I'm still not sure if it's just a workaround that just fixes a symptom
of a bigger issue, or actually needed, but this is good until I figure
this out completely.
git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@57761
7cbeb6ba-43b4-40fd-8cce-
4c39aea84d33
{
index = evas_common_font_glyph_search(fn, &fi, REPLACEMENT_CHAR);
}
+
+ /* Should we really set the size per fi? This fixes a bug for Korean
+ * because for some reason different fis are chosen for different
+ * chars in some cases. But we should find the source of the problem
+ * and not just fix the symptom. */
+ if (fi->src->current_size != fi->size)
+ {
+ FTLOCK();
+ FT_Activate_Size(fi->ft.size);
+ FTUNLOCK();
+ fi->src->current_size = fi->size;
+ }
+
LKL(fi->ft_mutex);
fg = evas_common_font_int_cache_glyph_get(fi, index);
if (!fg)