On Tue, 26 Sep 2006, Lieven Govaerts wrote:
...
> We need some more status regression tests, do you mind writing one
> for this scenario? I'll probably add some more later.
I modified the test for issue #2533 (status test #24) to also use
'-vN', which is a good regression test variation in the first place.
New patch attached/log message below.
Paul indicated on IRC that he's found additional problems with r21638,
but had to run before he could elaborate. He indicated that he'd pick
this up again tomorrow morning.
[[[
A follow-up to r21638 (for issue #2533), preventing double status
reporting of a changed WC root when 'svn status' is invoked with the
'-v' flag.
* subversion/libsvn_wc/status.c
(close_directory): When root WC dir is out of date, avoid reporting
it twice (once correctly, once incompletely) when
DIR_BATON->edit_baton->get_all is set (which is always the case
with '-v').
* subversion/tests/cmdline/stat_tests.py
(status_update_with_incoming_props): Add trailing comma on the last
element in a dict definition to reduce change in case additional
elements are added.
Patch by: pburba
dlr
]]]
- application/pgp-signature attachment: stored
Received on Tue Sep 26 22:06:55 2006