Re: [XEN PATCH v1] tools/python: Add python3 compatibility

[Javi Merino <javi.merino@xxxxxxxxx>]

On Wed, Oct 11, 2023 at 12:34:27AM +0800, Andrew Cooper wrote:
> On 10/10/2023 10:18 pm, Javi Merino wrote:
> > Most of the work for python3 compatibility was done in
> > 1430c5a8cad4 (tools/python: Python 3 compatibility, 2019-12-18).  This
> > just adds a few builtins needed for python3.
> >
> > Resolves: xen-project/xen#156
> >
> > Signed-off-by: Javi Merino <javi.merino@xxxxxxxxx>
> > ---
> >
> > I haven't tested it.
> >
> >  README                                     | 1 +
> >  tools/python/scripts/convert-legacy-stream | 3 +++
> >  2 files changed, 4 insertions(+)
> >
> > diff --git a/README b/README
> > index 855db01d36..44ed88c392 100644
> > --- a/README
> > +++ b/README
> > @@ -51,6 +51,7 @@ provided by your OS distributor:
> >      * POSIX compatible awk
> >      * Development install of zlib (e.g., zlib-dev)
> >      * Development install of Python 2.7 or later (e.g., python-dev)
> > +      - If using Python 2.7, you also need the future module.  This is not 
> > needed for Python 3.
> >      * Development install of curses (e.g., libncurses-dev)
> >      * Development install of openssl (e.g., openssl-dev)
> >      * Development install of x11 (e.g. xorg-x11-dev)
> > diff --git a/tools/python/scripts/convert-legacy-stream 
> > b/tools/python/scripts/convert-legacy-stream
> > index 7fe375a668..26a66c50fc 100755
> > --- a/tools/python/scripts/convert-legacy-stream
> > +++ b/tools/python/scripts/convert-legacy-stream
> > @@ -8,6 +8,9 @@ Convert a legacy migration stream to a v2 stream.
> >  from __future__ import print_function
> >  from __future__ import division
> >  
> > +from builtins import zip
> > +from builtins import range
> > +from builtins import object
> 
> It can't be object because most other scripts use it just fine in py2
> and py3.

This just makes the VM() class behave like python3: str(vm) will
return a unicode string.  After a quick glance at the code I don't
think the code currently does this, but at least the behaviour will
now be consistent between the two versions.

> There's only one single use of zip and range in this script, and range
> is clearly fine (although it wants to be xrange() on py2 and we do
> opencode that places), so I'm guessing the problem is with zip(), but
> it's not exactly clear why?

These changes just make the code be more consistent between python2
and python3.  As I said under the commit message, I have not tested
the changes.

Cheers,
Javi