Mark H Weaver <m...@netris.org> skribis:
> l...@gnu.org (Ludovic Courtès) writes:
[...]
>> BTW, why does ‘set-field’ has the record as its 2nd argument instead of
>> 1st (unlike ‘set-fields’)?
>
> Good question. I followed the syntax of 'set-field' from your original
> patch, but that argument order did not make sense for 'set-fields'.
In the meantime we concurred on IRC that keeping the record as the first
argument in both cases may be best.
>> +The @code{(srfi srfi-9 gnu)} module extends SRFI-9 with facilities to
>> +return new record instances based on existing ones, only with one or
>> +more field values changed---@dfn{functional setters}. First, the
>> +@code{define-immutable-record-type} works like
>> +@code{define-record-type}, except that setters are defined as functional
>> +setters.
>
> "except that the fields are immutable and the setters are ..."
OK.
>> +@deffn {Scheme Syntax} set-field (field sub-fields ...) record value
>> +Return a new record of @var{record}'s type whose fields are equal to
>> +the corresponding fields of @var{record} except for the one specified by
>> +@var{field}.
>> +
>> +@var{field} must be the name of the getter corresponding to the field of
>> +@var{record} being ``set''. Subsequent @var{sub-fields} must be record
>
> This is the first time that "getter" is used, but it has not been made
> clear that you mean what has been called an "accessor" elsewhere in the
> doc. More generally, there is a confusing mixture of the
> accessor/modifier and getter/setter terminology. I wonder if it would
> made sense to do some kind of find/replace in this section.
Yeah. I ended up leaving “getter”, because that’s the term used in the
SRFI-9 node from the beginning. But I agree we might need to do some
find/replace at some point.
Ludo’.
