Re: [emacs-bidi] mirroring of glyphs

[Eli Zaretskii]

On Sun, 18 Nov 2001, Alex Schroeder wrote:

> I have actually resigned myself to the inevitable.  It is not
> possible.

I don't think so.

> I think Ehud already mentioned this in a previous post.

Not if I understood what Ehud said.

> Anyway, here are some examples to illustrate this.  There are certain
> logical representations which yield the same visual representation:
> 
> (string= (bidi-logical-to-visual "my CAR123")
>        (bidi-logical-to-visual "my 123CAR"))
>   => t

True.

> Therefore the reverse operation must only return one of the many
> possible logical representations in order to be correct:
> 
> (bidi-visual-to-logical
>  (bidi-logical-to-visual "my CAR123") 'L2R)
>   => "my 123CAR"

True.  But where's the problem?

> The same is true for the one remaining problem test case from the PGBA
> page:
> 
> (bidi-visual-to-logical "<123H/>shalom<123H>" 'R2L)
>   => "<Hshalom<123</H123>"
> 
> What I didn't appreciate is that this is just one of the many possible
> logical representations for the same string:
> 
> (bidi-logical-to-visual (bidi-visual-to-logical "<123H/>shalom<123H>" 'R2L))
>   => "<123H/>shalom<123H>"
> 
> Therefore the current bidi implementation is as good as it gets.  :)

Sorry, no.  The output should be one of the possible inputs, but we
should try to make it reasonable.  "<Hshalom<123</H123>" isn't.
Reversing the input before the conversion would make in reasonable, as
your other mail demonstrates.

Note that "my 123CAR" as the result of the example above _is_
reasonable.