Re: occur slower than grep by a longshot

[Dan Jacobson]

>>>>> "E" == Eli Zaretskii <eliz@is.elta.co.il> writes:
>> Fellas, it's just so obvious,
>> the difference in speed between
>> (shell-command-on-region (region-beginning) (region-end) "grep /" nil nil 
>> nil)
>> and
>> (occur "/" nil) [<--slow]
>> on say a 20000 line ls-R ftp listing.

E> What's your point?  Isn't it obvious that a Lisp code run by an
E> interpreter would be slower than a tight optimized C program?

I dunno, maybe internally have the real grep to do the hard work, by
default, but keep the original occur code around for in case the user
does not have grep on their operating system?
-- 
http://www.geocities.com/jidanni/ Tel+886-4-25854780