texi2dvi has a bad egrep line:
egrep: Invalid range end
/usr/bin/texi2dvi: cannot read .//some/path/and/file.texi, skipping.
The offending line is here:
1686: echo "$command_line_filename" | $EGREP '^(/|[A-z]:/)' >&6 \
The problem is locale-dependent. With LC_ALL=C it works, with LANG=de_DE.utf8 it doesn't. I suggest you replace 'A-z' with 'A-Za-z', or pass LC_ALL=C to the egrep invocation.
Probably this is a bug report for upstream as well. Will investigate.
Fix comitted there:
I just added an texinfo-4.13-r1 ebuild to my overlay 'stuge' which includes the fixed upstream texi2dvi revision.
And https://bugs.gentoo.org/show_bug.cgi?id=343907 is the -r1 ebuild for inclusion in portage
*** Bug 343907 has been marked as a duplicate of this bug. ***
*** Bug 348004 has been marked as a duplicate of this bug. ***
Created attachment 257504 [details, diff]
Fix from upstream
This is the fix taken from the upstream repository, but stripped of keyword substitutions so it will apply easily. Please include this in portage.
(In reply to comment #6)
> This is the fix taken from the upstream repository,
The -r1 ebuild and the patch that I attached to #343907 and have included in my overlay 'stuge' include this fix. I hope that the ebuild has some chance of going into portage despite the fact that the ebuild bug was closed.
Meanwhile, you can use layman to grab my overlay, to easily get the ebuild:
layman -a stuge
I just hit again this issue. How are the chances to add the patches to the main tree?
(In reply to comment #8)
> I just hit again this issue. How are the chances to add the patches to the main
Dunno, but you can use the ebuild in my overlay fairly easily:
layman -a stuge
echo 'source /var/lib/layman/make.conf' >> /etc/make.conf
emerge =sys-apps/texinfo-4.13-r1 -av
I have an patched ebuild, that is not the problem. But if I install gentoo at a new PC and the installation hangs before a create a new overlay texinfo, this is nasty. That's why I asked.
usually i ignore texinfo bugs (since i hate the format) and they magically seem to resolve themselves. i guess that didnt happen here. oh well.
should be all set with 4.13-r1.