[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: A target that's even more bootstrap?
From: |
Lars Ingebrigtsen |
Subject: |
Re: A target that's even more bootstrap? |
Date: |
Fri, 21 Jun 2019 12:51:38 +0200 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/27.0.50 (gnu/linux) |
Eli Zaretskii <address@hidden> writes:
> I don't see this as a catastrophe. If the user does this inside a
> repository, they can "git checkout src/macuvs.h", if they don't intend
> running a full build. And if they do run a full build, the file will
> be recreated. And in a tree that is not a repository, removing the
> file is not a problem at all.
I just wondered whether the inclusion of the file is a bug, or the
removal of the file was a bug...
We do include some generated files in git, like
lisp/cedet/semantic/grammar-wy.el. But that's because we can't
(re)generate it without having it. :-) (wisent uses its own machinery
to parse its .wy files, from which this is generated, so it's a chicken
or egg problem.)
But Emacs does generate the macuvs.h file just fine, so it just seemed
odd to me and I wondered whether anybody could remember why it's this
way...
I had a look in both the file itself:
/* This file was automatically generated from admin/unidata/IVD_Sequences.txt
by the script admin/unidata/uvs.el */
and the file that generates the file, but I couldn't see any
explanation.
Here's the commit that added the generation of the file (and the removal
of the file in extraclean), but it didn't remove the file from git.
Paul, do you remember the logic behind this?
commit cf2f54c4e3d8adb26e3f4b8e9e51364165877444
Author: Paul Eggert <address@hidden>
Date: Mon May 26 08:48:28 2014 -0700
Include sources used to create macuvs.h.
* admin/unidata/IVD_Sequences.txt: New file.
* admin/unidata/Makefile.in (${top_srcdir}/src/macuvs.h): New rule.
(all): Build it.
(extraclean): Remove it.
* admin/unidata/README: Mention BidiMirroring.txt and IVD_Sequences.txt.
* admin/unidata/copyright.html: Update to current version from Unicode
Consortium.
* admin/unidata/uvs.el: Rename from admin/mac/uvs.el.
(uvs-print-table-ivd): Output a header in the form that
unidata-gen.el generates.
* lisp/international/README: Refer to the Unicode Terms of Use rather
than copying it bodily here, as that simplifies maintenance.
* src/Makefile.in ($(srcdir)/macuvs.h): New rule.
* src/macuvs.h: Use automatically-generated header.
--
(domestic pets only, the antidote for overdose, milk.)
bloggy blog: http://lars.ingebrigtsen.no
- Re: A target that's even more bootstrap?, (continued)
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/19
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/19
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/19
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/19
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/19
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/20
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/20
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/20
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/20
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/20
- Re: A target that's even more bootstrap?,
Lars Ingebrigtsen <=
- A target that's even more bootstrap?, Paul Eggert, 2019/06/21
- Re: A target that's even more bootstrap?, Richard Stallman, 2019/06/22
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/20
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/20
- Re: A target that's even more bootstrap?, Eli Zaretskii, 2019/06/20
- Re: A target that's even more bootstrap?, Lars Ingebrigtsen, 2019/06/20
- Re: A target that's even more bootstrap?, Paul Eggert, 2019/06/20
- Re: A target that's even more bootstrap?, Phillip Lord, 2019/06/19
- Re: A target that's even more bootstrap?, Óscar Fuentes, 2019/06/19
- Re: A target that's even more bootstrap?, Stefan Monnier, 2019/06/19