axiom-developer
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Axiom-developer] postprop.lisp


From: Gabriel Dos Reis
Subject: Re: [Axiom-developer] postprop.lisp
Date: 26 Oct 2006 05:05:58 +0200

root <address@hidden> writes:

| > > | > 
| > > | > Tim --
| > > | > 
| > > | >   The source file postprop.lisp is compiled into depsys, but loaded
| > > | > in interpreted form in AXIOMsys.  What is the reason for that?
| > > | > 
| > > | 
| > > | AFAIK postprop.lisp is unused -- IMHO it should be removed.
| > > 
| > > It is compiled into depsys -- which is used to compile "old" Boot.
| > > >From what I can see from the source file, it is used to map tokens to
| > > functions that parse corresponding grammar production (i.e. to support
| > > the zipper parser).
| > > 
| > 
| > Well, beauty of Lisp: the same things are defined in 'property.lisp'
| > (except for |special| property, which is unused). You are probably
| > right that depsys picks definitions from 'postprop.lisp'.  In my
| > experiments (IIRC using AXIOMsys) changing 'postprop.lisp' did not
| > work (had no effect), but changing 'property.lisp' worked.  Also data
| > in 'property.lisp' is slightly different, but for me it looks better
| > than what is in 'postprop.lisp'.  And bootstrap with 'postprop.lisp' 
| > removed worked fine.
| 
| you're both asking questions i can't answer off the top of my head.
| there is certain to be a lot of cruft in axiom.
| 
| however, i take the most conservative approach and PROVE that
| each function removed CANNOT every be called. be aware that
| the compiler sometimes dynamically constructs function names.

If I understand Waldek correctly, "unused" is probably the wrong
term.  Rather, "redundant" or "duplicate" would be appropriate.
Essentially, he is saying that property.lisp has the same definitions
and take over postprop.lisp.  I hve not run the "mental" diff yet.

-- Gaby




reply via email to

[Prev in Thread] Current Thread [Next in Thread]