unencodable-char-position bug?

unencodable-char-position bug?

[Paul Pogonyshev]

Hi,

Is it correct that

	(unencodable-char-position 0 1 'utf-8 nil "ы")

evaluates to 0?  Since UTF-8 can encode anything, I presume the
correct return value would be nil...  The result is the same if
tried on a Russian letter character in a buffer instead of a
string.

In general, what is the best way to determine if given character
is encodable in given coding system?

Paul

Re: unencodable-char-position bug?

[Kenichi Handa]

In article <200602150019.25230.pogonyshev@gmx.net>, Paul Pogonyshev <pogonyshev@gmx.net> writes:

> Is it correct that

> 	(unencodable-char-position 0 1 'utf-8 nil "ы")

> evaluates to 0?

No.  But my Emacs (CVS HEAD) evaluates that to nil, which is
correct.  Which version of Emacs are you using?

> In general, what is the best way to determine if given character
> is encodable in given coding system?

Using unencodable-char-position is good.  You can also use
these:

(memq 'SOME_CODING_SYSTEM (find-coding-systems-region FROM TO))
  or
(memq 'SOME_CODING_SYSTEM (find-coding-systems-string STRING))

---
Kenichi Handa
handa@m17n.org

Re: unencodable-char-position bug?

[Paul Pogonyshev]

Kenichi Handa wrote:
> In article <200602150019.25230.pogonyshev@gmx.net>, Paul Pogonyshev <pogonyshev@gmx.net> writes:
> 
> > Is it correct that
> 
> > 	(unencodable-char-position 0 1 'utf-8 nil "ы")
> 
> > evaluates to 0?
> 
> No.  But my Emacs (CVS HEAD) evaluates that to nil, which is
> correct.  Which version of Emacs are you using?

Just update to CVS HEAD, the same result, zero.  Weird...

> > In general, what is the best way to determine if given character
> > is encodable in given coding system?
> 
> Using unencodable-char-position is good.  You can also use
> these:
> 
> (memq 'SOME_CODING_SYSTEM (find-coding-systems-region FROM TO))
>   or
> (memq 'SOME_CODING_SYSTEM (find-coding-systems-string STRING))

This works, i.e. (memq 'mule-utf-8 (find-coding-systems-string "ы"))
evaluates to non-nil...

Paul

Re: unencodable-char-position bug?

[Kenichi Handa]

In article <200602152248.59556.pogonyshev@gmx.net>, Paul Pogonyshev <pogonyshev@gmx.net> writes:

>> > Is it correct that
>> 
>> > 	(unencodable-char-position 0 1 'utf-8 nil "ы")
>> 
>> > evaluates to 0?
>> 
>> No.  But my Emacs (CVS HEAD) evaluates that to nil, which is
>> correct.  Which version of Emacs are you using?

> Just update to CVS HEAD, the same result, zero.  Weird...

Very weird...  Does someone else reproduce this bug?

---
Kenichi Handa
handa@m17n.org

Re: unencodable-char-position bug?

[Paul Pogonyshev]

Kenichi Handa wrote:
> In article <200602152248.59556.pogonyshev@gmx.net>, Paul Pogonyshev <pogonyshev@gmx.net> writes:
> 
> >> > Is it correct that
> >> 
> >> > 	(unencodable-char-position 0 1 'utf-8 nil "ы")
> >> 
> >> > evaluates to 0?
> >> 
> >> No.  But my Emacs (CVS HEAD) evaluates that to nil, which is
> >> correct.  Which version of Emacs are you using?
> 
> > Just update to CVS HEAD, the same result, zero.  Weird...
> 
> Very weird...  Does someone else reproduce this bug?

Can you give me some debugging hints?  I could try to look into it,
but I'm not familiar with Emacs C code.

Paul