Hi Drew.
On 04/28/11 02:54 PM, Drew Fisher wrote:
I also grabbed and fixed
7040051 <http://monaco.us.oracle.com/detail.jsf?cr=7040051> Unit
test failure in test_p5i.py even after 7013385
<http://monaco.us.oracle.com/detail.jsf?cr=7013385> is fixed.
At the same time.
According to my understanding of this bug report's evaluation, if
Publisher() gets no args it will bomb.
A search through slim_source reveals many more places where
Publisher() is called without args, that the code review shows are
fixed. Are these all supposed to get no args (as for testing bogus
conditions, etc)?
Files I saw include:
./lib/install_transfer/test/test_ips.py (1)
- This is inside a method called test_default_publisher(), which
makes me wonder whether calling Publisher here without args is OK and
that it takes a default?
TransferIPS will set a default publisher if you don't specify one.
TransferP5I will not. This is why I had to label the publisher for
the test.
OK.
./lib/install_transfer/test/test_svr4.py (many)
This file's contents look suspicious. (But is this file on the way
out as it's SVR4?)
The svr4 code wasn't touched. I think Jean recycled the same names
for these tests.
Looks like ips.py, svr4.py and p5i.py (from solaris_install.transfer in
proto area) all import Publisher from the same place:
from solaris_install.transfer.info import Publisher
That said, if you have run all of the tests and they pass, I won't argue
with that.
Are these all supposed to get no args (as for testing bogus
conditions, etc)?
I think the rest of the tests cover this.
Also, all transfer tests run successfully, even the ones I didn't
touch. :)
OK.
Thanks,
Jack
Thanks for looking, Jack.
-Drew
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
