Re: [PATCH v1 02/36] Guile extension language: doc additions

[ludo@gnu.org (Ludovic Courtès)]

Doug Evans <xdje42@gmail.com> skribis:

> On Sat, Jan 18, 2014 at 12:42 PM, Ludovic Courtès <ludo@gnu.org> wrote:
>> Doug Evans <xdje42@gmail.com> skribis:
>>
>>> On Fri, Jan 3, 2014 at 1:30 PM, Ludovic Courtès <ludo@gnu.org> wrote:
>>>>>> +The optional @var{errors} argument is either @code{"strict"}
>>>>>> +or @code{"replace"}.  A value of @code{"strict"} corresponds to
>>>>>> +Guile's @code{SCM_FAILED_CONVERSION_ERROR} and a value of @code{"replace"}
>>>>>> +corresponds to Guile's @code{SCM_FAILED_CONVERSION_QUESTION_MARK}.
>>>>>
>>>>> Suggest a cross-reference to Guile documentation here.
>>>>
>>>> Agreed.  Also, Guile talks of “conversion strategy” and “conversion
>>>> error handler”, with values ‘error’, ‘substitute’, and ‘escape’ (at the
>>>> Scheme level), and I’d recommend sticking to those names and terminology.
>>>
>>> The values chosen were to be consistent with the python support.
>>> OTOH I *do* like being more consistent with the particular extension
>>> language at hand.
>>> I've tentatively changes things to use "error" and "substitute".
>>> Question: How about exporting the SCM_FAILED_CONVERSION_* constants
>>> and using those instead?
>>>
>>> E.g, (value->string foo #:errors SCM_FAILED_CONVERSION_ERROR) ?
>>
>> I’d rather use a symbol:
>
> Yeah, I thought of that, but the encoding is a string,
> so it'd be "#:encoding string #:errors symbol".

Right.  Looks good to me.

> I don't have a strong preference, but using a symbol here while
> feeling Schemey feels too weird.
> It's not a strong preference though.

Using a symbol for #:errors?  It would be natural and consistent with
the rest of Guile’s API (notably and ‘string->bytevector’,
‘set-port-conversion-strategy!’.)

>>   (value->string foo #:conversion-strategy 'error)
>>
>> So that has to be converted in C but I think that’s OK.
>
> #:conversion-strategy is more to type than #:errors but I'm happy to
> change it if you want.
> Though this is a case where I would not want to support both
> #:conversion-strategy and #:errors so whatever we pick is it.

Right.  Or ‘value->string’ could have this signature:

  value->string VAL [ENCODING [ERRORS]]

The precedent being ‘string->pointer’ and ‘string->bytevector’.

> We can create a 'error symbol at start up and just use scm_eq so the
> comparison is easy enough.

Exactly, that’s what I would suggest.

Thanks,
Ludo’.