2to3-2.7 crashed with UnicodeDecodeError in read(): 'utf8' codec can't decode byte 0xf6 in position 22890: invalid start byte

[Scott Kitterman]

Binary package hint: python2.7

Ran 2to3 -wv on the DNS directory in python-dns.

ProblemType: Crash
DistroRelease: Ubuntu 11.04
Package: python2.7 2.7.1-3
ProcVersionSignature: Ubuntu 2.6.38-1.28-generic 2.6.38-rc2
Uname: Linux 2.6.38-1-generic i686
NonfreeKernelModules: wl
Architecture: i386
Date: Sun Feb 6 01:41:58 2011
EcryptfsInUse: Yes
ExecutablePath: /usr/bin/2to3-2.7
InstallationMedia: Kubuntu 11.04 "Natty Narwhal" - Alpha i386 (20110202)
InterpreterPath: /usr/bin/python2.7
ProcCmdline: /usr/bin/python2.7 /usr/bin/2to3 -wv DNS
ProcEnviron:
 SHELL=/bin/bash
 LANG=en_US.UTF-8
 LANGUAGE=
PythonArgs: ['/usr/bin/2to3', '-wv', 'DNS']
SourcePackage: python2.7
Title: 2to3-2.7 crashed with UnicodeDecodeError in read(): 'utf8' codec can't decode byte 0xf6 in position 22890: invalid start byte
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare

[Scott Kitterman (kitterman)]

The same issue affects all four 2to3 versions in the archive.

[Matthias Klose (doko)]

which file was processed?

[Scott Kitterman (kitterman)]

It was DNS/Lib.py. Reduced test case about to be attached.

[Scott Kitterman (kitterman)]

2to3 testcase will reproduce the issue.

[Matthias Klose (doko)]

this seems to be an iso encoding, not utf, as the header suggests.

[Scott Kitterman (kitterman)]

It may be encoded incorrectly, but that doesn't mean crashing is the right thing to do. It should be handled more gracefully.

[dino99 (9d9)]

This version has expired long ago; no more supported