# HG changeset patch
# User Igor Chubin <igor@chub.in>
# Date 1288984066 -3600
# Node ID 753fb84437aa0b09fee9831149d0148161982b50
# Parent  48ca8248e9cc157b89a9757ca8333c25a94bfe41
link between words support; link is created using @

diff -r 48ca8248e9cc -r 753fb84437aa new-words.sh
--- a/new-words.sh	Tue Aug 17 21:35:57 2010 +0200
+++ b/new-words.sh	Fri Nov 05 20:07:46 2010 +0100
@@ -257,6 +257,7 @@
 
 use Encode;
 use utf8;
+use Lingua::Stem::Snowball qw(stem);
 
 eval {
 # http://stackoverflow.com/questions/251694/how-can-i-check-if-i-have-a-perl-module-before-using-it
@@ -268,6 +269,22 @@
     our $HAVE_String_Similarity=1;
 }
 
+
+sub load_notes_dict()
+{
+    my %dict;
+    if (open(NOTES, $ENV{NOTES_FILE})) {
+        while(<NOTES>) {
+            $_ = decode( "utf8", $_);
+            chomp;
+            s/^\s+//;
+            my ($a,$b)=split /\s+/,$_,2;
+            $dict{$a}=$b;
+        }
+    }
+    return %dict;
+}
+
 sub similar($$){
     my $a=shift;
     my $b=shift;
@@ -279,7 +296,8 @@
     }
 }
 
-sub normalize($)
+
+sub normalize_without_linked($)
 {
     if   ( $ENV{LANGUAGE} eq "en" ) { return normalize_english(shift); }
     elsif ( $ENV{LANGUAGE} eq "de" ) { return normalize_german(shift); }
@@ -288,6 +306,23 @@
     else { return shift ; }
 }
 
+sub normalize_with_linked($)
+{
+    my $word = normalize_without_linked(shift);
+        #return $word;
+    if ($linked_words{$word}) {
+        return $linked_words{$word};
+    }
+    else {
+        return $word;
+    }
+}
+
+sub normalize($)
+{
+    return normalize_with_linked(shift);
+}
+
 sub normalize_ukrainian($)
 {
     $_=lc(shift);
@@ -309,6 +344,12 @@
 
 sub normalize_german($)
 {
+    @stems = stem('de', \@_);
+    return $stems[0];
+}
+
+sub normalize_german_($)
+{
     $_=lc(shift);
 
     s/heit$//;  s/keit$//; s/tum$//; s/ung$//; s/nis$//;s/schaft$//; s/ist$//; 
@@ -379,6 +420,35 @@
     }
 }
 
+sub log_($)
+{
+    return;
+    open(LOG, ">>", "/tmp/log1");
+    print LOG $_[0];
+    close(LOG);
+}
+
+sub find_linked_words($)
+{
+    my %linked_words;
+    my $dict = shift;
+    log_("1");
+    log_(join(" ", keys(%$dict)));
+
+    for $key (keys(%$dict)) {
+        $val = $dict->{$key};
+        log_($key."\n");
+        if ($val =~ /\@([a-z]*)/) {
+            $linked_words{normalize($key)} = normalize($1);
+            log_(normalize($key)." = ".normalize($1)."\n");
+        }
+    }
+    return %linked_words;
+}
+
+our %dict = load_notes_dict();
+our %linked_words = find_linked_words(\%dict);
+
 our %Vocabulary;
 open(VOC, $ENV{VOCABULARY})
  or die "Can't open VOCABULARY";
@@ -415,6 +485,8 @@
     export VOCABULARY
     export NEED_TO_USE_VOCABULARY_WHEN_SORT
     export LANGUAGE
+    [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
+    export NOTES_FILE
     perl $PERL_SCRIPT_TEMP_NAME
     rm $PERL_SCRIPT_TEMP_NAME
 }
@@ -430,17 +502,24 @@
     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
 use Encode;
 
+sub load_notes_dict()
+{
+    my %dict;
+    if (open(NOTES, $ENV{NOTES_FILE})) {
+        while(<NOTES>) {
+            $_ = decode( "utf8", $_);
+            chomp;
+            s/^\s+//;
+            my ($a,$b)=split /\s+/,$_,2;
+            $dict{$a}=$b;
+        }
+    }
+    return %dict;
+}
+
+%dict = load_notes_dict();
+
 $file = $ARGV[0];
-our $dict;
-if (open(NOTES, $ENV{NOTES_FILE})) {
-    while(<NOTES>) {
-        $_ = decode( "utf8", $_);
-        chomp;
-        s/^\s+//;
-        my ($a,$b)=split /\s+/,$_,2;
-        $dict{$a}=$b;
-    }
-}
 if (open(F, $file)) {
     @lines=<F>;
     close(F);