Unicode and XML regular expression changes integrated to next branch as of commit below. pathsep support, logging, and windows unit test infrastructure remain to be reviewed.
commit 829041ea34558c054ca434094929f5bf1d89e22e
Author: Jon Foster <jon@…>
Date: Sat May 26 00:27:35 2012 +0100
Fix handling of caret and dollar in regexps.
They are normal characters in XSD REs but control characters
in Python REs, so we have to escape them.
Signed-off-by: Peter A. Bigot <pab@…>
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
As of the commit below, all changes that have been accepted have been merged. This includes the Unicode and XML regular expression changes, and most fixes for Windows support. In some cases the specific solution was changed slightly.
The windows test scripts were not taken. I don't want a solution that requires maintaining a list of known tests.
The modifications to utility.NormalizeLocation for Windows were also not taken. I agree with the patch's comment that the function should be used only on URIs. If something that is not a URI is given to it, then that something should be fixed. I believe the existing code is essentially correct for normal processing of schemas, and it was not easy to get that working so I don't want to change its behavior unless it's wrong for its intended purpose.
commit b9099d21956c3135756cf6b4a04357ae90f19c0e
Author: Jon Foster <jon@…>
Date: Sat Jun 23 01:03:31 2012 +0100
Normalize schema paths in tests.
Change schema paths to be normal absolute paths, with consistent
slashes and without "..".
Signed-off-by: Peter A. Bigot <pab@…>
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Unicode and XML regular expression changes integrated to next branch as of commit below. pathsep support, logging, and windows unit test infrastructure remain to be reviewed.
commit 829041ea34558c054ca434094929f5bf1d89e22e
Author: Jon Foster <jon@…>
Date: Sat May 26 00:27:35 2012 +0100
As of the commit below, all changes that have been accepted have been merged. This includes the Unicode and XML regular expression changes, and most fixes for Windows support. In some cases the specific solution was changed slightly.
The windows test scripts were not taken. I don't want a solution that requires maintaining a list of known tests.
The modifications to utility.NormalizeLocation for Windows were also not taken. I agree with the patch's comment that the function should be used only on URIs. If something that is not a URI is given to it, then that something should be fixed. I believe the existing code is essentially correct for normal processing of schemas, and it was not easy to get that working so I don't want to change its behavior unless it's wrong for its intended purpose.
commit b9099d21956c3135756cf6b4a04357ae90f19c0e
Author: Jon Foster <jon@…>
Date: Sat Jun 23 01:03:31 2012 +0100