Re: [PATCH v9 2/2] Add generic C.UTF-8 locale (Bug 17318)

[Florian Weimer via Libc-alpha <libc-alpha@sourceware.org>]

* Carlos O'Donell:

> diff --git a/NEWS b/NEWS
> index 79c895e382..807105a596 100644
> --- a/NEWS
> +++ b/NEWS
> @@ -9,7 +9,15 @@ Version 2.35
>  
>  Major new features:
>  
> -  [Add new features here]
> +* Support for the C.UTF-8 locale has been added to glibc.  The locale
> +  supports full code-point sorting for all valid Unicode code points.
> +  A limitation in the framework for fnmatch, regexec, and regcomp requires
> +  a compromise to save space and only ASCII-based range expressions are
> +  supported for now (see bug 28255).  The full size of the locale is only
> +  ~400KiB, with 346KiB coming from LC_CTYPE information for Unicode. This
> +  locale harmonizes downstream C.UTF-8 already shipping in Gentoo, Debian,
> +  Ubuntu, Fedora, CentOS Stream, and RHEL.  The locale is not built into
> +  glibc, and must be installed.

I would say “various downstream distributions”.  You left out SUSE's
distributions, and they have C.UTF-8 as well:

<https://build.opensuse.org/package/view_file/openSUSE:Factory/glibc/glibc-c-utf8-locale.patch?expand=1>

> --- /dev/null
> +++ b/iconv/tst-iconv9.c

> +  /* From ISO-8859-1 to ASCII. */

> +  /* From UTF-8 to ASCII. */

Missing spaces after “.”.

> diff --git a/posix/transbug.c b/posix/transbug.c
> index d0983b4d44..71632b7976 100644
> --- a/posix/transbug.c
> +++ b/posix/transbug.c
> @@ -116,14 +116,30 @@ do_test (void)
>    static const char lower[] = "[[:lower:]]+";
>    static const char upper[] = "[[:upper:]]+";
>    struct re_registers regs[4];
> +  int result;
>  
> +#define CHECK(exp) \
> +  if (exp) { puts (#exp); result = 1; }
> +
> +  printf ("INFO: Checking C.\n");
>    setlocale (LC_ALL, "C");
>  
>    (void) re_set_syntax (RE_SYNTAX_GNU_AWK);
>  
> -  int result;
> -#define CHECK(exp) \
> -  if (exp) { puts (#exp); result = 1; }
> +  result = run_test (lower, regs);
> +  result |= run_test (upper, &regs[2]);
> +  if (! result)
> +    {
> +      CHECK (regs[0].start[0] != regs[2].start[0]);
> +      CHECK (regs[0].end[0] != regs[2].end[0]);
> +      CHECK (regs[1].start[0] != regs[3].start[0]);
> +      CHECK (regs[1].end[0] != regs[3].end[0]);
> +    }
> +
> +  printf ("INFO: Checking C.UTF-8.\n");
> +  setlocale (LC_ALL, "C.UTF-8");
> +
> +  (void) re_set_syntax (RE_SYNTAX_GNU_AWK);
>  
>    result = run_test (lower, regs);
>    result |= run_test (upper, &regs[2]);

The second-to-last line overwrites the previous test results.

I think this can go in if you address those nits.

Thanks,
Florian