Release: GnuPG 1.2.1

Environment

SuSE Linux 8.1 on Pentium II
GnuPG 1.2.1 as downloaded from www.gnupg.org
idea.c put in the ciphers directory

Reproduced with AIX 5.1 on IBM RS/6000.

Description

With a PGP2 test key, running "gpg --charset iso-8859-1 --no-utf8-strings --quiet --yes --edit-key 'Rainer Perske <rainer.perske@uni-muenster.de>' creates a list of UIDs where those UIDs containing 8-bit characters are displayed truncated, e.g.
...
(2) Rainer der Gro\xdf\x65 Bruder <rainer@persk
...
This line should read:
(2) Rainer der Gro\xdf\x65 Bruder <rainer@perske.net>
It is truncated exactly 6 characters before its end.
Thus obviously, when the output length is calculated, it is forgotten to take the \x-encoding into account.

Either the \x-encoding should not be done, or the length calculation should take the encoding into account.

(I've set the severity to serious instead of non-critical because this wrong display breaks the key trust editing routines of my e-mail program.)

How To Repeat

I have created my test key with PGP 2.6.3in.
To include 8-bit characters in the userid, I had to compile PGP with -funsigned-char, otherwise a bug in
prohibits including 8-bit characters.
I did not test it with GnuPG generated keys.

Fix

Fixed in 1.2.2.