> Date: Mon, 10 Oct 2011 18:24:09 +0800 > From: li bo <[email protected]> > Cc: [email protected] > > ‘You cannot do this in the first place, because the parts of the UBA > before that need to distinguish between segment separators, paragraph > separators, and whitespace.’ > I think you lose a ‘not’ behind 'need'. Is that right?
No, the text is correct as written. >From the beginning of UBA, where you see the first call to `classify', and up until L1, segment separators, paragraph separators, and whitespace characters are processed differently. So the first call to `classify' returns their nominal types. In L1, segment separators, paragraph separators, and whitespace characters are treated in the same way, and so `classify' returns their types folded to a single type "N".
