Re: [evolution-patches] #42952 (GAL) second try

[Mike Kestner <mkestner ximian com>]

On Tue, 2003-05-20 at 17:05, Ettore Perazzoli wrote:
> The first fix didn't quite fix it...

Countering with an altered patch that should clean up some valgrindage
I've seen from Mark and Jeff recently related to this function.

Mike

Index: gal/e-table/e-tree.c
===================================================================
RCS file: /cvs/gnome/gal/gal/e-table/e-tree.c,v
retrieving revision 1.101
diff -u -p -r1.101 e-tree.c
--- gal/e-table/e-tree.c	15 May 2003 20:58:15 -0000	1.101
+++ gal/e-table/e-tree.c	20 May 2003 22:22:02 -0000
@@ -776,10 +776,15 @@ tree_canvas_size_allocate (GtkWidget *wi
 		g_source_remove(e_tree->priv->reflow_idle_id);
 	tree_canvas_reflow_idle(e_tree);
 
-	if (path)
-		e_tree_get_cell_geometry (e_tree, e_tree_row_of_node(e_tree, path), 0, &x, &y, &w, &h);
-	else
-		y = h = 0;
+	x = y = w = h = 0;
+	if (path) {
+		int row = e_tree_row_of_node(e_tree, path);
+		int col = 0;
+
+		if (row >= 0)
+			e_table_item_get_cell_geometry (E_TABLE_ITEM (e_tree->priv->item),
+							&row, &col, &x, &y, &w, &h);
+	}
 
  	if (y < adj->value || y + h > adj->value + adj->page_size)
 		gtk_adjustment_set_value(adj, CLAMP(y - adj->page_size / 2, adj->lower, adj->upper - adj->page_size));