source: trunk/gsdl/bin/script/import.pl@ 2328

Last change on this file since 2328 was 2328, checked in by sjboddie, 23 years ago

Added a -OIDtype option to import.pl to allow an alternative (and much
faster) OID creation technique to be used (i.e. a simple incremented
count). Hashing the contents of the document is still the default
behaviour. -- Note that this option is not yet implemented in buildcol.pl

  • Property svn:executable set to *
  • Property svn:keywords set to Author Date Id Revision
File size: 8.7 KB
Line 
1#!/usr/bin/perl -w
2
3###########################################################################
4#
5# import.pl --
6# A component of the Greenstone digital library software
7# from the New Zealand Digital Library Project at the
8# University of Waikato, New Zealand.
9#
10# Copyright (C) 1999 New Zealand Digital Library Project
11#
12# This program is free software; you can redistribute it and/or modify
13# it under the terms of the GNU General Public License as published by
14# the Free Software Foundation; either version 2 of the License, or
15# (at your option) any later version.
16#
17# This program is distributed in the hope that it will be useful,
18# but WITHOUT ANY WARRANTY; without even the implied warranty of
19# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20# GNU General Public License for more details.
21#
22# You should have received a copy of the GNU General Public License
23# along with this program; if not, write to the Free Software
24# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
25#
26###########################################################################
27
28
29# This program will import a number of files into a particular collection
30
31package import;
32
33BEGIN {
34 die "GSDLHOME not set\n" unless defined $ENV{'GSDLHOME'};
35 die "GSDLOS not set\n" unless defined $ENV{'GSDLOS'};
36 unshift (@INC, "$ENV{'GSDLHOME'}/perllib");
37 unshift (@INC, "$ENV{'GSDLHOME'}/perllib/plugins");
38 unshift (@INC, "$ENV{'GSDLHOME'}/perllib/classify");
39}
40
41use arcinfo;
42use colcfg;
43use plugin;
44use docprint;
45use util;
46use parsargv;
47use FileHandle;
48
49sub print_usage {
50 print STDERR "\n";
51 print STDERR "import.pl: Converts documents in collections -importdir directory into\n";
52 print STDERR " gml documents which are written to the -archivedir directory.\n\n";
53 print STDERR " usage: $0 [options] collection-name\n\n";
54 print STDERR " options:\n";
55 print STDERR " -verbosity number 0=none, 3=lots\n";
56 print STDERR " -importdir directory Where the original material lives\n";
57 print STDERR " -archivedir directory Where the converted material ends up\n";
58 print STDERR " -keepold Will not destroy the current contents of the\n";
59 print STDERR " archives directory (the default)\n";
60 print STDERR " -removeold Will remove the old contents of the archives\n";
61 print STDERR " directory -- use with care\n";
62 print STDERR " -gzip Use gzip to compress resulting gml documents\n";
63 print STDERR " (don't forget to include ZIPPlug in your plugin\n";
64 print STDERR " list when building from compressed documents)\n";
65 print STDERR " -maxdocs number Maximum number of documents to import\n";
66 print STDERR " -groupsize number Number of GML documents to group into one file\n";
67 print STDERR " -OIDtype hash|incremental The method to use when generating unique\n";
68 print STDERR " identifiers for each document. \"hash\" (the\n";
69 print STDERR " default) hashes the contents of the file and so\n";
70 print STDERR " will be the same every time the collection is\n";
71 print STDERR " imported. \"incremental\" is a simple document\n";
72 print STDERR " count and so will be significantly faster than\n";
73 print STDERR " \"hash\". It is not guaranteed to always assign\n";
74 print STDERR " the same identifier to a given document though\n";
75 print STDERR " and does not allow further documents to be added\n";
76 print STDERR " to existing gml archives\n";
77 print STDERR " -sortmeta metadata Sort documents alphabetically by metadata for\n";
78 print STDERR " building. This will be disabled if groupsize > 1\n";
79 print STDERR " -debug Print imported text to STDOUT\n";
80 print STDERR " -collectdir directory Collection directory (defaults to " .
81 &util::filename_cat ($ENV{'GSDLHOME'}, "collect") . ")\n";
82 print STDERR " -out Filename or handle to print output status to.\n";
83 print STDERR " The default is STDERR\n\n";
84}
85
86&main();
87
88sub main {
89 my ($verbosity, $importdir, $archivedir, $keepold,
90 $removeold, $gzip, $groupsize, $OIDtype, $debug,
91 $maxdocs, $collection, $configfilename, $collectcfg,
92 $pluginfo, $sortmeta, $archive_info_filename,
93 $archive_info, $processor, $out, $collectdir);
94 if (!parsargv::parse(\@ARGV,
95 'verbosity/\d+/2', \$verbosity,
96 'importdir/.*/', \$importdir,
97 'archivedir/.*/', \$archivedir,
98 'keepold', \$keepold,
99 'removeold', \$removeold,
100 'gzip', \$gzip,
101 'groupsize/\d+/1', \$groupsize,
102 'OIDtype/^(hash|incremental)$/hash', \$OIDtype,
103 'sortmeta/.*/', \$sortmeta,
104 'debug', \$debug,
105 'maxdocs/^\-?\d+/-1', \$maxdocs,
106 'collectdir/.*/', \$collectdir,
107 'out/.*/STDERR', \$out)) {
108 &print_usage();
109 die "\n";
110 }
111
112 my $close_out = 0;
113 if ($out !~ /^(STDERR|STDOUT)$/i) {
114 open (OUT, ">$out") || die "Couldn't open output file $out\n";
115 $out = 'import::OUT';
116 $close_out = 1;
117 }
118 $out->autoflush(1);
119
120 # set removeold to false if it has been defined
121 $removeold = 0 if ($keepold);
122
123 # get and check the collection name
124 if (($collection = &util::use_collection(@ARGV, $collectdir)) eq "") {
125 &print_usage();
126 die "\n";
127 }
128
129 # check sortmeta
130 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
131 if (defined $sortmeta && $groupsize > 1) {
132 print $out "WARNING: import.pl cannot sort documents when groupsize > 1\n";
133 print $out " sortmeta option will be ignored\n\n";
134 $sortmeta = undef;
135 }
136
137 # dynamically load 'docsave' module so it can pick up on a collection
138 # specific docsave.pm is specified.
139
140 unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
141 require docsave;
142
143
144 # get the list of plugins for this collection
145 my $plugins = [];
146 $configfilename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "etc", "collect.cfg");
147 if (-e $configfilename) {
148 $collectcfg = &colcfg::read_collect_cfg ($configfilename);
149 if (defined $collectcfg->{'plugin'}) {
150 $plugins = $collectcfg->{'plugin'};
151 }
152 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
153 $importdir = $collectcfg->{'importdir'};
154 }
155 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
156 $archivedir = $collectcfg->{'archivedir'};
157 }
158 if (defined $collectcfg->{'removeold'}) {
159 if ($collectcfg->{'removeold'} =~ /^true$/i && !$keepold) {
160 $removeold = 1;
161 }
162 if ($collectcfg->{'removeold'} =~ /^false$/i && !$removeold) {
163 $removeold = 0;
164 }
165 }
166 } else {
167 die "Couldn't find the configuration file $configfilename\n";
168 }
169
170 # fill in the default import and archives directories if none
171 # were supplied, turn all \ into / and remove trailing /
172 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
173 $importdir =~ s/[\\\/]+/\//g;
174 $importdir =~ s/\/$//;
175 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives") if $archivedir eq "";
176 $archivedir =~ s/[\\\/]+/\//g;
177 $archivedir =~ s/\/$//;
178
179 # load all the plugins
180 $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out);
181 if (scalar(@$pluginfo) == 0) {
182 print $out "No plugins were loaded.\n";
183 die "\n";
184 }
185
186 # remove the old contents of the archives directory if needed
187 if ($removeold && -e $archivedir) {
188 print $out "Warning - removing current contents of the archives directory\n";
189 print $out " in preparation for the import\n";
190 sleep(5); # just in case...
191 &util::rm_r ($archivedir);
192 }
193
194 # read the archive information file
195 if (!$debug) {
196 $archive_info_filename = &util::filename_cat ($archivedir, "archives.inf");
197 $archive_info = new arcinfo ();
198 $archive_info->load_info ($archive_info_filename);
199
200 # create a docsave object to process the documents
201 $processor = new docsave ($collection, $archive_info, $verbosity, $gzip, $groupsize, $out);
202 $processor->setarchivedir ($archivedir);
203 $processor->set_sortmeta ($sortmeta) if defined $sortmeta;
204 $processor->set_OIDtype ($OIDtype);
205 } else {
206 $processor = new docprint ();
207 }
208
209 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs);
210
211 # process the import directory
212 &plugin::read ($pluginfo, $importdir, "", {}, $processor, $maxdocs);
213
214 &plugin::end($pluginfo, $processor);
215
216 # write out the archive information file
217 if (!$debug) {
218 $processor->close_file_output() if $groupsize > 1;
219 $archive_info->save_info($archive_info_filename);
220 }
221 close OUT if $close_out;
222}
Note: See TracBrowser for help on using the repository browser.