On Fri, Oct 01, 2021 at 11:55:47AM +0200, David Marchand wrote:
> On Fri, Oct 1, 2021 at 11:51 AM Dmitry Kozlyuk <dmitry.kozl...@gmail.com>
> wrote:
> >
> > 2021-10-01 12:36 (UTC+0300), Dmitry Kozlyuk:
> > > 2021-10-01 09:27 (UTC+0200), David Marchand:
> > > > [...]
> > > > I just have a concern that headers get broken again if we have no check.
> > > > Could buildtools/chkincs do the job (if we make this check work on
> > > > Windows)?
> > >
> > > It's a valid caoncern and yes, chkincs would be a solution.
>
> Talked with Thomas, in any case, we can proceed with this patch.
>
> The check is something to add/make work the sooner the better.
>
> > >
> > > I discovered that the real issue with chkincs was a space after shebang in
> > > gen_c_file_for_header.py. If it is removed, find_program() is able to
> > > parse
> > > the line and call the script with Python interpreter. However, spaces
> > > *are* allowed after shebangs. Would you like me to submit a patch with a
> > > fix
> > > just chkincs or shall I *instead* work with meson developers (I'll do that
> > > anyway)? I'm asking because we have other files with extra spaces, so
> > > there's
> > > a question if we should fix them, add a rule for the space, etc.
> >
> > Just re-checked with meson 0.57.0 (as recommended in Windows GSG) and hit no
> > issue even with a space after "#!". Maybe it was a meson 0.47 issue on
> > Windows at the time you disabled chkins there?
>
> Maybe, I am not sure, Bruce had put this limitation.
>
I'm not sure either, sorry.
