NOTE -- RFC 1866 section 4.2.2 specifies that an HTML user agent should treat an end of line as a word space, except in preformatted text. This should be interpreted in the context of the script being processed, as the way words are separated in writing is script-dependent. For some scripts (e.g. Latin), a word space is just a space, but in other scripts (e.g. Thai) it is a zero-width word separator, whereas in yet other scripts (e.g. Japanese) it is nothing at all, i.e. totally ignored.
