[slime-devel] readtable-case :invert

[Helmut Eller]

"Thomas F. Burdick" <tfb at OCF.Berkeley.EDU> writes:

> I've been using SBCL with a copy of SLIME that I checked out on March 4.
> I was quite happy to see that when I evaluated
> 
>    (setf (readtable-case *readtable*) :invert)
> 
> SLIME didn't die.  However, there are some weird bits, especially
> symbol completion, which seems to work for oneword symbols, but not
> multi-word-symbols.  I've slowly come to the realization that I'm one
> of only a few Lispers who regularly works with case-inverted
> readtables, so I'm guessing not much thought has been put into this
> question.  For you all who know the insides of SLIME, I'm wondering if
> it would be much work to get SLIME to be inverted-case freindly, or if
> a couple well-placed WITH-STANDARD-IO-SYNTAX's would do it.

No, we didn't think much about this issue.  The problems with symbol
completion are probably not so much related to the reader/printer
settings, but to some inconsistent use of char-equal/char=/eql in our
code.

It's not entirely obvious to me what inverted-case friendly completion
means, so here some questions.  Assume we have the symbols foo, Foo,
FOO and FLOAT.  What are the preferred expansions for f?
 
 f => foo
 f => Foo, FOO, FLOAT
 f => Foo, foo, float
 f => FOO, Foo, foo, float