[text] Use cached length when possible

Since clutter_text_set_text() measures the length of the text
each time, we should use the cached length instead of recomputing
the text length each time. This should save us some time when
dealing with long, multi-byte texts.

diff --git a/clutter/clutter-text.c b/clutter/clutter-text.c
index 0b8deb1..da74de1 100644 (file)
--- a/clutter/clutter-text.c
+++ b/clutter/clutter-text.c
@@ -1143,7 +1143,7 @@ clutter_text_real_move_left (ClutterText         *self,
   gint pos = priv->position;
   gint len;
 
-  len = g_utf8_strlen (priv->text, -1);
+  len = priv->n_chars;
 
   if (pos != 0 && len !=0)
     {
@@ -1169,7 +1169,7 @@ clutter_text_real_move_right (ClutterText         *self,
   gint pos = priv->position;
   gint len;
 
-  len = g_utf8_strlen (priv->text, -1);
+  len = priv->n_chars;
 
   if (pos != -1 && len !=0)
     {
@@ -3190,7 +3190,7 @@ clutter_text_set_cursor_position (ClutterText *self,
 
   priv = self->priv;
 
-  len = g_utf8_strlen (priv->text, -1);
+  len = priv->n_chars;
 
   if (position < 0 || position >= len)
     priv->position = -1;
@@ -3335,7 +3335,7 @@ clutter_text_set_max_length (ClutterText *self,
   if (priv->max_length != max)
     {
       if (max < 0)
-        max = g_utf8_strlen (priv->text, -1);
+        max = priv->n_chars;
 
       priv->max_length = max;
 
@@ -3528,7 +3528,7 @@ clutter_text_delete_chars (ClutterText *self,
   if (!priv->text)
     return;
 
-  len = g_utf8_strlen (priv->text, -1);
+  len = priv->n_chars;
   new = g_string_new (priv->text);
 
   if (priv->position == -1)