I need to use UTF8... sometimes

I need to use UTF8... sometimes

[Gary]

Okay, here is the situation. My local environment uses ISO-8559-1
("latin-1") -
,----
| $ env | grep -iE "(LANG|LC_)"
| LC_MESSAGES=en_US.ISO-8859-1
| LANG=de_CH.ISO-8859-1
| LC_TIME=en_GB.ISO-8859-1
`----

I now need to be able to transfer files *inside emacs* to a remote
server. I intended to use the following simple function to do that:
,----
| (defun foo-upload-buffer ()
|   "Uploads the current (foo) buffer."
|   (interactive)
| 
|   (eshell-command
|    (concat foo-scp buffer-file-name " "
|            foo-server-user "@" foo-server ":" foo-server-path))
| )
`----

and then found out that the server ("foo-server", above) expects the
files to be in UTF-8 :(

I don't really want to change my use of ISO-8559-1 for various reasons,
not least of all being that I deal with lots of other systems that
expect files in that encoding. Nor do I really want to use UTF-8 locally
for these particular files, because it means I can't grep them
properly.

So what I am wondering is what the easiest way is to do the transfer
using UTF-8 but still edit locally in ISO-8559-1. I wondered if
foo-upload-buffer should write the buffer to a temporary file, convert
it to UTF-8, and then transfer that, but it seems a bit longwinded. Any
other ideas?

-- 
Gary        Please do NOT send me 'courtesy' replies off-list.
GNU Emacs 23.2.1
emacsclient 23.2

Re: I need to use UTF8... sometimes

[Eli Zaretskii]

> From: Gary <help-gnu-emacs@garydjones.name>
> Date: Wed, 06 Oct 2010 13:35:22 +0200
> Mail-Copies-To: never
> 
> So what I am wondering is what the easiest way is to do the transfer
> using UTF-8 but still edit locally in ISO-8559-1. I wondered if
> foo-upload-buffer should write the buffer to a temporary file, convert
> it to UTF-8, and then transfer that, but it seems a bit longwinded. Any
> other ideas?

Bind coding-system-for-write to 'utf-8 before invoking the Eshell
command.  That should do the trick.