Changeset 1287 for trunk/gsdl/bin/script/import.pl

Timestamp:

2000-07-14T12:24:20+12:00 (24 years ago)

Author:

sjboddie

Message:

Implemented a -sortmeta option for import.pl to sort archives.inf file
(generated at end of import process) alphabetically by the given
metadata element. This may be useful for some collections as boolean
queries currently return matches in build (fairly random) order. Changing
the order of archives.inf changes the order that documents are built.
This option has a couple of important limitations:

1. Can't be used in conjunction with the groupsize option as it would then only change the build order of groups of documents which doesn't seem very useful.
2. Is of limited use when building indexes at a section level as the build order is only sorted by document, not by section.

File:

• trunk/gsdl/bin/script/import.pl (modified) (6 diffs)

trunk/gsdl/bin/script/import.pl

@@ -60 +60 @@
     print STDERR "   -maxdocs number        Maximum number of documents to import\n";
     print STDERR "   -groupsize number      Number of GML documents to group into one file\n";
+    print STDERR "   -sortmeta metadata     Sort documents alphabetically by metadata for\n";
+    print STDERR "                          building. This will be disabled if groupsize > 1\n";
     print STDERR "   -debug                 Print imported text to STDOUT\n\n";
 }
@@ -69 +71 @@
     my ($verbosity, $importdir, $archivedir, $keepold, 
     $removeold, $gzip, $groupsize, $debug, $maxdocs, $collection,
-    $configfilename, $collectcfg, $pluginfo, 
+    $configfilename, $collectcfg, $pluginfo, $sortmeta,
     $archive_info_filename, $archive_info, $processor);
     if (!parsargv::parse(\@ARGV, 
@@ -79 +81 @@
              'gzip', \$gzip,
              'groupsize/\d+/1', \$groupsize,
+             'sortmeta/.*/', \$sortmeta,
              'debug', \$debug,
              'maxdocs/^\-?\d+/-1', \$maxdocs)) {
@@ -92 +95 @@
     &print_usage();
     die "\n";
+    }
+
+    # check sortmeta
+    $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
+    if (defined $sortmeta && $groupsize > 1) {
+    print STDERR "WARNING: import.pl cannot sort documents when groupsize > 1\n";
+    print STDERR "         sortmeta option will be ignored\n\n";
+    $sortmeta = undef;
     }
 
@@ -160 +171 @@
     $processor = new docsave ($collection, $archive_info, $verbosity, $gzip, $groupsize);
     $processor->setarchivedir ($archivedir);
+    $processor->set_sortmeta ($sortmeta) if defined $sortmeta;
     } else {
     $processor = new docprint ();
@@ -173 +185 @@
     # write out the archive information file
     if (!$debug) {
-    $processor->close_file_output();
+    $processor->close_file_output() if $groupsize > 1;
     $archive_info->save_info($archive_info_filename);
     }