Re: will we ever have zero width assertions in regexps?

[Ilya Zakharevich <nospam-abuse@ilyaz.org>]

On 2011-02-07, Stefan Monnier <monnier@iro.umontreal.ca> wrote:
>> Hmm, I thought about a different scenario: if the bit is set, then one
>> switches to a DIFFERENT REx designed for a multi-line case.  Otherwise
>> why not just run it against the rest of the buffer, instead of
>> one-line?
>
> Because we don't want to match those regexps against the whole buffer
> every time the buffer is modified (the buffer may be large).

I think there is some confusion here: I did not want to mean that you
run "another REx" via re-search(), but via looking-at() at the
corresponding position...  Essentially: do you want the REx to start
match on the current line, or end match on the current line?

Or maybe you mean a different thing, and
WANT e.g. [^a-z] to actually "mean" [^a-z\n] ?  (So only an explicit
non-negated \n in a REx may match \n in a buffer?)

> BTW, another reason to want a non-backtracking matcher can be seen in
> the recent thread "Stack overflow in regexp matcher".

I think this is a red herring.  Try to stack-overflow the Perl
matcher...  (Possible, but one must be very malicious to hit these
situations.)

Ilya