[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Re: UTF-8 [was Re: LuaSocket http ftp smpt...]
- From: Klaus Ripke <paul-lua@...>
- Date: Sun, 6 Feb 2005 13:27:42 +0100
On Sunday 06 February 2005 05:13, Diego Nehab wrote:
> > So, yes, I'll also add Tcl-style charset conversion to/from UTF-8.
> I don't think this should be up to a MIME library.
d'accord -- the actual charset conversion must in no way be
"required" by a MIME package, just like it need not include
a "handler" for application/msword.
What IS MIME's business, however, is to (be able to)
strip the MIME header encoding in two parts:
here are the bytes and this is the name of the charset.
Then the app may see whether the claimed charset is known
at all and whether it's interested in doing some recoding
(to whatever other charset).
> Perhaps someone can bind librecode...
Again I'd go for the Tcl crap since it's *MUCH* smaller.
> The support for headers with non-ASCII characters
> doesn't have to know what is the charset.
Just to the extend it needs to "know" what any Content-Type is
(and especially a charset attribute within a content type):
locate and return that name.
> It can simply applie Base64 or
> Quoted-Printable to binary data.
and return it silently assuming it was ISO-8859-1?
You guys in the Americas don't see other charsets all
too often, do you? Maybe we should donate a few Euro,
so you convert to 8859-15 like many folks over here ;)