source: trunk/gsdl/bin/script/import.pl@ 1970

Last change on this file since 1970 was 1970, checked in by sjboddie, 23 years ago

Added more usage information to all perl programs and removed a few
programs that are no longer useful.
  • Property svn:executable set to *
  • Property svn:keywords set to Author Date Id Revision
File size: 7.7 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: $0 [options] collection-name\n\n";
54 print STDERR " options:\n";
55 print STDERR " -verbosity number 0=none, 3=lots\n";
56 print STDERR " -importdir directory Where the original material lives\n";
57 print STDERR " -archivedir directory Where the converted material ends up\n";
58 print STDERR " -keepold Will not destroy the current contents of the\n";
59 print STDERR " archives directory (the default)\n";
60 print STDERR " -removeold Will remove the old contents of the archives\n";
61 print STDERR " directory -- use with care\n";
62 print STDERR " -gzip Use gzip to compress resulting gml documents\n";
63 print STDERR " (don't forget to include ZIPPlug in your plugin\n";
64 print STDERR " list when building from compressed documents)\n";
65 print STDERR " -maxdocs number Maximum number of documents to import\n";
66 print STDERR " -groupsize number Number of GML documents to group into one file\n";
67 print STDERR " -sortmeta metadata Sort documents alphabetically by metadata for\n";
68 print STDERR " building. This will be disabled if groupsize > 1\n";
69 print STDERR " -debug Print imported text to STDOUT\n";
70 print STDERR " -collectdir directory Collection directory (defaults to " .
71 &util::filename_cat ($ENV{'GSDLHOME'}, "collect") . ")\n";
72 print STDERR " -out Filename or handle to print output status to.\n";
73 print STDERR " The default is STDERR\n\n";
74}
75
76&main();
77
78sub main {
79 my ($verbosity, $importdir, $archivedir, $keepold,
80 $removeold, $gzip, $groupsize, $debug, $maxdocs, $collection,
81 $configfilename, $collectcfg, $pluginfo, $sortmeta,
82 $archive_info_filename, $archive_info, $processor,
83 $out, $collectdir);
84 if (!parsargv::parse(\@ARGV,
85 'verbosity/\d+/2', \$verbosity,
86 'importdir/.*/', \$importdir,
87 'archivedir/.*/', \$archivedir,
88 'keepold', \$keepold,
89 'removeold', \$removeold,
90 'gzip', \$gzip,
91 'groupsize/\d+/1', \$groupsize,
92 'sortmeta/.*/', \$sortmeta,
93 'debug', \$debug,
94 'maxdocs/^\-?\d+/-1', \$maxdocs,
95 'collectdir/.*/', \$collectdir,
96 'out/.*/STDERR', \$out)) {
97 &print_usage();
98 die "\n";
99 }
100
101 my $close_out = 0;
102 if ($out !~ /^(STDERR|STDOUT)$/i) {
103 open (OUT, ">$out") || die "Couldn't open output file $out\n";
104 $out = 'import::OUT';
105 $close_out = 1;
106 }
107 $out->autoflush(1);
108
109 # set removeold to false if it has been defined
110 $removeold = 0 if ($keepold);
111
112 # get and check the collection name
113 if (($collection = &util::use_collection(@ARGV, $collectdir)) eq "") {
114 &print_usage();
115 die "\n";
116 }
117
118 # check sortmeta
119 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
120 if (defined $sortmeta && $groupsize > 1) {
121 print $out "WARNING: import.pl cannot sort documents when groupsize > 1\n";
122 print $out " sortmeta option will be ignored\n\n";
123 $sortmeta = undef;
124 }
125
126 # dynamically load 'docsave' module so it can pick up on a collection
127 # specific docsave.pm is specified.
128
129 unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
130 require docsave;
131
132
133 # get the list of plugins for this collection
134 my $plugins = [];
135 $configfilename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "etc", "collect.cfg");
136 if (-e $configfilename) {
137 $collectcfg = &colcfg::read_collect_cfg ($configfilename);
138 if (defined $collectcfg->{'plugin'}) {
139 $plugins = $collectcfg->{'plugin'};
140 }
141 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
142 $importdir = $collectcfg->{'importdir'};
143 }
144 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
145 $archivedir = $collectcfg->{'archivedir'};
146 }
147 if (defined $collectcfg->{'removeold'}) {
148 if ($collectcfg->{'removeold'} =~ /^true$/i && !$keepold) {
149 $removeold = 1;
150 }
151 if ($collectcfg->{'removeold'} =~ /^false$/i && !$removeold) {
152 $removeold = 0;
153 }
154 }
155 } else {
156 die "Couldn't find the configuration file $configfilename\n";
157 }
158
159 # fill in the default import and archives directories if none
160 # were supplied, turn all \ into / and remove trailing /
161 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
162 $importdir =~ s/[\\\/]+/\//g;
163 $importdir =~ s/\/$//;
164 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives") if $archivedir eq "";
165 $archivedir =~ s/[\\\/]+/\//g;
166 $archivedir =~ s/\/$//;
167
168 # load all the plugins
169 $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out);
170 if (scalar(@$pluginfo) == 0) {
171 print $out "No plugins were loaded.\n";
172 die "\n";
173 }
174
175 # remove the old contents of the archives directory if needed
176 if ($removeold && -e $archivedir) {
177 print $out "Warning - removing current contents of the archives directory\n";
178 print $out " in preparation for the import\n";
179 sleep(5); # just in case...
180 &util::rm_r ($archivedir);
181 }
182
183 # read the archive information file
184 if (!$debug) {
185 $archive_info_filename = &util::filename_cat ($archivedir, "archives.inf");
186 $archive_info = new arcinfo ();
187 $archive_info->load_info ($archive_info_filename);
188
189 # create a docsave object to process the documents
190 $processor = new docsave ($collection, $archive_info, $verbosity, $gzip, $groupsize, $out);
191 $processor->setarchivedir ($archivedir);
192 $processor->set_sortmeta ($sortmeta) if defined $sortmeta;
193 } else {
194 $processor = new docprint ();
195 }
196
197 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs);
198
199 # process the import directory
200 &plugin::read ($pluginfo, $importdir, "", {}, $processor, $maxdocs);
201
202 &plugin::end($pluginfo, $processor);
203
204 # write out the archive information file
205 if (!$debug) {
206 $processor->close_file_output() if $groupsize > 1;
207 $archive_info->save_info($archive_info_filename);
208 }
209 close OUT if $close_out;
210}
Note: See TracBrowser for help on using the repository browser.