Re: [patch #4596] "egrep" and "fgrep" as minimal argv[0]-independent executables

[Charles Levert]

* On Friday 2005-11-04 at 13:07:32 +0000, Julian Foad wrote:
> Charles Levert wrote:
> >* On Friday 2005-11-04 at 04:22:53 +0000, Charles Levert wrote:
> >
> >> <http://savannah.gnu.org/patch/?func=detailitem&item_id=4596>
> 
> >>Full ChangeLog entry within.  Apply relative to current CVS.
> 
> I get errors doing that:
> 
> ~/src/grep> patch -p1 < ~/tmp/gefp.patch
> patching file ChangeLog
> Hunk #1 FAILED at 10.
> 1 out of 1 hunk FAILED -- saving rejects to file ChangeLog.rej
> patching file src/grep.c
> Hunk #18 FAILED at 2081.
> 1 out of 18 hunks FAILED -- saving rejects to file src/grep.c.rej

I'll see if I can't regenerate it.  It's strange
since I used "cvs diff -u" to generate it.

Are you sure you're not working out of an earlier
version of what I checked in for mb_icase_keys()?
I also added something to tests/foad1.sh
without asking since more tests should not be
controversial; this affected the ChangeLog.

I have made improvements to the patch since then:

   -- Remove no-dependencies from AUTOMAKE_OPTIONS
      in src/Makefile.am, since #including another
      C file needs to be properly tracked.

   -- Big change:  merge Gcompile() with
      EAcompile() to form GEAcompile().
      They were nearly identical, i.e., there
      was lots of duplicated code, which makes it
      too easy to fix a bug in one duplicate but
      not the other.  The new function becomes
      just compile() for egrep.