Re: RRI patches for grep

[Paolo Bonzini]

Il 27/04/2012 11:07, Aharon Robbins ha scritto:
> From 5c7665f2ced46d2e830958bce1bf46469995d3de Mon Sep 17 00:00:00 2001
> From: Arnold D. Robbins <address@hidden>
> Date: Fri, 27 Apr 2012 12:04:22 +0300
> Subject: [PATCH] Implement Rational Range Interpretation in Gnulib.
> 
> ---
>  lib/regcomp.c |   13 +++----------
>  lib/regexec.c |   12 ++----------
>  2 files changed, 5 insertions(+), 20 deletions(-)
> 
> diff --git a/lib/regcomp.c b/lib/regcomp.c
> index b51a9a6..7748535 100644
> --- a/lib/regcomp.c
> +++ b/lib/regcomp.c
> @@ -2702,7 +2702,6 @@ build_range_exp (const reg_syntax_t syntax,
>      wchar_t wc;
>      wint_t start_wc;
>      wint_t end_wc;
> -    wchar_t cmp_buf[6] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
>  
>      start_ch = ((start_elem->type == SB_CHAR) ? start_elem->opr.ch
>               : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0]
> @@ -2716,11 +2715,7 @@ build_range_exp (const reg_syntax_t syntax,
>             ? __btowc (end_ch) : end_elem->opr.wch);
>      if (start_wc == WEOF || end_wc == WEOF)
>        return REG_ECOLLATE;
> -    cmp_buf[0] = start_wc;
> -    cmp_buf[4] = end_wc;
> -
> -    if (BE ((syntax & RE_NO_EMPTY_RANGES)
> -            && wcscoll (cmp_buf, cmp_buf + 4) > 0, 0))
> +    else if ((syntax & RE_NO_EMPTY_RANGES) && start_wc > end_wc)
>        return REG_ERANGE;
>  
>      /* Got valid collation sequence values, add them as a new entry.
> @@ -2761,10 +2756,8 @@ build_range_exp (const reg_syntax_t syntax,
>      /* Build the table for single byte characters.  */
>      for (wc = 0; wc < SBC_MAX; ++wc)
>        {
> -     cmp_buf[2] = wc;
> -     if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
> -         && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
> -       bitset_set (sbcset, wc);
> +         if (start_wc <= wc && wc <= end_wc)
> +           bitset_set (sbcset, wc);
>        }
>    }
>  # else /* not RE_ENABLE_I18N */
> diff --git a/lib/regexec.c b/lib/regexec.c
> index 92efb44..5a6a0dc 100644
> --- a/lib/regexec.c
> +++ b/lib/regexec.c
> @@ -3986,18 +3986,10 @@ check_node_accept_bytes (const re_dfa_t *dfa, Idx 
> node_idx,
>  # endif /* _LIBC */
>       {
>         /* match with range expression?  */
> -#if __GNUC__ >= 2 && ! (__STDC_VERSION__ < 199901L && defined 
> __STRICT_ANSI__)
> -       wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'};
> -#else
> -       wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
> -       cmp_buf[2] = wc;
> -#endif
>         for (i = 0; i < cset->nranges; ++i)
>           {
> -           cmp_buf[0] = cset->range_starts[i];
> -           cmp_buf[4] = cset->range_ends[i];
> -           if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
> -               && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
> +              if (cset->range_starts[i] <= wc
> +                  && wc <= cset->range_ends[i])
>               {
>                 match_len = char_len;
>                 goto check_node_accept_bytes_match;
> -- 1.7.1

These look good.

Paolo