How to find "bad" characters that cannot be saved using some encoding

[Steven Tolkin]

Sometimes emacs tells me it cannot save a buffer because there
are characters that cannot use the current coding system. 
How can I have it show me the offending character or characters?  

Suppose I chose raw-text and save the buffer.
Then later when I read the file, it seems to have all ordinary characters,
and after a trivial change such as adding a blank it now can be saved.
Was some "bad" character replaced with a different character?
I am using emacs version 21.2.1 on Windows XP.

According to this old message fixing this was on some TODO list.
Was this ever done?
<quote>
From: Eli Zaretskii 
Subject: Re: what file was it that caused a coding system warning 
Newsgroups: gnu.emacs.bug
Date: 2002-01-17 03:33:12 PST 

> Of course it could
> also zero in on the problem spot in the file that caused the warning,
> so we would know what to fix.

Alas, it cannot do that, currently: it simply doesn't know where's that 
spot.  A feature to add the (non-trivial) code that will find that spot 
is on our TODO list.
</quote>

Thanks,
Steve
-- 
Steven Tolkin    Steve . Tolkin at Fmr . Com   617-563-0516 
Fidelity Investments   82 Devonshire St. V13C    Boston MA 02109
There is nothing so practical as a good theory.  Comments are by me, 
not Fidelity Investments, its subsidiaries or affiliates.