Re: Dealing with character ranges in grep

[Johannes Meixner]

Hello,

On Jun 16 13:51 Stanislav Brabec wrote (excerpt):
> grep in openSUSE uses glibc regex by default.

Yes.

Currently grep in openSUSE is built using
"configure --without-included-regex"
as it was built for openSUSE all the time.


Perhaps there is a misunderstanding what I mean.

What I mean is:

Currently there is inconsistence and any kind
of unexpected behaviour and subtle differences.

Therefore I support and appreciate any change from
upstream which leads to consistence in the future.


> So if we will want consistent support for [=e=], then
> it should be in glibc.

Isn't support for [=e=] already in glibc?

But there is "surprising consequence ...  when the tool is built
to use glibc's regexp code ... end up causing subtle problems".

If the glibc upstream authors would agree to eliminate
the "surprising consequences causing subtle problems"
in glibc, there would be no problem - but as far
as I know they didn't agree.

Therefore the upstream authors of grep, gawk, and sed
could agree on their own consistent way for regex handling
(without "surprising consequences causing subtle problems")
and they could eliminate the --with-included-regex and
--without-included-regex configure options to enforce
using their own included regex handling in any case.

I would support and appreciate such a change.


> Note that libpcre documentation cleanly states, that [=e=]
> is not supported tere.

It is o.k. for me if grep, gawk, and sed are consistent with
Perl Compatible Regular Expressions (at least in this case).

If each and any special stuff in each and any tool is sacrosanct,
nothing can be changed towards more consistency.


Again:
I do not care if this or that special feature is supported or not
because I think that consistent behaviour has topmost priority.


Kind Regards
Johannes Meixner
--
SUSE LINUX Products GmbH -- Maxfeldstrasse 5 -- 90409 Nuernberg -- Germany
HRB 16746 (AG Nuernberg) GF: Jeff Hawn, Jennifer Guild, Felix Imendoerffer