rfc2231 bug

rfc2231 bug

[Katsumi Yamaoka]

Hi,

The most recent changes in rfc2231-parse-string cause a bug.
This is the test case:

(rfc2231-parse-string
 (concat "Content-Disposition: attachment;"
	 (rfc2231-encode-string
	  "filename"
	  "一二三四五六七八九零一二三四五六七八九零.pdf")))
 => ("content-disposition:attachment"
     (filename . "一二三四五六七八九零一二三四8^O;<7H,6eNm\e(B.pdf"))

It is because each segmented encoded text is decoded before
concatenating them.  I once fixed this problem as follows:

2004-12-21  Katsumi Yamaoka  <yamaoka@jpl.org>

	* rfc2231.el (rfc2231-parse-string): Decode encoded value after
	concatenating segments rather than before concatenating them.
	Suggested by ARISAWA Akihiro <ari@mbf.ocn.ne.jp>.

Re: rfc2231 bug

[Katsumi Yamaoka]

>>>>> In <b4mzmijvl3y.fsf@jpl.org>
>>>>>	Katsumi Yamaoka <yamaoka@jpl.org> wrote:

> The most recent changes in rfc2231-parse-string cause a bug.

I found there's one more bug in rfc2231.el, though it is
irrelevant to the recent changes because yesterday's version
doesn't solve it.  The bug was reported by Reiner Steib and I
once fixed[1].  However, I might have enbugged after that.

[1]
http://news.gmane.org/group/gmane.emacs.gnus.general/thread=60638/force_load=t

I'll look into it (but I have no time nowadays, so I appreciate
someone debugging it).

Re: rfc2231 bug

[Katsumi Yamaoka]

>>>>> In <b4mzmijvl3y.fsf@jpl.org>
>>>>>	Katsumi Yamaoka <yamaoka@jpl.org> wrote:

> The most recent changes in rfc2231-parse-string cause a bug.
> This is the test case:

> (rfc2231-parse-string
>  (concat "Content-Disposition: attachment;"
> 	 (rfc2231-encode-string
> 	  "filename"
> 	  "一二三四五六七八九零一二三四五六七八九零.pdf")))
>  => ("content-disposition:attachment"
>      (filename . "一二三四五六七八九零一二三四8^O;<7H,6eNm\e(B.pdf"))

I've fixed it.  Yay!

>>>>> In <b4mk69nu4qo.fsf@jpl.org> Katsumi Yamaoka wrote:

> I found there's one more bug in rfc2231.el, though it is
> irrelevant to the recent changes because yesterday's version
> doesn't solve it.  The bug was reported by Reiner Steib and I
> once fixed[1].  However, I might have enbugged after that.

Please ignore it.  I seem likely to have misunderstood it.

Re: rfc2231 bug

[Lars Magne Ingebrigtsen]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

> It is because each segmented encoded text is decoded before
> concatenating them.  I once fixed this problem as follows:
>
> 2004-12-21  Katsumi Yamaoka  <yamaoka@jpl.org>
>
> 	* rfc2231.el (rfc2231-parse-string): Decode encoded value after
> 	concatenating segments rather than before concatenating them.
> 	Suggested by ARISAWA Akihiro <ari@mbf.ocn.ne.jp>.

Oh, sorry.  I wondered why the decoding was done that way, but
obviously not enough.  :-)  Thanks for fixing this again.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen