[gnote] Fix a regression in the URL detection introduced in 0.5.0

[Hubert Figuière <hub src gnome org>]

commit 26449802d19b8c9ce88a8855fce9c3f219e27f7c
Author: Hubert Figuiere <hub figuiere net>
Date:   Fri Jun 19 17:48:24 2009 -0400

    Fix a regression in the URL detection introduced in 0.5.0
    when the URL is in a list or withing text with non-ASCII
    characters.

 NEWS             |    4 ++++
 src/watchers.cpp |    6 ++++--
 2 files changed, 8 insertions(+), 2 deletions(-)
---
diff --git a/NEWS b/NEWS
index b5e3813..49c60d2 100644
--- a/NEWS
+++ b/NEWS
@@ -2,6 +2,10 @@
 
 Fixes:
 
+  * Fix a regression in the URL detection introduced in 0.5.0
+    when the URL is in a list or withing text with non-ASCII
+    characters.
+
 Translations:
 
   * Updated translations:
diff --git a/src/watchers.cpp b/src/watchers.cpp
index a5c0095..91d4fc5 100644
--- a/src/watchers.cpp
+++ b/src/watchers.cpp
@@ -519,7 +519,8 @@ namespace gnote {
       // must construct the Glib::ustring from a char *.
       // otherwise it expect the num of chars (UTF-8) instead of bytes.
       // here we compute the index of the URL. It is anchor - start - match.
-      Glib::ustring segment(p, input.data() - p - match1.size());
+      Glib::ustring::size_type len = input.data() - p - match1.size();
+      Glib::ustring segment(p, p + len);
       start_cpy.forward_chars (segment.size());
 
       end = start_cpy;
@@ -1036,7 +1037,8 @@ namespace gnote {
       if (!is_patronymic_name (match)) {
       
         Gtk::TextIter start_cpy = start;
-        Glib::ustring segment(std::string(p, input.data() - p - match.size()));
+        Glib::ustring::size_type len = input.data() - p - match.size();
+        Glib::ustring segment(p, p + len);
         start_cpy.forward_chars (segment.length());
 
         DBG_OUT("Highlighting wikiword: '%s' at offset %d",