Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

[Mark H Weaver]

Hi Mike,

Thanks for the Unicode 6.1 update!  Now, however:

  FAIL: srfi-14.test: Latin-1 (8-bit charset): char-set:symbol

Would you be willing to investigate?

  Many thanks!
     Mark


"Mike Gran" <spk121-/E1597aS9LQAvxtiuMwx3w@public.gmane.org> writes:
> commit bf8d845468fa71debf45e91a4e40f4b219dab4b0
> Author: Mike Gran <spk121-/E1597aS9LQAvxtiuMwx3w@public.gmane.org>
> Date:   Wed Feb 1 19:51:35 2012 -0800
>
>     Update srfi-14 character sets to Unicode 6.1
>     
>     * libguile/srfi-14.i.c (cs_lower_case_ranges, cs_lower_case)
>       (cs_upper_case_ranges, cs_upper_case, cs_letter_ranges, cs_letter)
>       (cs_digit_ranges, cs_digit, cs_letter_plus_digit_ranges, cs_letter_plus_digit)
>       (cs_graphic_ranges, cs_graphic, cs_printing_ranges, cs_printing)
>       (cs_punctuation_ranges, cs_punctuation, cs_symbol_ranges, cs_symbol)
>       (cs_designated_ranges, cs_designated): modified
>     * doc/ref/api-data.texi: modified

Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

[Mike Gran]

Hi Mark-
 
> Thanks for the Unicode 6.1 update!  Now, however:
> 
>   FAIL: srfi-14.test: Latin-1 (8-bit charset): char-set:symbol
> 
> Would you be willing to investigate?

Strange.  I'll check it out today.

-Mike

Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

[Mike Gran]

> Hi Mike,
> 
> Thanks for the Unicode 6.1 update!  Now, however:
> 
>   FAIL: srfi-14.test: Latin-1 (8-bit charset): char-set:symbol
> 
> Would you be willing to investigate?

Looks like Unicode 6.1 has recategorized some of the symbols, including
a few in Latin-1.
 
"§" U+00A7 SECTION SIGN from Symbol_Other to Punctuation_Other
"¶" U+00B6 PILCROW SIGN from Symbol_Other to Punctuation_Other
 
It seems that the correct response would be just to change
the Latin-1 test cases.
 
A wrinkle, though, is that in SRFI-14, they call out "§" and "¶"
as symbols.  But my interpretation of the text in SRFI-14 is that
they intended to follow Unicode's categorization.
 
http://srfi.schemers.org/srfi-14/srfi-14.html
 
WDYT?
 
Thanks,
 
Mike

Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

[Mark H Weaver]

Mike Gran <spk121@yahoo.com> writes:

>> Thanks for the Unicode 6.1 update!  Now, however:
>> 
>>   FAIL: srfi-14.test: Latin-1 (8-bit charset): char-set:symbol
>> 
>> Would you be willing to investigate?
>
> Looks like Unicode 6.1 has recategorized some of the symbols, including
> a few in Latin-1.
>  
> "§" U+00A7 SECTION SIGN from Symbol_Other to Punctuation_Other
> "¶" U+00B6 PILCROW SIGN from Symbol_Other to Punctuation_Other
>  
> It seems that the correct response would be just to change
> the Latin-1 test cases.
>  
> A wrinkle, though, is that in SRFI-14, they call out "§" and "¶"
> as symbols.  But my interpretation of the text in SRFI-14 is that
> they intended to follow Unicode's categorization.

Agreed.

> http://srfi.schemers.org/srfi-14/srfi-14.html

SRFI-14 states:

  char-set:symbol

  In Unicode, a symbol is any character that has one of the symbol
  categories in the Unicode character database (Sm, Sc, Sk, or So).

and I think that this is intended to be the normative definition.
SRFI-14 then proceeds to list the symbols of ASCII and Latin-1, but I
interpret that as non-normative, to save the reader the trouble of
consulting Unicode.  IMHO, anyway.

   Thanks!
     Mark

Re: GNU Guile branch, stable-2.0, updated. v2.0.5-5-gbf8d845

[David Kastrup]

Mike Gran <spk121@yahoo.com> writes:

>> Thanks for the Unicode 6.1 update!  Now, however:
>> 
>>   FAIL: srfi-14.test: Latin-1 (8-bit charset): char-set:symbol
>> 
>> Would you be willing to investigate?
>
> Looks like Unicode 6.1 has recategorized some of the symbols, including
> a few in Latin-1.
>  
> "§" U+00A7 SECTION SIGN from Symbol_Other to Punctuation_Other

Utter nonsense.

> "¶" U+00B6 PILCROW SIGN from Symbol_Other to Punctuation_Other

Makes sense.

> It seems that the correct response would be just to change
> the Latin-1 test cases.

Differing from the standard would be asking for even more trouble.  I
have no idea what sense declaring § as punctuation makes (it is not used
in that manner), but at least for ¶, there is some motivation.

-- 
David Kastrup