How to use base64-encode-region with MULE?

How to use base64-encode-region with MULE?

[Kai.Grossjohann]

Currently, I use write-region to save stuff to a tmp file, then
execute mimencode on the tmp file.  I assume that write-region does
some MULE magic if the region contains non-ASCII things.

I would like to do away with the tmp file and call
base64-encode-region instead.  But how to deal with the MULE magic in
this case?

As base64.el is part of Gnus, I thought this would be the right place
to ask...

tia,
kai
-- 
Abort this operation?   [Abort]  [Cancel]

Re: How to use base64-encode-region with MULE?

[Lars Magne Ingebrigtsen]

Kai.Grossjohann@CS.Uni-Dortmund.DE writes:

> I would like to do away with the tmp file and call
> base64-encode-region instead.  But how to deal with the MULE magic in
> this case?

You have to de-Mulify the buffer before calling base64-encode-region.
And, in Emacs 20.3 you have to be in a unibyte buffer.

So you basically would do something like

(encode-coding-region (point-min) (point-max) 'iso-8859-1) ; or something
(setq r (buffer-string))
(mm-with-unibyte-buffer
  (insert t)
  (base64-encode-region (point-min) (point-max))
  (setq base64 (buffer-string)))

Presto!  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: How to use base64-encode-region with MULE?

[Hrvoje Niksic]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> And, in Emacs 20.3 you have to be in a unibyte buffer.

Why?  Isn't it sufficient that all the characters are within [0,255]
range?

Re: How to use base64-encode-region with MULE?

[Lars Magne Ingebrigtsen]

Hrvoje Niksic <hniksic@srce.hr> writes:

> > And, in Emacs 20.3 you have to be in a unibyte buffer.
> 
> Why?  Isn't it sufficient that all the characters are within [0,255]
> range?

No -- é, for instance, is encoded internally as \201\040 (or
something).  Well, after you've encoded, it's no longer an é, but
there's still no actual character > 0x80 in the buffer.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: How to use base64-encode-region with MULE?

[Hrvoje Niksic]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Hrvoje Niksic <hniksic@srce.hr> writes:
> 
> > > And, in Emacs 20.3 you have to be in a unibyte buffer.
> > 
> > Why?  Isn't it sufficient that all the characters are within [0,255]
> > range?
> 
> No -- é, for instance, is encoded internally as \201\040 (or
> something).

Well yes, but shouldn't the Base64 code proceed character by character
and read the "decoded" characters from the buffer?  That's what the
code in XEmacs does, and to me it seems like the only sensible thing
to do.  Otherwise, you get the \201 junk in otherwise perfectly fine
buffers.

Re: How to use base64-encode-region with MULE?

[Lars Magne Ingebrigtsen]

Hrvoje Niksic <hniksic@srce.hr> writes:

> Well yes, but shouldn't the Base64 code proceed character by character
> and read the "decoded" characters from the buffer?  That's what the
> code in XEmacs does, and to me it seems like the only sensible thing
> to do.  Otherwise, you get the \201 junk in otherwise perfectly fine
> buffers.

Yup.  The Emacs implementation encodes the internal representation.
One the one hand, that may seem totally weird.  On the other hand,
Emacs has both multibyte and unibyte buffers, so this isn't all that
catastrophic for Emacs, while doing the same in XEmacs would be, er,
catastrophic.  

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen