281156 - PyXPCOM build fails if python uses 4-byte unicode chars

Status: RESOLVED FIXED

(Core :: XPCOM, defect)

Description

[Sebastian Tusk]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20050110 Firefox/1.0 (Debian package 1.0+dfsg.1-2)
Build Identifier: 

The current implementation of PyXPCOM uses a set of functions (*UCS2) to convert
from and to python string objects.

These function are only present if python is build to use 2-byte unicode
characters. On many linux machines python is build with 4-byte unicode chars.
The build of pyXPCOM then fails with unresolved symbol errors for the mentioned
functions.

PyXPCOM should use public functions from the python library only. The *UCS2
function aren't public.

Reproducible: Always

Steps to Reproduce:
build PyXPCOM on debian/unstable

Comment 1

[Sebastian Tusk]

Attachment: patch writteb by Shane Hathaway

See http://aspn.activestate.com/ASPN/Mail/Message/pyxpcom/2484016 for a
discussion of the problem.

Comment 2

[Mark Hammond [:markh] [:mhammond]]

Best I can tell this patch has already been applied to the trunk, so is out of date.

Comment 3

[Christian Persch (GNOME) (away; not receiving bug mail)]

(In reply to comment #2)
> Best I can tell this patch has already been applied to the trunk, so is out of
date.

Yes, the patch from bug 297598 incorporated this patch.

Comment 4

[Mark Hammond [:markh] [:mhammond]]

Closing as per comments.