[racket] drracket and arity checks

From: Robby Findler (robby at eecs.northwestern.edu)
Date: Fri Oct 18 15:44:44 EDT 2013

Could it be that errortrace is interfering with the optimizers ability to
spot these errors? That is, if you turn on the logging window and turn off
debugging (in the language dialog), do you then see them? Also, there are
some checkboxes nearby there that also can affect the optimizer's ability
(also for the purposes of trading off speed for better error messages when
runtime errors do occur).

Robby


On Fri, Oct 18, 2013 at 2:30 PM, Erik Pearson <erik at adaptations.com> wrote:

> Hi, I saw the recent question regarding arity mismatch reporting, but have
> a more pointed question for drracket usage.
>
> I can get a warning message from racket (command line tool) with the "-W
> warning" option, and from drracket by opening the log view and filtering by
> warning. However, in drracket I need to do something like create an
> executable in order to generate the warning, a simple press of the Run
> button doesn't do it. It appears that a simple "load" in cli "racket -W
> warning" does invoke the optimizer, so I guess that is not what drracket
> does (or does in a different way.)
>
> So, is a way to get drracket to invoke the optimizer (or do a full load of
> the file) or do whatever magic needs to be done when the Run button is
> pressed to generate the warnings? I could use emacs, but drracket is such a
> low-stress, enjoyable mode of working, especially on smallish files.
>
> Before delivering apps I do a raco exe, raco distribute so I can get this
> done somewhere in the workflow, it would just be nice to get those pesky
> little arity things squashed early in the development cycle. I'm using a
> lot of json data extraction and manipulation now, and like the *-ref
> functions the functions I'm using have a default value in case the hash key
> or list position is not found. It is easy to leave off the default value
> and not have it crop up until the app is run. Ouch! My functions require
> the default value argument, just so I can ensure that something is done to
> handle missing data, and thus the arity check is handy to ensure that this
> argument is indeed coded.
>
> Yes, and I'm learning TR, and am starting to work on TR adapters for
> racket libs and my own, which will help. But for more ad-hoc apps which may
> never be TR, the drracket assistance is welcome.
>
> Thanks,
> Erik.
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20131018/9c17641e/attachment.html>

Posted on the users mailing list.