Re: `.newsrc.eld' saves chinese group name in wrong coding

[Chong Yidong]

Kenichi Handa <address@hidden> writes:

> It seems that Gnus prints a unibyte string (perhaps created
> by encoding the original string "流星99.本地测试" by gb2312,
> or created by reading an external gb2312 data (filename?)
> without decoding) in a work buffer by `print', and then
> saves that buffer.
>
> There are two problems here.  The first one is to use
> `print' (or `prin1') on a form that contains a unibyte
> string.  This leads to the byte sequence '\\' '3' '0' '1'
> for the byte \301 of an unibyte string.  Next, the encoding
> of unibyte string (gb2312) doesn't match with the coding tag
> (utf-8-emacs).
>
> I think the correct fix is to print a form containing a
> multibyte string, and then encode the buffer on saving with
> a coding system that match with the coding tag.

Where is this code located?