09.09.2019 16:25, Max Reitz wrote:
> On 30.08.19 18:12, Vladimir Sementsov-Ogievskiy wrote:
>> After backup-top filter appearing it's not possible to see dirty
>> bitmaps in top node, so use node-name instead.
>>
>> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsement...@virtuozzo.com>
>> ---
>> tests/qemu-iotests/124 | 83 ++++----
>> tests/qemu-iotests/257 | 49 ++---
>> tests/qemu-iotests/257.out | 364 +++++++++++++---------------------
>> tests/qemu-iotests/iotests.py | 27 +++
>> 4 files changed, 219 insertions(+), 304 deletions(-)
>
> [...]
>
>> diff --git a/tests/qemu-iotests/iotests.py b/tests/qemu-iotests/iotests.py
>> index 84438e837c..1906eb72f3 100644
>> --- a/tests/qemu-iotests/iotests.py
>> +++ b/tests/qemu-iotests/iotests.py
>> @@ -643,6 +643,33 @@ class VM(qtest.QEMUQtestMachine):
>
> [...]
>
>> + def check_bitmap_status(self, node_name, bitmap_name, fields):
>> + ret = self.get_bitmap(node_name, bitmap_name)
>> +
>> + return fields.items() <= ret.items()
>
> Why the <=? AFAIU, it will compare each of the arrays’ elements one by
> one and return true if all of the ones in @fields are less than or equal
> to the ones in @ret.
No, it's a valid way to check that one dict is subdict of another,
as dict_view works like Set class:
$ python3
Python 3.5.2 (default, Nov 23 2017, 16:37:01)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> {1,2,3} <= {4,5,6}
False
>>> {1,2,3} <= {0,3,2,1}
True
>>> {'c':1, 'b':2}.items() <= {'a':4, 'b':2, 'c':1}.items()
True
>>> {'c':1, 'b':2}.items() <= {'a':4, 'b':2, 'c':2}.items()
False
>>> {'c':1, 'b':2}.items() <= {'a':4, 'b':2, 'c':0}.items()
False
>>> {'c':1, 'b':2}.items() <= {'d':4, 'b':2}.items()
False
>>> {'c':1, 'b':2}.items() <= {'c':1, 'b':3}.items()
False
Bad thing is that it's not so for python2, in which we need to use .viewitems()
instead..
$ python2
Python 2.7.12 (default, Dec 4 2017, 14:50:18)
[GCC 5.4.0 20160609] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> {1,2,3} <= {0,3,2,1}
True
>>> {'c':1, 'b':2}.items() <= {'a':4, 'b':2, 'c':1}.items()
False
>>> {'c':1, 'b':2}.items() <= {'d':4, 'b':2}.items()
False
>>>
>>> {'c':1, 'b':2}.items() <= {'c':1, 'b':2}.items()
True
>>> {'c':1, 'b':2}.items() <= {'c':1, 'b':3}.items()
True
>>> {'c':1, 'b':2}.items() <= {'c':1, 'b':3}.items()
True
Should I care about python2 now, or we already dropped its support in iotests?
[and, honestly, I learned it all mostly from here
https://stackoverflow.com/questions/30818694/test-if-dict-contained-in-dict]
>
> But that would mean that the values given in @fields no longer need to
> be equal to the ones in @ret. On top of that, I suppose if
> @fields.items() at some index contains a key that is not equal to the
> key in @ret.items() at the same index, luck will determine whether the
> comparison passes or not.
>
> Why not just loop through all keys of @fields and check that @fields and
> @ret contain the same value for all of them?
>
> Max
>
--
Best regards,
Vladimir
