[grep] about 200 times slower than Sed or Perl

[Eduard Bloch]

Hello,

today I tried to weed out some useles lines from a large log file and
was amazed to see that the current grep versions (and fgrep) are
slow like hell.

debian:/# time fgrep -v MAC /var/log/kern.log  | tail
...
real    0m25.092s
user    0m24.902s
sys     0m0.006s
debian:/# time perl -pe 'undef $_ if /MAC/' < /var/log/kern.log  | tail
...
real    0m0.148s
user    0m0.065s
sys     0m0.007s

I don't think that this is a bug in POSIX regexp implementation, since
Sed is much faster (but still 7 times slower than Perl, lol):

debian:/# time sed -e 's/MAC/foobar/' < /var/log/kern.log  | tail
...
real    0m0.747s
user    0m0.719s
sys     0m0.016s
debian:/# time perl -pe 's/MAC/foobar/' < /var/log/kern.log  | tail
...
real    0m0.123s
user    0m0.090s
sys     0m0.016s

MfG,
Eduard.