--- loncom/interface/loncommon.pm	2001/12/21 17:06:56	1.19
+++ loncom/interface/loncommon.pm	2001/12/25 21:02:35	1.20
@@ -1,7 +1,7 @@
 # The LearningOnline Network with CAPA
 # a pile of common routines
 #
-# $Id: loncommon.pm,v 1.19 2001/12/21 17:06:56 harris41 Exp $
+# $Id: loncommon.pm,v 1.20 2001/12/25 21:02:35 www Exp $
 #
 # Copyright Michigan State University Board of Trustees
 #
@@ -29,6 +29,8 @@
 # 2/13-12/7 Guy Albertelli
 # 12/11,12/12,12/17 Scott Harrison
 # 12/21 Gerd Kortemeyer
+# 12/21 Scott Harrison
+# 12/25 Gerd Kortemeyer
 
 # Makes a table out of the previous attempts
 # Inputs result_from_symbread, user, domain, course_id
@@ -41,12 +43,22 @@ use POSIX qw(strftime);
 use Apache::Constants qw(:common);
 use Apache::lonmsg();
 
-
+# ----------------------------------------------- Filetypes/Languages/Copyright
 my %language;
 my %cprtag;
 my %fe; my %fd;
 my %fc;
 
+# -------------------------------------------------------------- Thesaurus data
+my @therelated=();
+my @theword=();
+my @thecount=();
+my %theindex=();
+my $thetotalcount=0;
+my $thefuzzy=2;
+my $thethreshold=0.1/$thefuzzy;
+my $theavecount;
+
 # ----------------------------------------------------------------------- BEGIN
 BEGIN {
 # ------------------------------------------------------------------- languages
@@ -104,6 +116,82 @@ BEGIN {
 	    }
 	}
     }
+# -------------------------------------------------------------- Thesaurus data
+    {
+	my $fh=Apache::File->new($Apache::lonnet::perlvar{'lonTabDir'}.
+	       '/thesaurus.dat');
+	if ($fh) {
+            while (<$fh>) {
+               my ($tword,$tindex,$tcount,$trelated)=split(/\@/,$_);
+               $theindex{$tword}=$tindex;
+               $theword[$tindex]=$tword;
+               $thecount[$tindex]=$tcount;
+               $thetotalcount+=$tcount;
+               $therelated[$tindex]=$trelated;
+	   }
+        }
+        $theavecount=$thetotalcount/$#thecount;
+    }
+}
+# ============================================================= END BEGIN BLOCK
+
+
+# ---------------------------------------------------------- Is this a keyword?
+
+sub keyword {
+    my $newword=shift;
+    $newword=~s/\W//g;
+    $newword=~tr/A-Z/a-z/;
+    my $tindex=$theindex{$newword};
+    if ($tindex) {
+        if ($thecount[$tindex]>$theavecount) {
+           return 1;
+        }
+    }
+    return 0;
+}
+# -------------------------------------------------------- Return related words
+
+sub related {
+    my $newword=shift;
+    $newword=~s/\W//g;
+    $newword=~tr/A-Z/a-z/;
+    my $tindex=$theindex{$newword};
+    if ($tindex) {
+        my %found=();
+        foreach (split(/\,/,$therelated[$tindex])) {
+# - Related word found
+            my ($ridx,$rcount)=split(/\:/,$_);
+# - Direct relation index
+            my $directrel=$rcount/$thecount[$tindex];
+            if ($directrel>$thethreshold) {
+               foreach (split(/\,/,$therelated[$ridx])) {
+                  my ($rridx,$rrcount)=split(/\:/,$_);
+                  if ($rridx==$tindex) {
+# - Determine reverse relation index
+                     my $revrel=$rrcount/$thecount[$ridx];
+# - Calculate full index
+                     $found{$ridx}=$directrel*$revrel;
+                     if ($found{$ridx}>$thethreshold) {
+                        foreach (split(/\,/,$therelated[$ridx])) {
+                            my ($rrridx,$rrrcount)=split(/\:/,$_);
+                            unless ($found{$rrridx}) {
+                               my $revrevrel=$rrrcount/$thecount[$ridx];
+                               if (
+                          $directrel*$revrel*$revrevrel>$thethreshold
+                               ) {
+                                  $found{$rrridx}=
+                                       $directrel*$revrel*$revrevrel;
+                               }
+                            }
+                        }
+                     }
+                  }
+               }
+            }
+        }
+    }
+    return ();
 }
 
 # ---------------------------------------------------------------- Language IDs