RFR 8048619: Implement tests for converting PKCS12 keystores

zaiyao liu zaiyao.liu at oracle.com
Thu Dec 4 08:22:40 UTC 2014


Hi Max,

Please help to check again: 
http://cr.openjdk.java.net/~rhalade/8048619/webrev.04/

Also please help to push it if ok:
Full comments:
8048619: Implement tests for converting PKCS12 keystores
Reviewed-by: weijun
Contributed-by: Zaiyao Liu <zaiyao.liu at oracle.com>

Thanks and Regards.

Kevin


于 2014/12/3 15:37, Wang Weijun 写道:
> Looks fine.
>
> One nit, line 231, remove "Certs don't match ".
>
> Thanks
> Max
>
>> On Dec 3, 2014, at 10:26, zaiyao liu <zaiyao.liu at oracle.com> wrote:
>>
>> Hi Max,
>>
>> Please help to review the update:
>> http://cr.openjdk.java.net/~rhalade/8048619/webrev.02/
>>
>> Thanks
>> Kevin
>> 于 2014/9/23 15:08, zaiyao liu 写道:
>>> Hi Max,
>>>
>>> Please help to check this update:http://sqeweb.us.oracle.com/net/sqenfs-1/export1/comp/jsn/users/kevin1/webrev/8048619/webrev/
>>>
>>> Thanks
>>>
>>> Kevin
>>> 于 2014/9/22 17:10, Wang Weijun 写道:
>>>> On Sep 22, 2014, at 16:11, zaiyao liu <zaiyao.liu at oracle.com> wrote:
>>>>
>>>>> Hi Max,
>>>>>
>>>>> Thanks for review. please review the update: http://sqeweb.us.oracle.com/net/sqenfs-1/export1/comp/jsn/users/kevin1/webrev/8048619/webrev/, I will ask Rajan to create open webrev link after you passed.
>>>>   194         if (a.size() != b.size()) {
>>>>
>>>> How about "if (a.size() != 1 || b.size() != 1)"?
>>>>
>>>>   239-250
>>>>
>>>> This can be as simple as Arrays.equals(certsA, certsB).
>>>>
>>>> I would still like a single compareKeyStore() no matter what the size of the keystore is.
>>>>
>>>> --Max
>>>>
>>>>> Thanks
>>>>>
>>>>> Kevin
>>>>> 于 2014/9/19 15:34, Wang Weijun 写道:
>>>>>> In compareCerts(), you should not compare Certificate.toString(), its equals() method is more reliable. There is no need to define compareKeys() and compareCerts(). Instead, you should try not to repeat lines 223-231 on 252-260.
>>>>>>
>>>>>> The two calls on lines 204 and 205 have keystore names exchanged but keypass order is the same. Also, with two calls it means the comparisons of keys and certs are duplicated. The else block on line 206 looks strange. Do you mean the size is always 1 here? If so, check it. Otherwise, there is no guarantee the aliases appear in the same order.
>>>>>>
>>>>>> I would be glad to see a generalized comparison method no matter what the keystore size is.
>>>>>>
>>>>>> Minor issues:
>>>>>>
>>>>>> It will be clear if you divide the constant strings at the beginning into 2 parts, one for provider names, and one for algorithms.
>>>>>>
>>>>>> Line 143 and 145, there should be spaces around the testCase name.
>>>>>>
>>>>>> --Max
>>>>>>
>>>>>> On Sep 19, 2014, at 11:04, zaiyao liu <zaiyao.liu at oracle.com> wrote:
>>>>>>
>>>>>>> Hi Max,
>>>>>>>
>>>>>>> Can you help to review it?
>>>>>>>
>>>>>>> Thanks
>>>>>>>
>>>>>>> Kevin
>>>>>>> 于 2014/9/1 13:25, Wang Weijun 写道:
>>>>>>>> On vacation now. Can you look for someone else? I will be back in Sep 17 if you are not in a hurry.
>>>>>>>>
>>>>>>>> --Max
>>>>>>>>
>>>>>>>> On Sep 1, 2014, at 9:37, zaiyao liu <zaiyao.liu at oracle.com> wrote:
>>>>>>>>
>>>>>>>>> Hi Max,
>>>>>>>>>
>>>>>>>>> Please review the code change,the purpose of this fix is implement tests that convert PKCS12 keystores to other formats.
>>>>>>>>>
>>>>>>>>> Webrev: http://cr.openjdk.java.net/~tyan/kevin/JDK-8048619/webrev01/
>>>>>>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8048619
>>>>>>>>>
>>>>>>>>> Thanks
>>>>>>>>>
>>>>>>>>> Kevin



More information about the security-dev mailing list