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