Changeset 215 for trunk/gsdl/perllib/mgbuildproc.pm

Timestamp:

1999-03-30T17:08:18+12:00 (25 years ago)

Author:

rjmcnab

Message:

Added code to build a new gdbm format.

File:

• trunk/gsdl/perllib/mgbuildproc.pm (modified) (4 diffs)

trunk/gsdl/perllib/mgbuildproc.pm

@@ -5 +5 @@
 package mgbuildproc;
 
+use classify;
+use doc;
 use docproc;
 use util;
@@ -22 +24 @@
     $self->{'build_dir'} = $build_dir;
     $self->{'verbosity'} = $verbosity;
+    $self->{'classifiers'} = [];
     $self->{'mode'} = "text";
     $self->{'index'} = "section:text";
@@ -81 +84 @@
 }
 
+sub set_classifiers {
+    my $self = shift (@_);
+    my ($classifiers) = @_;
+
+    $self->{'classifiers'} = $classifiers;
+}
+
 sub process {
     my $self = shift (@_);
@@ -87 +97 @@
     $self->$method(@_);
 }
+
+sub newinfodb {
+    my $self = shift (@_);
+    my ($doc_obj, $filename) = @_;
+    my $handle = $self->{'output_handle'};
+#    $handle = "main::STDOUT";
+
+    # this was used in the old version
+    return if ($doc_obj eq 'classifications');
+
+    my $doctype = $doc_obj->get_doc_type();
+
+    # only output this document if it is one to be indexed
+    return if ($doctype ne "indexed_doc");
+
+    # this is another document
+    $self->{'num_docs'} += 1 unless ($doctype eq "classification");
+
+    my $section = $doc_obj->get_top_section ();
+    my $doc_OID = $doc_obj->get_OID();
+    while (defined $section) {
+    # update a few statistics
+    $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
+    $self->{'num_sections'} += 1 unless ($doctype eq "classification");
+
+    # output the section name
+    if ($section eq "") { print $handle "[$doc_OID]\n"; }
+    else { print $handle "[$doc_OID.$section]\n"; }
+
+    # output the fact that this document is a document
+    print $handle "<doctype>doc\n";
+
+    # output whether this node contains text
+    if ($doc_obj->get_text_length($section) > 0) {
+        print $handle "<hastxt>1\n";
+    } else {
+        print $handle "<hastxt>0\n";
+    }
+
+    # output all the section metadata
+    my $metadata = $doc_obj->get_all_metadata ($section);
+    foreach $pair (@$metadata) {
+        my ($field, $value) = (@$pair);
+
+        if ($field ne "Identifier" && $field !~ /^gsdl/ && 
+        defined $value && $value ne "") {
+        # escape problematic stuff
+        $value =~ s/\\/\\\\/g;
+        $value =~ s/\n/\\n/g;
+        $value =~ s/\r/\\r/g;
+        
+        print $handle "<$field>$value\n";
+        }
+    }
+
+    # output a list of children
+    my $children = $doc_obj->get_children ();
+    if (scalar(@$children) > 0) {
+        print $handle "<contains>";
+        my $firstchild = 1;
+        foreach $child (@$children) {
+        print $handle ";" unless $firstchild;
+        $firstchild = 0;
+        if ($child eq "") { print $handle "$doc_OID"; }
+        else { print $handle "$doc_OID.$child"; }
+        }
+        print $handle "\n";
+    }
+
+    # output the matching document number
+    print $handle "<docnum>$self->{'num_sections'}\n";
+
+    print $handle '-' x 70, "\n";
+
+    
+    # output a database entry for the document number
+    print $handle "[$self->{'num_sections'}]\n";
+    if ($section eq "") { print $handle "<section>$doc_OID\n"; }
+    else { print $handle "<section>$doc_OID.$section\n"; }
+    print $handle '-' x 70, "\n";
+
+
+    $section = $doc_obj->get_next_section($section);
+    }
+
+    # classify this document
+    &classify::classify_doc ($self->{'classifiers'}, $doc_obj);
+
+}
+
+
 
 sub infodb {