This large (700+ line) patch contains all of the
changes I made to current CVS (22 March 2005) so all of
the regression tests pass for Python 2.1, 2.2, 2.3 and
2.4 on my Debian 'sarge' box.
I believe that most of these changes will be applicable
to other unix-like systems as well, as long as Python
is installed in a "standard" location in /usr. Many of
the tests will fail if this is not the case, and I
don't think there's a general way to solve that problem
without completely changing how the tests are run.
Some of these changes are just special cases for one
version of Python or the other. Others are changes to
line numbers as listed in the warnings, or to change
the order of warnings to match "reality". A few others
are changes to source line where the person who created
the test had Python in a non-standard location, i.e. in
/home/neal/build/python/c3/Lib rather than in
I have not made any judgements as to whether these
various changes are correct or not. I have briefly
looked them over to see that they look generally
sensible, and I have assumed that if the tests pass for
one version of Python, that the code is correct for
other versions of Python, but the expected results
might differ. I am just not familar enough with the
code to understand exactly what differences are
expected between Python versions.
Log in to post a comment.