Re: Bug with gawk regex interval processing

[Stepan Kasal]

Hello,

On Mon, Apr 21, 2003 at 10:26:07AM -0500, Mike Kistler wrote:
> # This problem occurs on RedHat 8.0.  awk is packages in gawk-3.1.1-4

> awk --posix '{sub(/^[ ]*([^ ]+[ ]+){3}/, "", $0); print $0}' /tmp/awkbug.dat
> awk --posix '{sub(/^[ ]*([^ ]+[ ]+)([^ ]+[ ]+)([^ ]+[ ]+)/, "", $0); print 
> $0}' /tmp/awkbug.dat

I've tried to reproduce the problem.  My system is a RedHat, mixed
versions.  I tested gawk 3.1.1, gawk 3.1.2, both compiled manually.
Then I tested gawk-3.1.1-4, compiled from src.rpm .

I have never experienced the bug you mention, sorry.

Perhaps it's because you have a locale variable set (LC_ALL, LANG, ...).
If that's the case, you might try to upgrade to 3.1.2, it has a completely
new regex library which might have the problem fixed.

HTH,
        Stepan Kasal