I'm pretty sure its to do with flex allocating + 2 onto the end of the
original size to look for 2 null bytes to indicate the last token. In
some cases the last token gets lost.
This may be something that needs handled in the parser rather than the
tokenizer. I'd need to investigate some more.
Hi Xuefer, all,
Yeah, the re2c patch did change (or add?) a bit for handling the "#"
character (though I haven't investigated your example...).
However, I wanted to report that I just started playing with some changes to
the scanner (re2c), and am seeing weird (or "wired" re: subject :-P) stuff
t
i'm having trouble with the bug http://bugs.php.net/bug.php?id=44654 i
just reported. i'm not sure if it's a expected behavoir or a bug
introduced by re2c or whatever patch.
On Sun, Apr 6, 2008 at 10:19 PM, PHP Bug Database
<[EMAIL PROTECTED]> wrote:
> From: xuefer at gmail dot com
>
