[plt-scheme] DrScheme online syntax highlighting

Excellent; thanks!

On Mon, 3 Nov 2003, Robby Findler wrote:
> You may be interested in the PLTNOTOOLS and PLTONLYTOOLS environment 
> variables in the drscheme manual.
> 
> On Nov 3, 2003, at 2:31 PM, Daniel Silva wrote:
> > On Mon, 2003-11-03 at 15:05, John Clements wrote:
> >> On Monday, November 3, 2003, at 02:59  PM, Daniel Silva wrote:
> >>> On Mon, 2003-11-03 at 14:42, John Clements wrote:
> >>>> On Monday, November 3, 2003, at 02:27  PM, Daniel Silva wrote:
> >>>>> Does anybody who tried highlight.ss get errors like these when
> >>>>> loading
> >>>>> scheme files?
> >>>>>
> >>>>> open-input-text-editor: end index outside the range [1997,3000]: 
> >>>>> 1024
> >>>>> open-input-text-editor: end index outside the range [2596,4000]: 
> >>>>> 2011
> >>>>> #<struct:object:/home/daniel/plt/collects/stepper/stepper-
> >>>>> tool.ss:525:8>:1:0: read: expected a ')'
> >>>>
> >>>> No idea, but the "#<struct....stepper-tool.ss:525:8>" is actually
> >>>> DrScheme's definitions-text%.  This is because the stepper tool
> >>>> applies
> >>>> a mixin to this text% to override several text% methods.
> >>>>
> >>>
> >>> It looks like stepper-definitions-text-mixin overrides on-insert and
> >>> on-delete.  I override after-insert and after-delete to change the
> >>> nearest sexp's style.  I didn't think that change-style would trigger
> >>> on-insert or on-delete.
> >>>
> >>> Do drscheme:eval:traverse-program/multiple, syntax-case, expand, or
> >>> drscheme:language:make-text/pos invoke on-insert or on-delete?  I'm
> >>> using those to read sexps, and the combination of get-backward-sexp 
> >>> and
> >>> get-forward-sexp to find the last sexp typed.
> >>
> >> No, no, let me clarify.  The error that you're getting in all
> >> likelihood has
> >> nothing whatsoever to do with the stepper.  Since the stepper is the
> >> last
> >> one to mix something into drscheme's definitions-text, the object that
> >> is
> >> the result of this composition (the text%) is given a source position 
> >> in
> >> the stepper.  Any error message in DrScheme that involves DrScheme's
> >> definitions-text% will refer to it as a struct that comes from the
> >> stepper.
> >>
> >> Put another way:  rm -rf the stepper collection.  I claim that (with
> >> high likelihood) the bug will still be present, and the error message
> >> will
> >> look the same, except that the source position given will come from 
> >> some
> >> other tool.
> >>
> >
> > Oh, you're right -- it came from syntax-color before and when I removed
> > the dir, the error came from the steppper.
> >
> > Do mixins make it impossible to debug classes then?  I can't remove the
> > stepper because DrScheme requires it.  The error message points to the
> > beginning of the mixin class, not a method, and a with-handlers guard
> > around my code isn't catching the error.  Can I catch errors with some
> > other way? Is there a with-all-inclusive,-even-internal-errors-handlers
> > form?
> >
> > Daniel
> >
> >
>