Instead of assuming that the first font in the cascade font list will have a glyph for space, we need to find it in the list taking into account unicode ranges.
