1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # inexport.pm -- useful utilities to support import.pl and export.pl
|
---|
4 | # A component of the Greenstone digital library software
|
---|
5 | # from the New Zealand Digital Library Project at the
|
---|
6 | # University of Waikato, New Zealand.
|
---|
7 | #
|
---|
8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
---|
9 | #
|
---|
10 | # This program is free software; you can redistribute it and/or modify
|
---|
11 | # it under the terms of the GNU General Public License as published by
|
---|
12 | # the Free Software Foundation; either version 2 of the License, or
|
---|
13 | # (at your option) any later version.
|
---|
14 | #
|
---|
15 | # This program is distributed in the hope that it will be useful,
|
---|
16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
18 | # GNU General Public License for more details.
|
---|
19 | #
|
---|
20 | # You should have received a copy of the GNU General Public License
|
---|
21 | # along with this program; if not, write to the Free Software
|
---|
22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | #
|
---|
24 | ###########################################################################
|
---|
25 |
|
---|
26 | package inexport;
|
---|
27 |
|
---|
28 | use strict;
|
---|
29 |
|
---|
30 | use File::Basename;
|
---|
31 |
|
---|
32 | use util;
|
---|
33 | use GDBMUtils;
|
---|
34 |
|
---|
35 |
|
---|
36 | sub prime_doc_oid_count
|
---|
37 | {
|
---|
38 | my ($archivedir) = @_;
|
---|
39 | my $oid_count_filename = &util::filename_cat ($archivedir, "OIDcount");
|
---|
40 |
|
---|
41 | if (-e $oid_count_filename) {
|
---|
42 | if (open(OIDIN,"<$oid_count_filename")) {
|
---|
43 | my $OIDcount = <OIDIN>;
|
---|
44 | chomp $OIDcount;
|
---|
45 | close(OIDIN);
|
---|
46 |
|
---|
47 | $doc::OIDcount = $OIDcount;
|
---|
48 | }
|
---|
49 | else {
|
---|
50 |
|
---|
51 | print STDERR "Warning: unable to read document OID count from $oid_count_filename\n";
|
---|
52 | print STDERR "Setting value to 0\n";
|
---|
53 | }
|
---|
54 | }
|
---|
55 |
|
---|
56 | }
|
---|
57 |
|
---|
58 | sub store_doc_oid_count
|
---|
59 | {
|
---|
60 | # Use the file "OIDcount" in the archives directory to record
|
---|
61 | # what value doc.pm got up to
|
---|
62 |
|
---|
63 | my ($archivedir) = @_;
|
---|
64 | my $oid_count_filename = &util::filename_cat ($archivedir, "OIDcount");
|
---|
65 |
|
---|
66 |
|
---|
67 | if (open(OIDOUT,">$oid_count_filename")) {
|
---|
68 | print OIDOUT $doc::OIDcount, "\n";
|
---|
69 |
|
---|
70 | close(OIDOUT);
|
---|
71 | }
|
---|
72 | else {
|
---|
73 | print STDERR "Warning: unable to store document OID count\n";
|
---|
74 | }
|
---|
75 | }
|
---|
76 |
|
---|
77 |
|
---|
78 |
|
---|
79 | sub new_vs_old_import_diff
|
---|
80 | {
|
---|
81 | my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
|
---|
82 |
|
---|
83 | # First convert all files to absolute form
|
---|
84 | # This is to support the situation where the import folder is not
|
---|
85 | # the default
|
---|
86 |
|
---|
87 | my $prev_all_files = $archive_info->{'prev_import_filelist'};
|
---|
88 | my $full_prev_all_files = {};
|
---|
89 |
|
---|
90 | foreach my $prev_file (keys %$prev_all_files) {
|
---|
91 |
|
---|
92 | if (!&util::filename_is_absolute($prev_file)) {
|
---|
93 | my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
|
---|
94 | $full_prev_all_files->{$full_prev_file} = $prev_file;
|
---|
95 | }
|
---|
96 | else {
|
---|
97 | $full_prev_all_files->{$prev_file} = $prev_file;
|
---|
98 | }
|
---|
99 | }
|
---|
100 |
|
---|
101 |
|
---|
102 | # Figure out which are the new files, existing files and so
|
---|
103 | # by implication the files from the previous import that are not
|
---|
104 | # there any more => mark them for deletion
|
---|
105 | foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
|
---|
106 |
|
---|
107 | my $full_curr_file = $curr_file;
|
---|
108 |
|
---|
109 | # entry in 'all_files' is moved to either 'existing_files',
|
---|
110 | # 'deleted_files', 'new_files', or 'new_metadata_files'
|
---|
111 |
|
---|
112 | if (!&util::filename_is_absolute($curr_file)) {
|
---|
113 | # add in import dir to make absolute
|
---|
114 | $full_curr_file = &util::filename_cat($importdir,$curr_file);
|
---|
115 | }
|
---|
116 |
|
---|
117 | if (defined $block_hash->{'file_blocks'}->{$full_curr_file}) {
|
---|
118 | # If in block list, we want to ignore it
|
---|
119 | delete $block_hash->{'all_files'}->{$curr_file};
|
---|
120 |
|
---|
121 | if (defined $full_prev_all_files->{$full_curr_file}) {
|
---|
122 | # also make sure it is gone from 'previous' list so
|
---|
123 | # not mistaken for a file that needs to be deleted
|
---|
124 | delete $full_prev_all_files->{$full_curr_file};
|
---|
125 | }
|
---|
126 | next;
|
---|
127 | }
|
---|
128 |
|
---|
129 | # figure out if new file or not
|
---|
130 | if (defined $full_prev_all_files->{$full_curr_file}) {
|
---|
131 |
|
---|
132 | if ($incremental_mode eq "all") {
|
---|
133 |
|
---|
134 | # had it before
|
---|
135 | $block_hash->{'existing_files'}->{$full_curr_file} = 1;
|
---|
136 |
|
---|
137 | # Now remove it, so by end of loop only the files
|
---|
138 | # that need deleting are left
|
---|
139 |
|
---|
140 | delete $full_prev_all_files->{$full_curr_file};
|
---|
141 | }
|
---|
142 | else {
|
---|
143 | # Warning in "onlyadd" mode, but had it before!
|
---|
144 | print STDERR "Warning: File $full_curr_file previously imported.\n";
|
---|
145 | print STDERR " Treating as new file\n";
|
---|
146 |
|
---|
147 | $block_hash->{'new_files'}->{$full_curr_file} = 1;
|
---|
148 | delete $full_prev_all_files->{$full_curr_file};
|
---|
149 | }
|
---|
150 | }
|
---|
151 | else {
|
---|
152 | if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
|
---|
153 | # the new file is the special sort of file greenstone uses
|
---|
154 | # to attach metadata to src documents
|
---|
155 | # i.e metadata.xml
|
---|
156 | # (but note, the filename used is not constrained in
|
---|
157 | # Greenstone to always be this)
|
---|
158 |
|
---|
159 | ## print STDERR "***** Detected new metadata file: $full_curr_file\n";
|
---|
160 | $block_hash->{'new_metadata_files'}->{$full_curr_file} = 1;
|
---|
161 | }
|
---|
162 | else {
|
---|
163 | $block_hash->{'new_files'}->{$full_curr_file} = 1;
|
---|
164 | }
|
---|
165 | }
|
---|
166 |
|
---|
167 |
|
---|
168 | delete $block_hash->{'all_files'}->{$curr_file};
|
---|
169 | }
|
---|
170 |
|
---|
171 |
|
---|
172 | # Deal with complication of new metadata.xml files by forcing
|
---|
173 | # everything from this point down in the file hierarchy to
|
---|
174 | # be freshly imported.
|
---|
175 | #
|
---|
176 | # This may mean files that have not changed are reindexed, but does
|
---|
177 | # guarantee by the end of processing all new metadata is correctly
|
---|
178 | # associated with the relevant document(s).
|
---|
179 |
|
---|
180 | foreach my $new_mdf (keys %{$block_hash->{'new_metadata_files'}}) {
|
---|
181 | my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
|
---|
182 |
|
---|
183 | $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
|
---|
184 |
|
---|
185 | # Go through existing_files, and mark anything that is contained
|
---|
186 | # within 'situated_dir' to be reindexed (in case some of the metadata
|
---|
187 | # attaches to one of these files)
|
---|
188 |
|
---|
189 | my $reindex_files = [];
|
---|
190 |
|
---|
191 | foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
|
---|
192 | # need to protect windows slash \ in regular expression?
|
---|
193 |
|
---|
194 | if ($existing_f =~ m/^$situated_dir/) {
|
---|
195 | push(@$reindex_files,$existing_f);
|
---|
196 | $block_hash->{'reindex_files'}->{$existing_f} = 1;
|
---|
197 |
|
---|
198 | }
|
---|
199 | }
|
---|
200 |
|
---|
201 | # Reindexing is accomplished by putting them in th list for reindexing (line above)
|
---|
202 | # and then tagging the arcinfo version as to be deleted.
|
---|
203 |
|
---|
204 | _mark_docs_for_deletion($archive_info,$block_hash,$reindex_files,$archivedir,$verbosity, "reindex");
|
---|
205 |
|
---|
206 | # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
|
---|
207 | # (or equivalent)
|
---|
208 | $block_hash->{'new_files'}->{$new_mdf} = 1;
|
---|
209 |
|
---|
210 | }
|
---|
211 |
|
---|
212 |
|
---|
213 | # By this point full_prev_all_files contains the files
|
---|
214 | # mentioned in archiveinf-src.db but are not in the 'import'
|
---|
215 | # folder (or whatever was specified through -importdir ...)
|
---|
216 |
|
---|
217 | # This list can contain files that were created in the 'tmp' or
|
---|
218 | # 'cache' areas (such as screen-size and thumbnail images).
|
---|
219 | #
|
---|
220 | # In building the final list of files to delete, we test to see if
|
---|
221 | # it exists on the filesystem and if it does (unusual for a "normal"
|
---|
222 | # file in import, but possible in the case of 'tmp' files),
|
---|
223 | # supress it from going into the final list
|
---|
224 |
|
---|
225 | my $collectdir = $ENV{'GSDLCOLLECTDIR'};
|
---|
226 |
|
---|
227 | my @deleted_files = values %$full_prev_all_files;
|
---|
228 | map { my $curr_file = $_;
|
---|
229 | my $full_curr_file = $curr_file;
|
---|
230 |
|
---|
231 | if (!&util::filename_is_absolute($curr_file)) {
|
---|
232 | # add in import dir to make absolute
|
---|
233 |
|
---|
234 | $full_curr_file = &util::filename_cat($collectdir,$curr_file);
|
---|
235 | }
|
---|
236 |
|
---|
237 |
|
---|
238 | if (!-e $full_curr_file) {
|
---|
239 | $block_hash->{'deleted_files'}->{$curr_file} = 1;
|
---|
240 | }
|
---|
241 | } @deleted_files;
|
---|
242 |
|
---|
243 |
|
---|
244 |
|
---|
245 | }
|
---|
246 |
|
---|
247 |
|
---|
248 | sub is_assoc_file
|
---|
249 | {
|
---|
250 | my ($file,$doc_rec) = @_;
|
---|
251 |
|
---|
252 | my ($file_root,$dirname,$suffix) = fileparse($file, "\\.[^\\.]+\$");
|
---|
253 |
|
---|
254 | foreach my $af (@{$doc_rec->{'assoc-file'}}) {
|
---|
255 | my $full_af = &util::filename_cat($dirname,$af);
|
---|
256 |
|
---|
257 | return 1 if ($full_af eq $file);
|
---|
258 | }
|
---|
259 |
|
---|
260 | return 0;
|
---|
261 | }
|
---|
262 |
|
---|
263 |
|
---|
264 | sub _mark_docs_for_deletion
|
---|
265 | {
|
---|
266 | my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode_text) = @_;
|
---|
267 |
|
---|
268 | my $doc_db = "archiveinf-doc.gdb";
|
---|
269 | my $src_db = "archiveinf-src.gdb";
|
---|
270 | my $arcinfo_doc_filename = &util::filename_cat ($archivedir, $doc_db);
|
---|
271 | my $arcinfo_src_filename = &util::filename_cat ($archivedir, $src_db);
|
---|
272 |
|
---|
273 |
|
---|
274 | # record files marked for deletion in arcinfo
|
---|
275 | foreach my $file (@$deleted_files) {
|
---|
276 | # use 'archiveinf-src' GDBM file to look up all the OIDs
|
---|
277 | # that this file is used in (note in most cases, it's just one OID)
|
---|
278 |
|
---|
279 | my $src_rec = GDBMUtils::gdbmRecordToHash($arcinfo_src_filename,$file);
|
---|
280 | my $oids = $src_rec->{'oid'};
|
---|
281 | foreach my $oid (@$oids) {
|
---|
282 | # Find out if it's a main doc, assoc file, or metadata
|
---|
283 |
|
---|
284 | my $doc_rec = GDBMUtils::gdbmRecordToHash($arcinfo_doc_filename,$oid);
|
---|
285 |
|
---|
286 | if (is_assoc_file($file,$doc_rec)) {
|
---|
287 | # assoc file => mark it for re-indexing (safest thing to do)
|
---|
288 | my $curr_status = $archive_info->get_status_info($oid);
|
---|
289 |
|
---|
290 |
|
---|
291 | if (defined($curr_status) && (($curr_status ne "D") && ($curr_status ne "R"))) {
|
---|
292 | if ($verbosity > 1) {
|
---|
293 | print STDERR "$oid marked for (potential) reindexing\n";
|
---|
294 | print STDERR " (because associated file $file deleted)\n";
|
---|
295 | }
|
---|
296 | $archive_info->set_status_info($oid,"R");
|
---|
297 |
|
---|
298 | my $val = &GDBMUtils::gdbmDatabaseGet($arcinfo_doc_filename,$oid);
|
---|
299 | $val =~ s/^<index-status>(.*)$/<index-status>R/m;
|
---|
300 | &GDBMUtils::gdbmDatabaseSet($arcinfo_doc_filename,$oid,$val);
|
---|
301 | }
|
---|
302 | GDBMUtils::gdbmDatabaseRemove($arcinfo_src_filename,$file);
|
---|
303 | }
|
---|
304 | else {
|
---|
305 | # either src-file or metadata.xml file linking to src-file
|
---|
306 |
|
---|
307 | my $src_file;
|
---|
308 |
|
---|
309 | if ($doc_rec->{'src-file'}->[0] ne $file) {
|
---|
310 | # it's a metadata file attached to this OID
|
---|
311 | # => workout the src-file it matches to
|
---|
312 |
|
---|
313 | $src_file = $doc_rec->{'src-file'}->[0];
|
---|
314 |
|
---|
315 | my $src_filename = $src_file;
|
---|
316 | if (!&util::filename_is_absolute($src_file)) {
|
---|
317 | $src_filename = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$src_file);
|
---|
318 | }
|
---|
319 |
|
---|
320 | $block_hash->{'reindex_files'}->{$src_filename} = 1;
|
---|
321 |
|
---|
322 | # remove the metadata file from the src-database
|
---|
323 | GDBMUtils::gdbmDatabaseRemove($arcinfo_src_filename,$file);
|
---|
324 | }
|
---|
325 | else {
|
---|
326 | # It's the main doc
|
---|
327 | # => make it the target and mark it for deletion
|
---|
328 | $src_file = $file;
|
---|
329 | }
|
---|
330 |
|
---|
331 | # Whether the main file directly or indirectly, mark for deletion/reindex
|
---|
332 |
|
---|
333 | my $val = &GDBMUtils::gdbmDatabaseGet($arcinfo_doc_filename,$oid);
|
---|
334 | my ($index_status) = ($val =~ m/^<index-status>(.*)$/m);
|
---|
335 |
|
---|
336 | if ($index_status ne "D") {
|
---|
337 | if ($verbosity>1) {
|
---|
338 | print STDERR "$oid ($src_file) marked to be $mode_text on next buildcol.pl\n";
|
---|
339 | }
|
---|
340 | $archive_info->set_status_info($oid,"D");
|
---|
341 |
|
---|
342 | $val =~ s/^<index-status>(.*)$/<index-status>D/m;
|
---|
343 | &GDBMUtils::gdbmDatabaseSet($arcinfo_doc_filename,$oid,$val);
|
---|
344 |
|
---|
345 | GDBMUtils::gdbmDatabaseRemove($arcinfo_src_filename,$src_file);
|
---|
346 | }
|
---|
347 |
|
---|
348 |
|
---|
349 | }
|
---|
350 |
|
---|
351 | }
|
---|
352 | }
|
---|
353 | }
|
---|
354 |
|
---|
355 |
|
---|
356 | sub mark_docs_for_deletion
|
---|
357 | {
|
---|
358 | _mark_docs_for_deletion(@_,"deleted from index");
|
---|
359 | }
|
---|
360 |
|
---|
361 |
|
---|
362 | sub mark_docs_for_reindex
|
---|
363 | {
|
---|
364 | my ($archive_info,$block_hash,$archivedir,$verbosity) = @_;
|
---|
365 |
|
---|
366 | # Reindexing is accomplished by deleting the previously indexed
|
---|
367 | # version of the document, and then allowing the new version to
|
---|
368 | # be indexed (as would a new document be indexed).
|
---|
369 | #
|
---|
370 | # The first step (marking for deletion) is implemented by this routine.
|
---|
371 | #
|
---|
372 | # By default in Greenstone a new version of an index will hash to
|
---|
373 | # a new unique OID, and the above strategy of reindex=delete+add
|
---|
374 | # works fine. A special case arises when a persistent OID is
|
---|
375 | # allocated to a document (for instance through a metadata field),
|
---|
376 | # and the second step to reindexing (see XXXX) detects this and
|
---|
377 | # deals with it appropriately.
|
---|
378 |
|
---|
379 | my @existing_files = sort keys %{$block_hash->{'existing_files'}};
|
---|
380 |
|
---|
381 | my $doc_db = "archiveinf-doc.gdb";
|
---|
382 | my $arcinfo_doc_filename = &util::filename_cat ($archivedir, $doc_db);
|
---|
383 |
|
---|
384 | my $archiveinf_timestamp = -M $arcinfo_doc_filename;
|
---|
385 |
|
---|
386 | my $reindex_files = [];
|
---|
387 |
|
---|
388 | foreach my $existing_filename (@existing_files) {
|
---|
389 | if (-M $existing_filename < $archiveinf_timestamp) {
|
---|
390 | # file is newer than last build
|
---|
391 |
|
---|
392 | my $existing_file = $existing_filename;
|
---|
393 | #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
|
---|
394 |
|
---|
395 | #my $collectdir_resafe = &util::filename_to_regex($collectdir);
|
---|
396 | #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
|
---|
397 |
|
---|
398 | print STDERR "**** Reindexing existing file: $existing_file\n";
|
---|
399 |
|
---|
400 | push(@$reindex_files,$existing_file);
|
---|
401 | $block_hash->{'reindex_files'}->{$existing_filename} = 1;
|
---|
402 | }
|
---|
403 |
|
---|
404 | }
|
---|
405 |
|
---|
406 | _mark_docs_for_deletion($archive_info,$block_hash,$reindex_files,$archivedir,$verbosity, "reindex");
|
---|
407 |
|
---|
408 | }
|
---|
409 |
|
---|
410 |
|
---|
411 |
|
---|
412 | 1;
|
---|