[clisp-list] Re: New, from-scratch implementation of backquote.

[Hoehle, Joerg-C. <Joe...@t-...>]

Hi,

please help me nuderstand some area of backquotes.


Kaz Kylheku wrote:
>Which reminds me that I should add some optimization for reducing
>a constant list to a constant vector; we should be spitting out
>the literal object:
>  #(1 2 (2 3) 3)
Yes please:
`(foo #(1 2 3) ,bar)
should not call (vector 1 2 3) at run-time.
Or at least I believe so.
(list 'foo '#(1 2 3) bar) is the code I'd expect.

[...]
>   (list 'system::backquote (vector 1 
>                                    (list 'system::unquote 'a) 
>                                    2))
>   ==> `#(1 ,A 2)
>This is a feature: we want the backquote notation to have a
>list-based target language such that we can write code which writes
>that language.

This is great! Actually, I believe I posted several years ago to comp.lang.lisp on the topic of standardizing backquote internals (like in Scheme). I now believe that this not possible in CL (your e-mails here helped). But it's fine if it works in CLISP.

Regards,
	Jorg Hohle.