From 1639272931024b42de273a9d4456ee1232448ea6 Mon Sep 17 00:00:00 2001
From: root <root>
Date: Sat, 21 Jan 2006 21:20:34 +0000
Subject: [PATCH] *** empty log message ***

---
 Changes            |  2 +-
 src/perl/selection | 57 +++++++++++++++++++++++++++++++---------------
 2 files changed, 40 insertions(+), 19 deletions(-)

diff --git a/Changes b/Changes
index 38905828..cedf0f30 100644
--- a/Changes
+++ b/Changes
@@ -19,7 +19,7 @@ WISH: just for fun, do shade and tint with XRender.
 WISH: support tex fonts
 
 7.2
-        - bugfix: the selection speed in in 7.1 unfortunately caused
+        - bugfix: the selection speedup in in 7.1 unfortunately caused
           non-ascii characters to enlarge/move the selection.
         - added "tabbed" extension that provides a crude tabbed terminal.
         - added "readline" extension that allows cursor positioning
diff --git a/src/perl/selection b/src/perl/selection
index 10e40011..fc4d2e38 100644
--- a/src/perl/selection
+++ b/src/perl/selection
@@ -69,26 +69,35 @@ sub on_sel_extend {
 
    my @matches;
 
-   # not doing matches in unicode mode helps speed
-   # enourmously here. working in utf-8 should be
-   # equivalent due to the magic of utf-8 encoding.
-   utf8::encode $text;
-   study $text; # _really_ helps, too :)
-
-   for my $regex (@mark_patterns, @{ $self->{patterns} }) {
-      while ($text =~ /$regex/g) {
-         if ($-[1] <= $markofs and $markofs <= $+[1]) {
-            my $ofs = $-[1];
-            my $match = $1;
-
-            for my $regex (@simplify_patterns) {
-               if ($match =~ $regex) {
-                  $match = $1;
-                  $ofs += $-[1];
+   if ($markofs < $line->l) {
+      # convert markofs form character to UTF-8 offset space
+      {
+         my $prefix = substr $text, 0, $markofs;
+         utf8::encode $prefix;
+         $markofs = length $prefix;
+      }
+
+      # not doing matches in unicode mode helps speed
+      # enourmously here. working in utf-8 should be
+      # equivalent due to the magic of utf-8 encoding.
+      utf8::encode $text;
+      study $text; # _really_ helps, too :)
+
+      for my $regex (@mark_patterns, @{ $self->{patterns} }) {
+         while ($text =~ /$regex/g) {
+            if ($-[1] <= $markofs and $markofs <= $+[1]) {
+               my $ofs = $-[1];
+               my $match = $1;
+
+               for my $regex (@simplify_patterns) {
+                  if ($match =~ $regex) {
+                     $match = $1;
+                     $ofs += $-[1];
+                  }
                }
-            }
 
-            push @matches, [$ofs, length $match];
+               push @matches, [$ofs, length $match];
+            }
          }
       }
    }
@@ -101,6 +110,18 @@ sub on_sel_extend {
 
       next if $len <= $curlen;
 
+      # convert back from UTF-8 offset space to character space
+      {
+         my $length = substr $text, $ofs, $len;
+         utf8::decode $length;
+         $len = length $length;
+      }
+      {
+         my $prefix = substr $text, 0, $ofs;
+         utf8::decode $prefix;
+         $ofs = length $prefix;
+      }
+
       $self->selection_beg ($line->coord_of ($ofs));
       $self->selection_end ($line->coord_of ($ofs + $len));
       return 1;
-- 
2.34.1