wrong @ conversion in iso-2022-jp text

Sugihara san (who has reported about the problem of
non-ASCII encoded filenames to this mailing list some days
ago) reported to me that v2.6.0 with iso2022jp.pl converts
Japanese text improperly.

For example,

| Subject: =?iso-2022-jp?B?GyRCQCQzJhsoQg==?=

should be decoded as

| ^[ $ B @ $ 3 & ^[ ( B         # SE-KAI, means WORLD
       ^^^ ^^^
       SE  KAI

but v2.6.0 converts this as

| ^[ $ B @ $ 3 & ^[ ( B    # BROKEN!


This problem occurs in mail body (i.e., between
<!--X-Body-of-Message--> and <!--X-Body-of-Message-End-->),
too.


I believe this is serious problem for Japanese users.
Would you fix this problem?
(In fact, I tried to fix this problem, but I couldn't...)

I attached sample mail.
My rcfile is just:

<CHARSETCONVERTERS>
iso-2022-jp; iso_2022_jp::str2html; iso2022jp.pl
</CHARSETCONVERTERS>


Best regards,

-- 
Takashi P.KATOH

> --- Begin Message ---
> \xA8焉扤扤, means star.
>
>
> --- End Message ---