1 |
#!/usr/bin/perl -w |
2 |
|
3 |
use strict; |
4 |
use IsisDB; |
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 |
|
15 |
$|=1; |
16 |
|
17 |
my $config_file = $0; |
18 |
$config_file =~ s/\.pl$/.conf/; |
19 |
$config_file = $ARGV[0] if ($ARGV[0] && -f $ARGV[0]); |
20 |
die "FATAL: can't find configuration file '$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 |
|
143 |
if ($field_name) { |
144 |
$field_name = x($field_name); |
145 |
if (! $last_field_name) { |
146 |
$last_field_name = $field_name; |
147 |
return $last_field_name; |
148 |
} elsif ($field_name ne $last_field_name) { |
149 |
$last_field_name = $field_name; |
150 |
return $last_field_name; |
151 |
} |
152 |
} |
153 |
} |
154 |
|
155 |
|
156 |
# init variables for different types |
157 |
sub init_visible_type($) { |
158 |
my $type = shift; |
159 |
|
160 |
# swish, swish_exact, display, index, index_lookup |
161 |
# swish and display defaults |
162 |
my ($s,$se,$d,$i,$il) = (1,0,1,0,0); |
163 |
if (lc($type) eq "display") { |
164 |
$s = 0; |
165 |
} elsif (lc($type) eq "swish") { |
166 |
$d = 0; |
167 |
} elsif (lc($type) eq "index") { |
168 |
($s,$se,$d,$i) = (0,1,0,1); |
169 |
} elsif (lc($type) eq "swish_exact") { |
170 |
($s,$se,$d,$i) = (0,1,0,0); |
171 |
} elsif (lc($type) =~ /^lookup/) { |
172 |
($s,$se,$d,$i,$il) = (0,1,0,0,1); |
173 |
} elsif ($type) { |
174 |
print STDERR "WARNING: unknown type: $type\n"; |
175 |
} |
176 |
return ($s,$se,$d,$i,$il); |
177 |
} |
178 |
|
179 |
|
180 |
# convert |
181 |
# |
182 |
# <tag> |
183 |
# <delimiter>, </delimiter> |
184 |
# <value>200a</value> |
185 |
# </tag> |
186 |
# |
187 |
# to |
188 |
# |
189 |
# <tag delimiter=", ">200a</tag> |
190 |
# |
191 |
# but without loosing spaces in delimiter (becasue |
192 |
# new XML::Simple strips spaces in attribute values |
193 |
# as defined in XML specification) |
194 |
# |
195 |
sub unroll_x($) { |
196 |
my $x = shift; |
197 |
|
198 |
if (defined $x->{value}) { |
199 |
my ($v,$d) = ($x->{value}->{content}, $x->{delimiter}->{content}); |
200 |
delete $x->{value}; |
201 |
delete $x->{delimiter}; |
202 |
$x->{content} = $v; |
203 |
$x->{delimiter} = $d; |
204 |
} |
205 |
return $x; |
206 |
} |
207 |
|
208 |
# begin real work: go field by field |
209 |
foreach my $field (@sorted_tags) { |
210 |
|
211 |
$field=x($field); |
212 |
$field_usage{$field}++; |
213 |
|
214 |
my $swish_data = ""; |
215 |
my $swish_exact_data = ""; |
216 |
my $display_data = ""; |
217 |
my @index_data; |
218 |
my $line_delimiter; |
219 |
|
220 |
my ($swish,$display); |
221 |
|
222 |
my $tag = $type2tag{$type} || die "can't find which tag to use for type $type"; |
223 |
|
224 |
# is this field page-by-page? |
225 |
my $iterate_by_page = $config->{indexer}->{$field}->{iterate_by_page}; |
226 |
push @page_fields,$field if ($iterate_by_page); |
227 |
my %page_max = (); |
228 |
# default line_delimiter if using |
229 |
my $page_line_delimiter = $config->{indexer}->{$field}->{page_line_delimiter} || '<br/>'; |
230 |
$cache->{index_delimiter}->{$field} = $config->{indexer}->{$field}->{index_delimiter}; |
231 |
|
232 |
my $format_name = $config->{indexer}->{$field}->{format_name}; |
233 |
my $format_delimiter = $config->{indexer}->{$field}->{format_delimiter}; |
234 |
if ($format_name && $format_delimiter) { |
235 |
$cache->{format}->{$field}->{format_name} = $format_name; |
236 |
$cache->{format}->{$field}->{format_delimiter} = $format_delimiter; |
237 |
} |
238 |
|
239 |
foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) { |
240 |
|
241 |
$x = unroll_x($x); |
242 |
|
243 |
my $format = x($x->{content}); |
244 |
my $delimiter = x($x->{delimiter}) || ' '; |
245 |
|
246 |
my $repeat_off = 0; # init repeatable offset |
247 |
|
248 |
my ($s,$se,$d,$i,$il) = init_visible_type($x->{type}); |
249 |
|
250 |
# what will separate last line from this one? |
251 |
if ($display_data && $x->{append}) { |
252 |
$line_delimiter = $delimiter; |
253 |
} elsif ($display_data) { |
254 |
$line_delimiter = '<br/>'; |
255 |
} |
256 |
|
257 |
# init vars so that we go into while... |
258 |
($swish,$display) = (1,1); |
259 |
|
260 |
# placeholder for all repeatable entries for index |
261 |
|
262 |
sub mkformat($$) { |
263 |
my $x = shift || die "mkformat needs tag reference"; |
264 |
my $data = shift || return; |
265 |
my $format_name = x($x->{format_name}) || return $data; |
266 |
my $fmt = x($config->{format}->{$format_name}->{content}) || die "<format name=\"$format_name\"> is not defined!"; |
267 |
my $format_delimiter = x($x->{format_delimiter}); |
268 |
my @data; |
269 |
if ($format_delimiter) { |
270 |
@data = split(/$format_delimiter/,$data); |
271 |
} else { |
272 |
push @data,$data; |
273 |
} |
274 |
|
275 |
if ($fmt) { |
276 |
my $nr = scalar $fmt =~ s/%s/%s/g; |
277 |
if (($#data+1) == $nr) { |
278 |
return sprintf($fmt,@data); |
279 |
} else { |
280 |
#print STDERR "mkformat: [$data] can't be split on [$format_delimiter] to $nr fields!\n"; |
281 |
return $data; |
282 |
} |
283 |
} else { |
284 |
print STDERR "usage of link '$format_name' without defined format (<link> tag)\n"; |
285 |
} |
286 |
} |
287 |
|
288 |
# while because of repeatable fields |
289 |
while ($swish || $display) { |
290 |
my $page = $repeat_off; |
291 |
$page_max{$field} = $page if ($iterate_by_page && $page > ($page_max{$field} || 0)); |
292 |
($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp); |
293 |
if ($repeat_off > 1000) { |
294 |
print STDERR "loop (more than 1000 repeatable fields) deteced in $row, $format\n"; |
295 |
last; |
296 |
} |
297 |
|
298 |
# is this field is lookup? |
299 |
if ($display && $x->{lookup}) { |
300 |
my $null = "<!-- null -->"; |
301 |
if ($use_lhash_cache) { |
302 |
if (!defined($cache->{lhash}->{$display})) { |
303 |
my $new_display = $lhash{$display}; |
304 |
if (defined($new_display)) { |
305 |
#print STDERR "lookup cache store '$display' = '$new_display'\n"; |
306 |
$display = $new_display; |
307 |
$cache->{lhash}->{$display} = $new_display; |
308 |
} else { |
309 |
# print STDERR "WARNING: lookup for '$display' didn't find anything.\n"; |
310 |
$display = ""; |
311 |
$cache->{lhash}->{$display} = $null; |
312 |
} |
313 |
} else { |
314 |
$display = $cache->{lhash}->{$display}; |
315 |
} |
316 |
} else { |
317 |
$display = $lhash{$display} || $null; |
318 |
} |
319 |
} |
320 |
|
321 |
# filter="name" ; filter this field through |
322 |
# filter/[name].pm |
323 |
my $filter = $x->{filter}; |
324 |
if ($filter && !$cache->{filter_loaded}->{$filter}) { |
325 |
require "filter/".$filter.".pm"; |
326 |
$cache->{filter_loaded}->{$filter}++; |
327 |
} |
328 |
# type="swish" ; field for swish |
329 |
if ($swish) { |
330 |
my $tmp = $swish; |
331 |
if ($filter && ($s || $se)) { |
332 |
no strict 'refs'; |
333 |
$tmp = join(" ",&$filter($tmp)) if ($s || $se); |
334 |
} |
335 |
|
336 |
$swish_data .= $tmp if ($s && $tmp); |
337 |
$swish_exact_data .= "xxbxx $tmp xxexx " if ($tmp && $tmp ne "" && $se); |
338 |
} |
339 |
|
340 |
# type="display" ; field for display |
341 |
if ($d && $display) { |
342 |
my $ldel = $delimiter; |
343 |
if ($line_delimiter && $display_data) { |
344 |
$ldel = $line_delimiter; |
345 |
} |
346 |
if ($filter) { |
347 |
no strict 'refs'; |
348 |
my @arr; |
349 |
foreach my $tmp (&$filter($display)) { |
350 |
my $tmp2 = mkformat($x,$tmp); |
351 |
push @arr,$tmp2 if ($tmp2); |
352 |
} |
353 |
$display_data .= $ldel if ($display_data && @arr); |
354 |
$display_data .= join($delimiter,@arr); |
355 |
} else { |
356 |
$display_data .= $ldel if ($display_data); |
357 |
my $tmp = mkformat($x,$display); |
358 |
$display_data .= $tmp if ($tmp); |
359 |
} |
360 |
} |
361 |
|
362 |
# type="index" ; insert into index |
363 |
my $idisplay; |
364 |
if ($i && $display) { |
365 |
$idisplay = $display; |
366 |
if ($filter) { |
367 |
no strict 'refs'; |
368 |
$idisplay = &$filter($idisplay); |
369 |
} |
370 |
push @index_data, $idisplay if ($idisplay && !$iterate_by_page); |
371 |
} |
372 |
|
373 |
# store fields in lookup |
374 |
if ($il && $display) { |
375 |
if (lc($x->{type}) eq "lookup_key") { |
376 |
if ($lookup_key) { |
377 |
print STDERR "WARNING: try to redefine lookup_key (keys shouldn't be repeatable fields!)"; |
378 |
} else { |
379 |
if ($filter) { |
380 |
no strict 'refs'; |
381 |
$lookup_key = &$filter($display); |
382 |
} else { |
383 |
$lookup_key = $display; |
384 |
} |
385 |
} |
386 |
} elsif (lc($x->{type}) eq "lookup_val") { |
387 |
if ($lookup_key) { |
388 |
if ($filter) { |
389 |
no strict 'refs'; |
390 |
$lhash{$lookup_key} = &$filter($display); |
391 |
} else { |
392 |
$lhash{$lookup_key} = $display; |
393 |
} |
394 |
} else { |
395 |
print STDERR "WARNING: no lookup_key defined for '$display'?"; |
396 |
} |
397 |
} |
398 |
|
399 |
} |
400 |
|
401 |
# store data for page-by-page repeatable fields |
402 |
if ($iterate_by_page) { |
403 |
sub iterate_fld($$$$$$) { |
404 |
my ($cache,$what,$field,$page,$data,$append) = @_; |
405 |
return if (!$data); |
406 |
|
407 |
my $ldel = $page_line_delimiter; |
408 |
$ldel = " " if ($append); |
409 |
#print STDERR "line delimiter: ",Dumper($ldel) if ($ldel); |
410 |
if (! $cache->{$what}->{$field}->[$page]) { |
411 |
$cache->{$what}->{$field}->[$page] = $data; |
412 |
} else { |
413 |
$cache->{$what}->{$field}->[$page] .= $ldel.$data; |
414 |
} |
415 |
} |
416 |
|
417 |
if ($display_data) { |
418 |
iterate_fld($cache,'display_data',$field,$page,$display_data,$x->{append}); |
419 |
} |
420 |
$display_data = ""; |
421 |
if ($swish_data) { |
422 |
iterate_fld($cache,'swish_data',$field,$page,$swish_data,$x->{append}); |
423 |
$swish_data = ""; |
424 |
} |
425 |
if ($swish_exact_data) { |
426 |
iterate_fld($cache,'swish_exact_data',$field,$page,$swish_exact_data,$x->{append}); |
427 |
$swish_exact_data = ""; |
428 |
} |
429 |
|
430 |
if ($idisplay) { |
431 |
my $ldel=$page_line_delimiter; |
432 |
my @index_data; |
433 |
if ($cache->{index_data}->{$field}->[$page]) { |
434 |
|
435 |
@index_data = @{$cache->{index_data}->{$field}->[$page]}; |
436 |
} |
437 |
if ($x->{append}) { |
438 |
if (@index_data) { |
439 |
$index_data[$#index_data] .= $idisplay; |
440 |
} else { |
441 |
push @index_data, $idisplay; |
442 |
} |
443 |
} else { |
444 |
push @index_data, $idisplay; |
445 |
} |
446 |
$idisplay = ""; |
447 |
@{$cache->{index_data}->{$field}->[$page]} = @index_data; |
448 |
} |
449 |
} |
450 |
} |
451 |
|
452 |
if (! $iterate_by_page) { |
453 |
my $idel = $x->{index_delimiter}; |
454 |
# fill data in index |
455 |
foreach my $tmp (@index_data) { |
456 |
my $i = $d = $tmp; |
457 |
if ($idel && $tmp =~ m/$idel/) { |
458 |
($i,$d) = split(/$idel/,$tmp); |
459 |
} |
460 |
$index->insert($field, $i, $d, $path); |
461 |
} |
462 |
@index_data = (); |
463 |
} |
464 |
} |
465 |
|
466 |
# now try to parse variables from configuration file |
467 |
foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) { |
468 |
|
469 |
$x = unroll_x($x); |
470 |
|
471 |
my $delimiter = x($x->{delimiter}) || ' '; |
472 |
my $val = $cfg->val($database, x($x->{content})); |
473 |
|
474 |
# FIXME index_lookup is not supported! |
475 |
my ($s,$se,$d,$i,$il) = init_visible_type($x->{type}); |
476 |
|
477 |
if ($val) { |
478 |
$display_data .= $delimiter.$val if ($d); |
479 |
$swish_data .= " ".$val if ($s); |
480 |
$index->insert($field, $val, $path) if ($i); |
481 |
} |
482 |
|
483 |
if ($iterate_by_page) { |
484 |
# FIXME data from config tag will appear just |
485 |
# on first page!!! |
486 |
my $page = 0; |
487 |
if ($display_data) { |
488 |
$cache->{display_data}->{$field}->[$page] = $display_data; |
489 |
$display_data = ""; |
490 |
} |
491 |
if ($swish_data) { |
492 |
$cache->{swish_data}->{$field}->[$page] = $swish_data; |
493 |
$swish_data = ""; |
494 |
} |
495 |
if ($swish_exact_data) { |
496 |
$cache->{swish_exact_data}->{$field}->[$page] = $swish_exact_data; |
497 |
$swish_exact_data = ""; |
498 |
} |
499 |
} |
500 |
} |
501 |
|
502 |
# save data page-by-page |
503 |
foreach my $field (@page_fields) { |
504 |
my $nr_pages = $page_max{$field} || next; |
505 |
#print STDERR "field '$field' iterate over ",($nr_pages || 0)," pages...\n"; |
506 |
#print STDERR Dumper($cache->{display_data}); |
507 |
for (my $page=0; $page <= $nr_pages; $page++) { |
508 |
my $display_data; |
509 |
if ($cache->{format}->{$field}) { |
510 |
my $tmp = mkformat($cache->{format}->{$field},$cache->{display_data}->{$field}->[$page]); |
511 |
$display_data=$tmp if ($tmp); |
512 |
} else { |
513 |
$display_data = $cache->{display_data}->{$field}->[$page]; |
514 |
} |
515 |
if ($display_data) { # default |
516 |
if ($field eq "headline") { |
517 |
$xml .= xmlify("headline", $display_data); |
518 |
} else { |
519 |
|
520 |
# fallback to empty field name if needed |
521 |
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
522 |
$html .= "#-#".$display_data."###\n"; |
523 |
} |
524 |
} |
525 |
|
526 |
my $swish_data = $cache->{swish_data}->{$field}->[$page]; |
527 |
if ($swish_data) { |
528 |
# remove extra spaces |
529 |
$swish_data =~ s/ +/ /g; |
530 |
$swish_data =~ s/ +$//g; |
531 |
|
532 |
$xml .= xmlify($field."_swish", my_unac_string($codepage,$swish_data)); |
533 |
} |
534 |
|
535 |
my $swish_exact_data = $cache->{swish_exact_data}->{$field}->[$page]; |
536 |
if ($swish_exact_data) { |
537 |
$swish_exact_data =~ s/ +/ /g; |
538 |
$swish_exact_data =~ s/ +$//g; |
539 |
|
540 |
# add delimiters before and after word. |
541 |
# That is required to produce exact match |
542 |
$xml .= xmlify($field."_swish_exact", my_unac_string($codepage,$swish_exact_data)); |
543 |
} |
544 |
|
545 |
my $idel = $cache->{index_delimiter}->{$field}; |
546 |
foreach my $tmp (@{$cache->{index_data}->{$field}->[$page]}) { |
547 |
my $i = $tmp; |
548 |
my $d = $tmp; |
549 |
if ($idel && $tmp =~ m/$idel/) { |
550 |
($i,$d) = split(/$idel/,$tmp); |
551 |
} |
552 |
$index->insert($field, $i, $d, $path); |
553 |
#print STDERR "index [$idel] $field: $i --> $d [$path]\n"; |
554 |
} |
555 |
} |
556 |
|
557 |
} |
558 |
|
559 |
if (! $iterate_by_page) { |
560 |
if ($display_data) { |
561 |
if ($field eq "headline") { |
562 |
$xml .= xmlify("headline", $display_data); |
563 |
} else { |
564 |
|
565 |
# fallback to empty field name if needed |
566 |
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
567 |
$html .= "#-#".$display_data."###\n"; |
568 |
} |
569 |
} |
570 |
if ($swish_data) { |
571 |
# remove extra spaces |
572 |
$swish_data =~ s/ +/ /g; |
573 |
$swish_data =~ s/ +$//g; |
574 |
|
575 |
$xml .= xmlify($field."_swish", my_unac_string($codepage,$swish_data)); |
576 |
} |
577 |
|
578 |
if ($swish_exact_data) { |
579 |
$swish_exact_data =~ s/ +/ /g; |
580 |
$swish_exact_data =~ s/ +$//g; |
581 |
|
582 |
# add delimiters before and after word. |
583 |
# That is required to produce exact match |
584 |
$xml .= xmlify($field."_swish_exact", my_unac_string($codepage,$swish_exact_data)); |
585 |
} |
586 |
} |
587 |
} |
588 |
|
589 |
# dump formatted output in <html> |
590 |
if ($html) { |
591 |
#$xml .= xmlify("html",$html); |
592 |
$xml .= "<html><![CDATA[ $html ]]></html>"; |
593 |
} |
594 |
|
595 |
if ($xml) { |
596 |
$xml .= $add_xml if ($add_xml); |
597 |
return "<xml>\n$xml</xml>\n"; |
598 |
} else { |
599 |
return; |
600 |
} |
601 |
} |
602 |
|
603 |
########################################################################## |
604 |
|
605 |
# read configuration for this script |
606 |
my $cfg = new Config::IniFiles( -file => $config_file ); |
607 |
|
608 |
# read global.conf configuration |
609 |
my $cfg_global = new Config::IniFiles( -file => 'global.conf' ); |
610 |
|
611 |
# open index |
612 |
$index = new index_DBI( |
613 |
$cfg_global->val('global', 'dbi_dbd'), |
614 |
$cfg_global->val('global', 'dbi_dsn'), |
615 |
$cfg_global->val('global', 'dbi_user'), |
616 |
$cfg_global->val('global', 'dbi_passwd') || '', |
617 |
); |
618 |
|
619 |
my $show_progress = $cfg_global->val('global', 'show_progress'); |
620 |
|
621 |
my $my_unac_filter = $cfg_global->val('global', 'my_unac_filter'); |
622 |
if ($my_unac_filter) { |
623 |
print STDERR "using $my_unac_filter to filter characters for search\n"; |
624 |
require $my_unac_filter; |
625 |
} else { |
626 |
print STDERR "### fallback to default my_unac_string!\n"; |
627 |
eval q{ |
628 |
sub main::my_unac_string($$) { |
629 |
my ($charset, $string) = (@_); |
630 |
return $string; |
631 |
} |
632 |
}; |
633 |
} |
634 |
|
635 |
foreach my $database ($cfg->Sections) { |
636 |
|
637 |
my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; |
638 |
my $add_xml = $cfg -> val($database, 'xml'); # optional |
639 |
|
640 |
# create new lookup file |
641 |
my $lookup_file = $cfg -> val($database, 'lookup_newfile'); # optional |
642 |
if ($lookup_file) { |
643 |
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_NEWDB, 0644; |
644 |
if (! -e $lookup_file) { |
645 |
open(LOOKUP, "> $lookup_file") || die "can't create $lookup_file': $!"; |
646 |
close(LOOKUP); |
647 |
} |
648 |
tie %lhash, 'TDB_File', $lookup_file, TDB_CLEAR_IF_FIRST, O_RDWR, 0644; |
649 |
print STDERR "creating lookup file '$lookup_file'\n"; |
650 |
# delete memory cache for lookup file |
651 |
delete $cache->{lhash}; |
652 |
} |
653 |
|
654 |
# open existing lookup file |
655 |
$lookup_file = $cfg -> val($database, 'lookup_open'); # optional |
656 |
if ($lookup_file) { |
657 |
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_READER, 0644; |
658 |
tie %lhash, 'TDB_File', $lookup_file, TDB_DEFAULT, O_RDWR, 0644; |
659 |
print STDERR "opening lookup file '$lookup_file'\n"; |
660 |
} |
661 |
|
662 |
print STDERR "reading ./import_xml/$type.xml\n"; |
663 |
|
664 |
# extract just type basic |
665 |
my $type_base = $type; |
666 |
$type_base =~ s/_.+$//g; |
667 |
|
668 |
$config=XMLin("./import_xml/$type.xml", ForceArray => [ $type2tag{$type_base}, 'config', 'format' ], ForceContent => 1 ); |
669 |
|
670 |
# helper for progress bar |
671 |
sub fmt_time { |
672 |
my $t = shift || 0; |
673 |
my $out = ""; |
674 |
|
675 |
my ($ss,$mm,$hh) = gmtime($t); |
676 |
$out .= "${hh}h" if ($hh); |
677 |
$out .= sprintf("%02d:%02d", $mm,$ss); |
678 |
$out .= " " if ($hh == 0); |
679 |
return $out; |
680 |
} |
681 |
|
682 |
# output current progress indicator |
683 |
my $last_p = 0; |
684 |
my $start_t = time(); |
685 |
sub progress { |
686 |
return if (! $show_progress); |
687 |
my $current = shift || 1; |
688 |
my $total = shift || 1; |
689 |
my $p = int($current * 100 / $total); |
690 |
if ($p != $last_p) { |
691 |
my $rate = ($current / (time() - $start_t || 1)); |
692 |
my $eta = ($total-$current) / ($rate || 1); |
693 |
printf STDERR ("%5d [%-38s] %-5d %0.1f/s %s\r",$current,"=" x ($p/3)."$p%>", $total, $rate, fmt_time($eta)); |
694 |
$last_p = $p; |
695 |
} |
696 |
} |
697 |
|
698 |
my $fake_dir = 1; |
699 |
sub fakeprogress { |
700 |
return if (! $show_progress); |
701 |
my $current = shift @_; |
702 |
|
703 |
my @ind = ('-','\\','|','/','-','\\','|','/', '-'); |
704 |
|
705 |
$last_p += $fake_dir; |
706 |
$fake_dir = -$fake_dir if ($last_p > 1000 || $last_p < 0); |
707 |
if ($last_p % 10 == 0) { |
708 |
printf STDERR ("%5d / %5s [%-51s]\r",$current,"?"," " x ($last_p/20).$ind[($last_p/20) % $#ind]); |
709 |
} |
710 |
} |
711 |
|
712 |
# now read database |
713 |
print STDERR "using: $type...\n"; |
714 |
|
715 |
# erase cache for tags by order in this database |
716 |
delete $cache->{tags_by_order}; |
717 |
|
718 |
if ($type_base eq "isis") { |
719 |
|
720 |
my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!"; |
721 |
|
722 |
$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage); |
723 |
my $db = new IsisDB( isisdb => $isis_db ); |
724 |
|
725 |
my $max_rowid = $db->{'maxmfn'} || die "can't find maxmfn"; |
726 |
|
727 |
print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; |
728 |
|
729 |
my $path = $database; |
730 |
|
731 |
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { |
732 |
my $row = $db->to_hash( $row_id ); |
733 |
if ($row) { |
734 |
|
735 |
$row->{mfn} = $row_id; |
736 |
$row->{record} = $db->{record}; |
737 |
|
738 |
progress($row->{mfn}, $max_rowid); |
739 |
|
740 |
my $swishpath = $path."#".int($row->{mfn}); |
741 |
|
742 |
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
743 |
$xml = $cp2utf->convert($xml); |
744 |
use bytes; # as opposed to chars |
745 |
print "Path-Name: $swishpath\n"; |
746 |
print "Content-Length: ".(length($xml)+1)."\n"; |
747 |
print "Document-Type: XML\n\n$xml\n"; |
748 |
} |
749 |
} |
750 |
} |
751 |
print STDERR "\n"; |
752 |
|
753 |
} elsif ($type_base eq "excel") { |
754 |
require Spreadsheet::ParseExcel; |
755 |
require Spreadsheet::ParseExcel::Utility; |
756 |
import Spreadsheet::ParseExcel::Utility qw(int2col); |
757 |
|
758 |
$import2cp = Text::Iconv->new($config->{excel_codepage},$codepage); |
759 |
my $excel_file = $cfg -> val($database, 'excel_file') || die "$database doesn't have 'excel_file' defined!"; |
760 |
|
761 |
my $sheet = x($config->{sheet}) || die "no sheet in $type.xml"; |
762 |
my $start_row = x($config->{start_row}) - 1 || die "no start_row in $type.xml"; |
763 |
|
764 |
my $oBook = Spreadsheet::ParseExcel::Workbook->Parse($excel_file) || die "can't open Excel file '$excel_file'"; |
765 |
|
766 |
my $sheet_nr = 0; |
767 |
foreach my $oWks (@{$oBook->{Worksheet}}) { |
768 |
#print STDERR "-- SHEET $sheet_nr:", $oWks->{Name}, "\n"; |
769 |
last if ($oWks->{Name} eq $sheet); |
770 |
$sheet_nr++; |
771 |
} |
772 |
|
773 |
my $oWorksheet = $oBook->{Worksheet}[$sheet_nr]; |
774 |
|
775 |
print STDERR "using sheet: ",$oWorksheet->{Name},"\n"; |
776 |
defined ($oWorksheet) || die "can't find sheet '$sheet' in $excel_file"; |
777 |
my $end_row = x($config->{end_row}) || $oWorksheet->{MaxRow}; |
778 |
|
779 |
for(my $iR = $start_row ; defined $end_row && $iR <= $end_row ; $iR++) { |
780 |
my $row; |
781 |
for(my $iC = $oWorksheet->{MinCol} ; defined $oWorksheet->{MaxCol} && $iC <= $oWorksheet->{MaxCol} ; $iC++) { |
782 |
my $cell = $oWorksheet->{Cells}[$iR][$iC]; |
783 |
if ($cell) { |
784 |
# this conversion is a cludge. |
785 |
# Files from Excell could have |
786 |
# characters which don't fit into |
787 |
# destination encoding. |
788 |
$row->{int2col($iC)} = $utf2cp->convert($cell->Value) || $cell->Value; |
789 |
} |
790 |
} |
791 |
|
792 |
progress($iR, $end_row); |
793 |
|
794 |
# print "row[$iR/$end_row] "; |
795 |
# foreach (keys %{$row}) { |
796 |
# print "$_: ",$row->{$_},"\t"; |
797 |
# } |
798 |
# print "\n"; |
799 |
|
800 |
my $swishpath = $database."#".$iR; |
801 |
|
802 |
next if (! $row); |
803 |
|
804 |
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
805 |
$xml = $cp2utf->convert($xml); |
806 |
use bytes; # as opposed to chars |
807 |
print "Path-Name: $swishpath\n"; |
808 |
print "Content-Length: ".(length($xml)+1)."\n"; |
809 |
print "Document-Type: XML\n\n$xml\n"; |
810 |
} |
811 |
} |
812 |
} elsif ($type_base eq "marc") { |
813 |
|
814 |
require MARC; |
815 |
|
816 |
$import2cp = Text::Iconv->new($config->{marc_codepage},$codepage); |
817 |
my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!"; |
818 |
|
819 |
# optional argument is format |
820 |
my $format = x($config->{marc_format}) || 'usmarc'; |
821 |
|
822 |
print STDERR "Reading MARC file '$marc_file'\n"; |
823 |
|
824 |
my $marc = new MARC; |
825 |
my $nr = $marc->openmarc({ |
826 |
file=>$marc_file, format=>$format |
827 |
}) || die "Can't open MARC file '$marc_file' with format '$format'"; |
828 |
|
829 |
# read MARC file in memory |
830 |
$marc->nextmarc(-1); |
831 |
|
832 |
my $max_rec = $marc->marc_count(); |
833 |
|
834 |
for(my $i=1; $i<=$max_rec; $i++) { |
835 |
|
836 |
progress($i,$max_rec); |
837 |
|
838 |
# store value for marc_sf.pm |
839 |
$main::cache->{marc_record} = $i; |
840 |
|
841 |
my $swishpath = $database."#".$i; |
842 |
|
843 |
if (my $xml = data2xml($type_base,$marc,$add_xml,$cfg,$database)) { |
844 |
$xml = $cp2utf->convert($xml); |
845 |
use bytes; # as opposed to chars |
846 |
print "Path-Name: $swishpath\n"; |
847 |
print "Content-Length: ".(length($xml)+1)."\n"; |
848 |
print "Document-Type: XML\n\n$xml\n"; |
849 |
} |
850 |
} |
851 |
|
852 |
print STDERR "\n"; |
853 |
|
854 |
} elsif ($type_base eq "feed") { |
855 |
|
856 |
$import2cp = Text::Iconv->new($config->{feed_codepage},$codepage); |
857 |
my $prog = x($config->{prog}) || die "$database doesn't have 'prog' defined!"; |
858 |
|
859 |
print STDERR "Reading feed from program '$prog'\n"; |
860 |
|
861 |
open(FEED,"feeds/$prog |") || die "can't start $prog: $!"; |
862 |
|
863 |
my $i=1; # record nr. |
864 |
|
865 |
my $data; |
866 |
my $line=1; |
867 |
|
868 |
while (<FEED>) { |
869 |
chomp; |
870 |
|
871 |
if (/^$/) { |
872 |
my $swishpath = $database."#".$i++; |
873 |
|
874 |
if (my $xml = data2xml($type_base,$data,$add_xml,$cfg,$database)) { |
875 |
$xml = $cp2utf->convert($xml); |
876 |
use bytes; # as opposed to chars |
877 |
print "Path-Name: $swishpath\n"; |
878 |
print "Content-Length: ".(length($xml)+1)."\n"; |
879 |
print "Document-Type: XML\n\n$xml\n"; |
880 |
} |
881 |
$line = 1; |
882 |
$data = {}; |
883 |
next; |
884 |
} |
885 |
|
886 |
$line = $1 if (s/^(\d+):\s*//); |
887 |
$data->{$line++} = $_; |
888 |
|
889 |
fakeprogress($i); |
890 |
|
891 |
} |
892 |
# close lookup |
893 |
untie %lhash if (%lhash); |
894 |
} |
895 |
} |
896 |
|
897 |
# call this to commit index |
898 |
$index->close; |
899 |
|
900 |
1; |
901 |
__END__ |
902 |
########################################################################## |
903 |
|
904 |
=head1 NAME |
905 |
|
906 |
all2xml.pl - read various file formats and dump XML for SWISH-E |
907 |
|
908 |
=head1 SYNOPSYS |
909 |
|
910 |
$ all2xml.pl [test.conf] |
911 |
|
912 |
=head1 DESCRIPTION |
913 |
|
914 |
This command will read ISIS data file using OpenIsis perl module, MARC |
915 |
records using MARC module and optionally Micro$oft Excel files to |
916 |
create one XML file for usage with I<SWISH-E> indexer. Dispite it's name, |
917 |
this script B<isn't general xml generator> from isis files (isis allready |
918 |
has something like that). Output of this script is tailor-made for SWISH-E. |
919 |
|
920 |
If no configuration file is specified, it will use default one called |
921 |
C<all2xml.conf>. |
922 |
|
923 |
=head1 BUGS |
924 |
|
925 |
Documentation is really lacking. However, in true Open Source spirit, source |
926 |
is best documentation. I even made considerable effort to comment parts |
927 |
which are not intuitively clear, so... |
928 |
|
929 |
=head1 AUTHOR |
930 |
|
931 |
Dobrica Pavlinusic <dpavlin@rot13.org> |
932 |
|
933 |
=head1 COPYRIGHT |
934 |
|
935 |
GNU Public License (GPL) v2 or later |
936 |
|
937 |
=head1 SEE ALSO |
938 |
|
939 |
SWISH-E web site at http://www.swish-e.org |
940 |
|
941 |
=cut |