This attempt at optimization is actually detrimental on modern CPUs.
Removing it helps playback speed ~0.3%. mlt_properties_get_data() is now
down to ~0.6%.
{
// Check if we're hashed
if ( list->count > 0 &&
- name[ 0 ] == list->name[ i ][ 0 ] &&
!strcmp( list->name[ i ], name ) )
value = list->value[ i ];
// Locate the item
for ( i = list->count - 1; value == NULL && i >= 0; i -- )
- if ( name[ 0 ] == list->name[ i ][ 0 ] && !strcmp( list->name[ i ], name ) )
+ if ( !strcmp( list->name[ i ], name ) )
value = list->value[ i ];
}
mlt_properties_unlock( self );