Re: Implementing encoded-character

[Kjetil Torgrim Homme <kjetilho@xxxxxxxxxx>]

On Wed, 2007-04-04 at 10:11 +0200, Michael Haardt wrote:
> Exim will implement encoded-character RSN,

excellent!

> but I wonder if I understood things correctly:
> 
> "$${hex:40}" -> "$@"

yes.

> A dollar does not have to be quoted, does it?

right, only for very unlikely sequences.

> "${hex:40" -> "${hex:40"
> "${hex: 40 }" -> "${hex: 40 }"

yes.  whitespace is only allowed between hex-pairs.  btw, how do you
feel about allowing CRLF as well as SPC and TAB between hex-pairs?

> "${hex:400}" -> "${hex:400}"

yes, hex-pair can't be three digits.

> "${unicode:40}" -> "${unicode:40}"

no, this is "@".

> "${unicode:1000000}" -> "${unicode:1000000}"

yes.

> There is no word of the encoded-character grammar inside the string,
> taking everything literal.

I don't understand this statement.

> "${hex:40${hex:40}}" -> "${hex:40$}

no, "${hex:40@}"

> It looks nested, but actually it's just some junk around a word of the
> grammar.
> 
> "${unicode:020000}" -> error
> 
> Unicode range violation.

no, U+20000 is inside the Unicode range.  ${unicode:0020000} fails due
to not matching unicode-hex (too many digits), ${unicode:200000} fails
due to being outside the Unicode range.

-- 
Kjetil T.