Re: "make distclean" does not bring back build tree to previous state

[Gavin Smith]

On Sun, Dec 10, 2023 at 04:00:56PM +0100, Preuße, Hilmar wrote:
> Hello,
> 
> I got a report telling that "make distclean" does not bring back the build
> tree into the original state. After running a build (configure line below)
> and calling "make distclean", we have a few differences. Some files were
> deleted, some files were re-generated and hence show a different time stamp
> inside the file:

We'll probably have to investigate each of these changes separately to
see where they occur.  I am probably not going to fix them all in one
go.

First, could you confirm which version of Texinfo you got these results
with.

I tested with Texinfo 7.1, ran configure with the same configure line
as you, then ran "make distclean".

> 
> Changes not staged for commit:
>   (use "git add/rm <file>..." to update what will be committed)
>   (use "git restore <file>..." to discard changes in working directory)
>         deleted:    INSTALL

This file was not deleted for me.  But it definitely should not be
deleted and if it is, there is a bug.  Please explain exactly what you
did to make this happen.

>         deleted:    build-aux/compile
>         deleted:    build-aux/config.rpath
>         deleted:    build-aux/depcomp
>         deleted:    build-aux/mdate-sh

I checked that all these files were still present.

>         modified:   doc/stamp-1
>         modified:   doc/stamp-2
>         modified:   doc/stamp-vti
>         modified:   doc/version-stnd.texi
>         modified:   doc/version-texi2any_api.texi
>         modified:   doc/version.texi

These files were not modified, although it is possible that this was because
I did not run "make" first.

>         modified:   man/makeinfo.1
>         modified:   man/texi2dvi.1
>         modified:   man/texindex.1

Likewise.

>         modified:   po_document/ca.po

Likewise.  po files being updated is a perennial nuisance that I thought
we had got right with documenting release procedures exactly, but evidently
it still isn't right.

>         modified:   tp/Texinfo/XS/TestXS.c

This file was modified.  It is probably because xsubpp ran to regenerate
TestXS.c.

This issue was previously reported:

https://lists.gnu.org/archive/html/bug-texinfo/2021-02/msg00160.html
https://lists.gnu.org/archive/html/bug-texinfo/2021-03/msg00030.html
https://lists.gnu.org/archive/html/bug-texinfo/2021-08/msg00021.html

- but I did not want to make any changes at that time, due to how
fragile the XS build code was and the risk of breaking something.

Obviously clobbering the pristine TestXS.c with no way to get back the
original version (different because of varying output for different
versions of xsubpp) is not acceptable.

I am going to make a change not to distribute TestXS.c, so that xsubpp
will always be required for building XS modules.

Later then I plan to add this and other fixes to the release branch and
make a new Texinfo 7.1.1 release where "make distclean" works correctly.

> (I do the test inside a git repo to make the difference easily visible).
> 
> Would be nice if you could have a look at the issue. Are the deleted /
> re-generated files really needed for building? Feel free to contact for
> questions, please keep me in Cc, I'm not subscribed.
> 
>         AWK=awk ./configure \
>           --with-external-Text-Unidecode=yes \
>           --prefix=/usr \
>           --libexecdir='$${prefix}/lib' \
>           --infodir='$${prefix}/share/info' --mandir='$${prefix}/share/man'
> 
> Many thanks,
>   Hilmar
> -- 
> sigfault
>