Re: SRFI-14 and locale settings

[Kevin Ryde]

address@hidden (Ludovic Courtès) writes:
>
> (in particular, `ispunct ()' does not match `char-set:punctuation').

Oh, it's a bit bigger.

> but `punctuation', for instance, is a superset of what
> SRFI-14 expects while `symbol' is (correspondingly) a subset of what it
> should be,

Does the srfi specified relation to graphic still hold?  Ie.

        graphic = letter + digit + punctuation + symbol

> -#define SCM_CHARSET_SET(cs, idx) \
> -  (((long *) SCM_SMOB_DATA (cs))[(idx) / SCM_BITS_PER_LONG] |= \
> +#define SCM_CHARSET_SET(cs, idx)                             \
> +  (((long *) SCM_SMOB_DATA (cs))[(idx) / SCM_BITS_PER_LONG] |=       \
>      (1L << ((idx) % SCM_BITS_PER_LONG)))

Is that a change?

> +      if (pred (c))                          \
> +     SCM_CHARSET_SET ((cset), (c));          \
> +      else                                   \
> +     SCM_CHARSET_UNSET ((cset), (c));        \

It may be possible to do a "set to a value" rather than separate
set/unset macros.

> -(use-modules (srfi srfi-14))
> +(use-modules (srfi srfi-14)
> +             (srfi srfi-1) ;; `every'
> +             (test-suite lib))

A "define-module" there can prevent srfi-1 leaking out to subsequent
tests.

> +(define (find-latin1-locale)
> +  ;; Try to find and install an ISO-8859-1 locale.  Return `#f' on failure.
> +  (if (defined? 'setlocale)
> +      (let loop ((locales (map (lambda (lang)
> +                              (string-append lang ".iso88591"))
> +                            '("de_DE" "en_GB" "en_US" "es_ES"
> +                              "fr_FR" "it_IT"))))

The posix "locale -a" program can print all available locales, if you
wanted to ask nl_langinfo(CODESET) or "locale -k charmap" what the
charset is for each of them, or just try the undotted ones with
8859-1, or whatever.