text_props->info = calloc(1, sizeof(Evas_Text_Props_Info));
fi = fn->fonts->data;
- /* evas_common_font_size_use(fn); */
- evas_common_font_int_reload(fi);
- if (fi->src->current_size != fi->size)
- {
- FTLOCK();
- FT_Activate_Size(fi->ft.size);
- FTUNLOCK();
- fi->src->current_size = fi->size;
- }
-
/* Load the glyph according to the first letter of the script, pretty
* bad, but will have to do */
{
base_char++)
;
if (!*base_char && (base_char > text)) base_char--;
+
evas_common_font_glyph_search(fn, &fi, *base_char);
}
text_props->font_instance = fi;
+ evas_common_font_int_reload(fi);
+ if (fi->src->current_size != fi->size)
+ {
+ FTLOCK();
+ FT_Activate_Size(fi->ft.size);
+ FTUNLOCK();
+ fi->src->current_size = fi->size;
+ }
+
#ifdef OT_SUPPORT
size_t char_index;
Evas_Font_Glyph_Info *gl_itr;
index = evas_common_get_char_index(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)