[Prev][Next][Index][Thread]
Re: Q: statements in module file?
Roland Paterson-Jones wrote in message
<9hfqc2$6seu$1@sp15at20.hursley.ibm.com>...
>
>"Andreas Bogk" <andreas@andreas.org> wrote in message
>87pubrno6k.fsf@teonanacatl.andreas.org">news:87pubrno6k.fsf@teonanacatl.andreas.org...
>> Hugh Greene <q@tardis.ed.ac.maps.uk> writes:
>>
>> > just "encouraged". I have a vague recollection of *very* occasionally
>> > seeing some "define class" calls which were sufficiently dynamic that
>they
>> > had to be moved before their first point of reference, for FunDev to
>> > accept them.
>>
>> d2c just bails out on them, and says it doesn't understand "hairy
>> classes".
>
>Do you have some code examples that would help my understanding, both of
>'hairy classes', and bad and good macro definition ordering?
>
This would probably do it, tho' I didn't try:
define variable shoot-me-now = make(<class-maker>);
define class <in-the-foot-please> (shoot-me-now) end;
If the exact object that the call to make returns can't be determined
at compile-time, then <in-the-foot> is completely dynamic. It seems
not unreasonable to me for some compilers to bail out on this.
References: