If we had a script to check for this, it could be set up as part of the
default Travis thing generated for packages.

There could be a package/tool to generate helpful error messages when you
try to used a function/MathConstant by the wrong (confusable) unicode
character. Something like "You're using the wrong pi; try this one or the
name `pi`".

-- Leah


On Fri, Jan 17, 2014 at 11:32 AM, Milan Bouchet-Valat <[email protected]>wrote:

> Le vendredi 17 janvier 2014 à 12:07 -0500, Jiahao Chen a écrit :
> > > It seems like a huge problem to me if, when I read your code, I
> > can’t tell which characters you’re using just by reading your code.
> >
> > Seems ridiculous, doesn't it? Welcome to the world of non-ASCII
> > character sets ;-)
> >
> > I think we can agree that it would be nice to have a better way of
> > detecting Unicode homograph collisions than manual verification. We
> > could try to standardize on various codepoints. But I think we'll just
> > have to agree to disagree on making this a prescription.
> If a tool is written to automatically format Julia code, it could also
> check that two homograph characters are not used at the same time in the
> same project. This would catch all problematic cases, and I don't think
> there would be many false positives.
>
> That wouldn't catch the case were you do not call the tool at all before
> running the code -- but doing the check in the compiler itself may slow
> down compilation.
>
>
> Regards
>

Reply via email to