I switched to the cygwin build of emacs and everything is working as
expected so thank you for your help. I usually do my work on a Linux box,
but they gave me a Windows machine at work recently. It's been such a pain
to work with sometimes.

Thanks again,

Andrzej

PS. And yes CMD is what I meant by "regular windows shell" :)

On Tue, Sep 14, 2010 at 9:21 PM, Eli Zaretskii <eliz@gnu.org> wrote:

> > Date: Tue, 14 Sep 2010 13:02:30 +0200
> > From: Andrzej Skiba <andskiba@gmail.com>
> >
> > (defun as/grep-project (project pattern)
> >   (interactive "sProject: \nsPattern: ")
> >     (grep-find (concat
> >                    "/usr/bin/find /cygdrive/c/projects/"
> >                    project
> >                    " -type f "
> >                    " -not -name \"*.svn-base\" "
> >                    "-and -not -name \"*.tmp\" "
> >                    "-and -not -name \"*.log\" -print0 "
> >                    "| xargs -0 -e grep -U -n -s -F \""
> >                    pattern
> >                    "\"")))
> >
> > All works great until I try to search for a word with Polish letters
> (such
> > as ą, ś, ć, ł, ń etc.). The files are all utf-8. When I run the command
> > searching for string "Usuń" in project test I get the following output in
> > the grep buffer:
> >
> > /usr/bin/find /cygdrive/c/projects/test -type f -not -name "*.svn-base"
> -and
> > -not -name "*.tmp" -and -not -name "*.log" -print0 | xargs -0 -e grep -U
> -n
> > -s -F "Usuń"
> > /usr/bin/bash: /usr/bin/find /cygdrive/c/projects/test -type f -not -name
> > "*.svn-base" -and -not -name "*.tmp" -and -not -name "*.log" -print0 |
> xargs
> > -0 -e grep -U -n -s -F "Usuń": No such file or directory
> >
> > It runs fine with any input without Polish characters.
>
> You seem to be using the native build of Emacs in conjunction with
> Cygwin tools (Grep, Bash, etc.).  If so, this is asking for trouble,
> because there are subtle incompatibilities between Cygwin programs and
> native Windows programs.  I/O encoding is one of these areas: whereas
> latest Cygwin versions use UTF-8, native Windows programs use the
> current Windows codepage.  The native build of Emacs cannot encode
> command lines it passes to programs in anything but the current
> codepage, which is no good for you if your files are encoded in UTF-8.
>
> I suggest to use the Cygwin build of Emacs instead.
>
> > The find command works fine in a regular window shell as well as cygwin
> > bash.
>
> What is the "regular window shell"?  If it's CMD, then I don't see how
> it could work, since CMD does not support UTF-8 keyboard input.
> Perhaps the Cygwin port of Grep transparently converts keyboard input
> into UTF-8 or something.
>