = 0 if (lc($x->{type}) eq "swish"); |
426 |
# no support for swish exact in config. |
427 |
# IMHO, it's useless |
428 |
($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index"); |
429 |
|
430 |
if ($val) { |
431 |
$display_data .= $delimiter.$val if ($d); |
432 |
$swish_data .= $val if ($s); |
433 |
$index->insert($field, $val, $path) if ($i); |
434 |
} |
435 |
|
436 |
if ($iterate_by_page) { |
437 |
# FIXME data from config tag will appear just |
438 |
# on first page!!! |
439 |
my $page = 0; |
440 |
if ($display_data) { |
441 |
$cache->{display_data}->{$field}->[$page] = $display_data; |
442 |
$display_data = ""; |
443 |
} |
444 |
if ($swish_data) { |
445 |
$cache->{swish_data}->{$field}->[$page] = $swish_data; |
446 |
$swish_data = ""; |
447 |
} |
448 |
if ($swish_exact_data) { |
449 |
$cache->{swish_exact_data}->{$field}->[$page] = $swish_exact_data; |
450 |
$swish_exact_data = ""; |
451 |
} |
452 |
} |
453 |
} |
454 |
|
455 |
# save data page-by-page |
456 |
foreach my $field (@page_fields) { |
457 |
my $nr_pages = $page_max{$field} || next; |
458 |
#print STDERR "field '$field' iterate over ",($nr_pages || 0)," pages...\n"; |
459 |
#print STDERR Dumper($cache->{display_data}); |
460 |
for (my $page=0; $page <= $nr_pages; $page++) { |
461 |
my $display_data; |
462 |
if ($cache->{format}->{$field}) { |
463 |
my $tmp = mkformat($cache->{format}->{$field},$cache->{display_data}->{$field}->[$page]); |
464 |
$display_data=$tmp if ($tmp); |
465 |
} else { |
466 |
$display_data = $cache->{display_data}->{$field}->[$page]; |
467 |
} |
468 |
if ($display_data) { # default |
469 |
if ($field eq "headline") { |
470 |
$xml .= xmlify("headline", $display_data); |
471 |
} else { |
472 |
|
473 |
# fallback to empty field name if needed |
474 |
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
475 |
$html .= "#-#".$display_data."###\n"; |
476 |
} |
477 |
} |
478 |
|
479 |
my $swish_data = $cache->{swish_data}->{$field}->[$page]; |
480 |
if ($swish_data) { |
481 |
# remove extra spaces |
482 |
$swish_data =~ s/ +/ /g; |
483 |
$swish_data =~ s/ +$//g; |
484 |
|
485 |
$xml .= xmlify($field."_swish", unac_string($codepage,$swish_data)); |
486 |
} |
487 |
|
488 |
my $swish_exact_data = $cache->{swish_exact_data}->{$field}->[$page]; |
489 |
if ($swish_exact_data) { |
490 |
$swish_exact_data =~ s/ +/ /g; |
491 |
$swish_exact_data =~ s/ +$//g; |
492 |
|
493 |
# add delimiters before and after word. |
494 |
# That is required to produce exact match |
495 |
$xml .= xmlify($field."_swish_exact", unac_string($codepage,$swish_exact_data)); |
496 |
} |
497 |
|
498 |
my $idel = $cache->{index_delimiter}->{$field}; |
499 |
foreach my $tmp (@{$cache->{index_data}->{$field}->[$page]}) { |
500 |
my $i = $tmp; |
501 |
my $d = $tmp; |
502 |
if ($idel && $tmp =~ m/$idel/) { |
503 |
($i,$d) = split(/$idel/,$tmp); |
504 |
} |
505 |
$index->insert($field, $i, $d, $path); |
506 |
#print STDERR "index [$idel] $field: $i --> $d [$path]\n"; |
507 |
} |
508 |
} |
509 |
|
510 |
} |
511 |
|
512 |
if (! $iterate_by_page) { |
513 |
if ($display_data) { |
514 |
if ($field eq "headline") { |
515 |
$xml .= xmlify("headline", $display_data); |
516 |
} else { |
517 |
|
518 |
# fallback to empty field name if needed |
519 |
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
520 |
$html .= "#-#".$display_data."###\n"; |
521 |
} |
522 |
} |
523 |
if ($swish_data) { |
524 |
# remove extra spaces |
525 |
$swish_data =~ s/ +/ /g; |
526 |
$swish_data =~ s/ +$//g; |
527 |
|
528 |
$xml .= xmlify($field."_swish", unac_string($codepage,$swish_data)); |
529 |
} |
530 |
|
531 |
if ($swish_exact_data) { |
532 |
$swish_exact_data =~ s/ +/ /g; |
533 |
$swish_exact_data =~ s/ +$//g; |
534 |
|
535 |
# add delimiters before and after word. |
536 |
# That is required to produce exact match |
537 |
$xml .= xmlify($field."_swish_exact", unac_string($codepage,$swish_exact_data)); |
538 |
} |
539 |
} |
540 |
} |
541 |
|
542 |
# dump formatted output in <html> |
543 |
if ($html) { |
544 |
#$xml .= xmlify("html",$html); |
545 |
$xml .= "<html><![CDATA[ $html ]]></html>"; |
546 |
} |
547 |
|
548 |
if ($xml) { |
549 |
$xml .= $add_xml if ($add_xml); |
550 |
return "<xml>\n$xml</xml>\n"; |
551 |
} else { |
552 |
return; |
553 |
} |
554 |
} |
555 |
|
556 |
########################################################################## |
557 |
|
558 |
# read configuration for this script |
559 |
my $cfg = new Config::IniFiles( -file => $config_file ); |
560 |
|
561 |
# read global.conf configuration |
562 |
my $cfg_global = new Config::IniFiles( -file => 'global.conf' ); |
563 |
|
564 |
# open index |
565 |
$index = new index_DBI( |
566 |
$cfg_global->val('global', 'dbi_dbd'), |
567 |
$cfg_global->val('global', 'dbi_dsn'), |
568 |
$cfg_global->val('global', 'dbi_user'), |
569 |
$cfg_global->val('global', 'dbi_passwd') || '', |
570 |
); |
571 |
|
572 |
my $show_progress = $cfg_global->val('global', 'show_progress'); |
573 |
|
574 |
my $unac_filter = $cfg_global->val('global', 'unac_filter'); |
575 |
if ($unac_filter) { |
576 |
require $unac_filter; |
577 |
} |
578 |
|
579 |
foreach my $database ($cfg->Sections) { |
580 |
|
581 |
my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; |
582 |
my $add_xml = $cfg -> val($database, 'xml'); # optional |
583 |
|
584 |
# create new lookup file |
585 |
my $lookup_file = $cfg -> val($database, 'lookup_newfile'); # optional |
586 |
if ($lookup_file) { |
587 |
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_NEWDB, 0644; |
588 |
tie %lhash, 'TDB_File', $lookup_file, TDB_CLEAR_IF_FIRST, O_RDWR, 0644; |
589 |
print STDERR "creating lookup file '$lookup_file'\n"; |
590 |
# delete memory cache for lookup file |
591 |
delete $cache->{lhash}; |
592 |
} |
593 |
|
594 |
# open existing lookup file |
595 |
$lookup_file = $cfg -> val($database, 'lookup_open'); # optional |
596 |
if ($lookup_file) { |
597 |
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_READER, 0644; |
598 |
tie %lhash, 'TDB_File', $lookup_file, TDB_DEFAULT, O_RDWR, 0644; |
599 |
print STDERR "opening lookup file '$lookup_file'\n"; |
600 |
} |
601 |
|
602 |
print STDERR "reading ./import_xml/$type.xml\n"; |
603 |
|
604 |
# extract just type basic |
605 |
my $type_base = $type; |
606 |
$type_base =~ s/_.+$//g; |
607 |
|
608 |
$config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base}, 'config', 'format' ], forcecontent => 1); |
609 |
|
610 |
# output current progress indicator |
611 |
my $last_p = 0; |
612 |
sub progress { |
613 |
return if (! $show_progress); |
614 |
my $current = shift; |
615 |
my $total = shift || 1; |
616 |
my $p = int($current * 100 / $total); |
617 |
if ($p != $last_p) { |
618 |
printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$current,$total,"=" x ($p/2).">", $p ); |
619 |
$last_p = $p; |
620 |
} |
621 |
} |
622 |
|
623 |
my $fake_dir = 1; |
624 |
sub fakeprogress { |
625 |
return if (! $show_progress); |
626 |
my $current = shift @_; |
627 |
|
628 |
my @ind = ('-','\\','|','/','-','\\','|','/', '-'); |
629 |
|
630 |
$last_p += $fake_dir; |
631 |
$fake_dir = -$fake_dir if ($last_p > 1000 || $last_p < 0); |
632 |
if ($last_p % 10 == 0) { |
633 |
printf STDERR ("%5d / %5s [%-51s]\r",$current,"?"," " x ($last_p/20).$ind[($last_p/20) % $#ind]); |
634 |
} |
635 |
} |
636 |
|
637 |
# now read database |
638 |
print STDERR "using: $type...\n"; |
639 |
|
640 |
# erase cache for tags by order in this database |
641 |
delete $cache->{tags_by_order}; |
642 |
|
643 |
if ($type_base eq "isis") { |
644 |
|
645 |
my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!"; |
646 |
|
647 |
$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage); |
648 |
my $db = OpenIsis::open( $isis_db ); |
649 |
|
650 |
# check if .txt database for OpenIsis is zero length, |
651 |
# if so, erase it and re-open database |
652 |
sub check_txt_db { |
653 |
my $isis_db = shift || die "need isis database name"; |
654 |
my $reopen = 0; |
655 |
|
656 |
if (-e $isis_db.".TXT") { |
657 |
print STDERR "WARNING: removing $isis_db.TXT OpenIsis database...\n"; |
658 |
unlink $isis_db.".TXT" || warn "FATAL: unlink error on '$isis_db.TXT': $!"; |
659 |
$reopen++; |
660 |
} |
661 |
if (-e $isis_db.".PTR") { |
662 |
print STDERR "WARNING: removing $isis_db.PTR OpenIsis database...\n"; |
663 |
unlink $isis_db.".PTR" || warn "FATAL: unlink error on '$isis_db.PTR': $!"; |
664 |
$reopen++; |
665 |
} |
666 |
return OpenIsis::open( $isis_db ) if ($reopen); |
667 |
} |
668 |
|
669 |
# EOF error |
670 |
if ($db == -1) { |
671 |
$db = check_txt_db($isis_db); |
672 |
if ($db == -1) { |
673 |
print STDERR "FATAL: OpenIsis can't open zero size file $isis_db\n"; |
674 |
next; |
675 |
} |
676 |
} |
677 |
|
678 |
# OpenIsis::ERR_BADF |
679 |
if ($db == -4) { |
680 |
print STDERR "FATAL: OpenIsis can't find file $isis_db\n"; |
681 |
next; |
682 |
# OpenIsis::ERR_IO |
683 |
} elsif ($db == -5) { |
684 |
print STDERR "FATAL: OpenIsis can't access file $isis_db\n"; |
685 |
next; |
686 |
} elsif ($db < 0) { |
687 |
print STDERR "FATAL: OpenIsis unknown error $db with file $isis_db\n"; |
688 |
next; |
689 |
} |
690 |
|
691 |
my $max_rowid = OpenIsis::maxRowid( $db ); |
692 |
|
693 |
# if 0 records, try to rease isis .txt database |
694 |
if ($max_rowid == 0) { |
695 |
# force removal of database |
696 |
$db = check_txt_db($isis_db); |
697 |
$max_rowid = OpenIsis::maxRowid( $db ); |
698 |
} |
699 |
|
700 |
print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; |
701 |
|
702 |
my $path = $database; |
703 |
|
704 |
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { |
705 |
my $row = OpenIsis::read( $db, $row_id ); |
706 |
if ($row && $row->{mfn}) { |
707 |
|
708 |
progress($row->{mfn}, $max_rowid); |
709 |
|
710 |
my $swishpath = $path."#".int($row->{mfn}); |
711 |
|
712 |
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
713 |
$xml = $cp2utf->convert($xml); |
714 |
use bytes; # as opposed to chars |
715 |
print "Path-Name: $swishpath\n"; |
716 |
print "Content-Length: ".(length($xml)+1)."\n"; |
717 |
print "Document-Type: XML\n\n$xml\n"; |
718 |
} |
719 |
} |
720 |
} |
721 |
# for this to work with current version of OpenIsis (0.9.0) |
722 |
# you might need my patch from |
723 |
# http://www.rot13.org/~dpavlin/projects/openisis-0.9.0-perl_close.diff |
724 |
OpenIsis::close($db); |
725 |
print STDERR "\n"; |
726 |
|
727 |
} elsif ($type_base eq "excel") { |
728 |
require Spreadsheet::ParseExcel; |
729 |
require Spreadsheet::ParseExcel::Utility; |
730 |
import Spreadsheet::ParseExcel::Utility qw(int2col); |
731 |
|
732 |
$import2cp = Text::Iconv->new($config->{excel_codepage},$codepage); |
733 |
my $excel_file = $cfg -> val($database, 'excel_file') || die "$database doesn't have 'excel_file' defined!"; |
734 |
|
735 |
my $sheet = x($config->{sheet}) || die "no sheet in $type.xml"; |
736 |
my $start_row = x($config->{start_row}) - 1 || die "no start_row in $type.xml"; |
737 |
|
738 |
my $oBook = Spreadsheet::ParseExcel::Workbook->Parse($excel_file) || die "can't open Excel file '$excel_file'"; |
739 |
|
740 |
my $sheet_nr = 0; |
741 |
foreach my $oWks (@{$oBook->{Worksheet}}) { |
742 |
#print STDERR "-- SHEET $sheet_nr:", $oWks->{Name}, "\n"; |
743 |
last if ($oWks->{Name} eq $sheet); |
744 |
$sheet_nr++; |
745 |
} |
746 |
|
747 |
my $oWorksheet = $oBook->{Worksheet}[$sheet_nr]; |
748 |
|
749 |
print STDERR "using sheet: ",$oWorksheet->{Name},"\n"; |
750 |
defined ($oWorksheet) || die "can't find sheet '$sheet' in $excel_file"; |
751 |
my $end_row = x($config->{end_row}) || $oWorksheet->{MaxRow}; |
752 |
|
753 |
for(my $iR = $start_row ; defined $end_row && $iR <= $end_row ; $iR++) { |
754 |
my $row; |
755 |
for(my $iC = $oWorksheet->{MinCol} ; defined $oWorksheet->{MaxCol} && $iC <= $oWorksheet->{MaxCol} ; $iC++) { |
756 |
my $cell = $oWorksheet->{Cells}[$iR][$iC]; |
757 |
if ($cell) { |
758 |
$row->{int2col($iC)} = $cell->Value; |
759 |
} |
760 |
} |
761 |
|
762 |
progress($iR, $end_row); |
763 |
|
764 |
# print "row[$iR/$end_row] "; |
765 |
# foreach (keys %{$row}) { |
766 |
# print "$_: ",$row->{$_},"\t"; |
767 |
# } |
768 |
# print "\n"; |
769 |
|
770 |
my $swishpath = $database."#".$iR; |
771 |
|
772 |
next if (! $row); |
773 |
|
774 |
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
775 |
$xml = $cp2utf->convert($xml); |
776 |
use bytes; # as opposed to chars |
777 |
print "Path-Name: $swishpath\n"; |
778 |
print "Content-Length: ".(length($xml)+1)."\n"; |
779 |
print "Document-Type: XML\n\n$xml\n"; |
780 |
} |
781 |
} |
782 |
} elsif ($type_base eq "marc") { |
783 |
|
784 |
require MARC; |
785 |
|
786 |
$import2cp = Text::Iconv->new($config->{marc_codepage},$codepage); |
787 |
my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!"; |
788 |
|
789 |
# optional argument is format |
790 |
my $format = x($config->{marc_format}) || 'usmarc'; |
791 |
|
792 |
print STDERR "Reading MARC file '$marc_file'\n"; |
793 |
|
794 |
my $marc = new MARC; |
795 |
my $nr = $marc->openmarc({ |
796 |
file=>$marc_file, format=>$format |
797 |
}) || die "Can't open MARC file '$marc_file' with format '$format'"; |
798 |
|
799 |
# read MARC file in memory |
800 |
$marc->nextmarc(-1); |
801 |
|
802 |
my $max_rec = $marc->marc_count(); |
803 |
|
804 |
for(my $i=1; $i<=$max_rec; $i++) { |
805 |
|
806 |
progress($i,$max_rec); |
807 |
|
808 |
# store value for marc_sf.pm |
809 |
$main::cache->{marc_record} = $i; |
810 |
|
811 |
my $swishpath = $database."#".$i; |
812 |
|
813 |
if (my $xml = data2xml($type_base,$marc,$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 "feed") { |
825 |
|
826 |
$import2cp = Text::Iconv->new($config->{feed_codepage},$codepage); |
827 |
my $prog = x($config->{prog}) || die "$database doesn't have 'prog' defined!"; |
828 |
|
829 |
print STDERR "Reading feed from program '$prog'\n"; |
830 |
|
831 |
open(FEED,"feeds/$prog |") || die "can't start $prog: $!"; |
832 |
|
833 |
my $i=1; # record nr. |
834 |
|
835 |
my $data; |
836 |
my $line=1; |
837 |
|
838 |
while (<FEED>) { |
839 |
chomp; |
840 |
|
841 |
if (/^$/) { |
842 |
my $swishpath = $database."#".$i++; |
843 |
|
844 |
if (my $xml = data2xml($type_base,$data,$add_xml,$cfg,$database)) { |
845 |
$xml = $cp2utf->convert($xml); |
846 |
use bytes; # as opposed to chars |
847 |
print "Path-Name: $swishpath\n"; |
848 |
print "Content-Length: ".(length($xml)+1)."\n"; |
849 |
print "Document-Type: XML\n\n$xml\n"; |
850 |
} |
851 |
$line = 1; |
852 |
$data = {}; |
853 |
next; |
854 |
} |
855 |
|
856 |
$line = $1 if (s/^(\d+):\s*//); |
857 |
$data->{$line++} = $_; |
858 |
|
859 |
fakeprogress($i); |
860 |
|
861 |
} |
862 |
# close lookup |
863 |
untie %lhash if (%lhash); |
864 |
} |
865 |
} |
866 |
|
867 |
# call this to commit index |
868 |
$index->close; |
869 |
|
870 |
1; |
871 |
__END__ |
872 |
########################################################################## |
873 |
|
874 |
=head1 NAME |
875 |
|
876 |
all2xml.pl - read various file formats and dump XML for SWISH-E |
877 |
|
878 |
=head1 DESCRIPTION |
879 |
|
880 |
This command will read ISIS data file using OpenIsis perl module, MARC |
881 |
records using MARC module and optionally Micro$oft Excel files to |
882 |
create one XML file for usage with I<SWISH-E> indexer. Dispite it's name, |
883 |
this script B<isn't general xml generator> from isis files (isis allready |
884 |
has something like that). Output of this script is tailor-made for SWISH-E. |
885 |
|
886 |
=head1 BUGS |
887 |
|
888 |
Documentation is really lacking. However, in true Open Source spirit, source |
889 |
is best documentation. I even made considerable effort to comment parts |
890 |
which are not intuitively clear, so... |
891 |
|
892 |
=head1 AUTHOR |
893 |
|
894 |
Dobrica Pavlinusic <dpavlin@rot13.org> |
895 |
|
896 |
=head1 COPYRIGHT |
897 |
|
898 |
GNU Public License (GPL) v2 or later |
899 |
|
900 |
=head1 SEE ALSO |
901 |
|
902 |
SWISH-E web site at http://www.swish-e.org |
903 |
|
904 |
=cut |