replacing ^M with emacs

[Giorgos Keramidas]

On 2006-10-30 10:03, Giorgos Keramidas <keramida at ceid.upatras.gr> wrote:
> On 2006-10-28 04:18, Tsampros Leonidas <ltsampros at upnet.gr> wrote:
> > I think there is something similar in emacs by using the
> > set-buffer-file-coding-system (binded at C-x RET f in default
> > configurations).
> > 
> > So to "cure" and succesfully "convert" DOS files into unix format, i
> > use C-x RET f unix RET.
> 
> I'm not sure `set-buffer-file-coding-system' will have any effect on an
> already opened file though.  I just tried this with a file which was
> created outside Emacs, and contained:
> 
>     $ cat -vte foo 
>     fooo^M$
>     $
> 
> Opening this file with `C-x C-f foo RET' and setting the buffer file
> coding system with `C-x RET f unix RET', marks the buffer as modified,
> but saving the file does not modify the contents of the file to use UNIX
> newlines only.
> 
> If you really want to use Emacs for the conversion, you have to
> *explicitly* replace ^M characters, either with `M-x replace-string RET
> C-q C-m RET RET' or some either way.

Oops...

Apparently, I have `inhibit-eol-conversion' modified locally.  This is
what makes Emacs avoid EOL conversion when `set-buffer-file-coding-system'
is called.

Sorry for the confusion.