#2054: Grep-2.5.3
------------------------------------------+---------------------------------
Reporter: [EMAIL PROTECTED] | Owner: [EMAIL PROTECTED]
Type: enhancement | Status: assigned
Priority: normal | Milestone: 7.0
Component: Book | Version: SVN
Severity: normal | Resolution:
Keywords: |
------------------------------------------+---------------------------------
Comment (by [EMAIL PROTECTED]):
> As it is, I've just stumbled across the grep CVS commits list. I wonder
whether simply backporting http://lists.gnu.org/archive/html/grep-
commit/2008-02/msg00005.html and http://lists.gnu.org/archive/html/grep-
commit/2008-02/msg00004.html and anything similar would be the quickest
option for now, while upstream work on fixing things up for Grep-2.5.4?
Thanks for that. I couldn't get 00004 to apply (line-wrapping didn't
help), but I found a link to the daily snapshots and applied the changes
to foad1 and yesno from that. I'll take the yesno change, thank you! For
foad1, with an armful of debian patches, I seem to get an ''extra''
failure with the patch. Ah, it's the (new) -H -h test and the output is
"(standard input):wA wB/" instead of "wA wB/" - the marker is a warning
somewhere in the debian patchset. Will try to fix the expected result.
There are also documentation updates for "-e PATTERN" which we might as
well have.
Current proposal - let's take all the debian patches except their manpage
changes (they omit reference to info pages because they remove them under
the dfsg), plus upstream docs, yesno, and (adjusted) foad1. This includes
debian's 55-bigfile patch - I can't find a testcase, and I don't intend to
create a file bigger than 4GB to search through (I suspect this might be
'bigfile' as in 'more than 32 bits long'), but it's definitely playing in
the multibyte charset space. So, a debian patch and an upstream patch
(with a note that it is adapted to play nice with the debian patch).
Omit the mandriva patches. I don't have a testcase for their mbcset
change. The change to the test scripts is mostly replaced by the yesno
patch.
This should leave us with a headline "2 of 14 tests failed", and 42 and 2
failures within them (we started with "3 of 14" and 58, 2, 3). It should
also fix the excessive slowness in multibyte locales.
I'll try and get the patches prepared this weekend, and retest them.
--
Ticket URL: <http://wiki.linuxfromscratch.org/lfs/ticket/2054#comment:18>
LFS Trac <http://wiki.linuxfromscratch.org/lfs/>
Linux From Scratch: Your Distro, Your Rules.
--
http://linuxfromscratch.org/mailman/listinfo/lfs-book
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
