Re: reader config (was 199.22)

[Matthew Flatt <mflatt@cs.utah.edu>]

Meant to reply to this long ago...

Quoting MichaelL@frogware.com:
> 
> > >  * EXPERIMENTAL: Added `#cs' and `#ci' prefixes for controlling
> > >    case sentivity during `read'. The sequence `#cs' can be used as a
> > >    prefix for any expression to make symbols within the expression
> > >    case-sensitive.  The `#ci' prefix similarly makes symbols
> > >    case-insensitive. For example,
> 
> Just a thought... is it worth taking the opportunity to provide a
> generalized mechanism for
> paramterizing the reader--rather along the lines of XML processing
> instructions? One idea
> might be have a mechanism for setting values (eg, #<case-sensitive=#t>)
> another might be
> to actually evaluate code during reading (eg, #<read-case-sensitive! #t>).
> Again, just a thought...

Yes, I thought about that quite a lot, mostly along the lines of
#<case-sensitive=#t>. (Evaluating arbitrary code at read time seems to
open doors that are too difficult to close.) I never came to a
conclusion on whether a general mechanism was worthwhile, though.

In any case, it seemed like a general mechanism would be too verbose
for the common case of turning on case sensitivity, and we'd want a
shorthand. So I went ahead with the shorthand for now.

Matthew