RFR:JDK-8145633-Adjacent value parsing not supported for Localized Patterns

Roger Riggs Roger.Riggs at Oracle.com
Tue Dec 20 21:41:31 UTC 2016


Hi Nadeesh,

On 12/20/2016 2:34 PM, nadeesh tv wrote:
>
> Hi Roger & Stephen ,
> Thanks for the comments.
>
> Please see the updated webrev 
> http://cr.openjdk.java.net/~ntv/8145633/webrev.12/
>
> Changes included :
>
> 1. Doc changes and cosmetic changes suggested by Roger (except the 
> note about delay..)
The comments at 4776-4779 are fine.
The issue came from passing null at line 4809 to the NumberPrinterParser 
constructor
that expects the field argument to be non-null, and wanting some 
explanation of that disconnect.

Other than that, I'm fine with the changes.

Roger

>
> 2.  Changed the  behavior of 'e' to parse only 1 digit as suggested by 
> Stephen.  Changed the existing test cases for this.
>
> Thanks and Regards,
> Nadeesh
>
> On 12/20/2016 10:48 PM, Stephen Colebourne wrote:
>> In the test provider_adjacentValuePatterns2(), please add
>>
>> {"YYYYwwe", Y, w, c, "20161201", 2016, 12, 1},
>>
>> This should succeed, because a single number is all that is needed to
>> parse day-of-week. (So, it will need to be removed from the invalid
>> patterns test).
>>
>> Line 1869 will need to change to "count, count, count" to make the 
>> tests pass.
>>
>> Otehrwise, looks fine, thanks.
>> Stephen
>>
>>
>> On 20 December 2016 at 09:55, nadeesh tv <nadeesh.tv at oracle.com> wrote:
>>> Hi all,
>>>
>>> BugId: https://bugs.openjdk.java.net/browse/JDK-8145633
>>>
>>> Issue:  Support adjacent value parsing for  Localized Patterns
>>>
>>> Webrev:  http://cr.openjdk.java.net/~ntv/8145633/webrev.10/
>>>
>>> Pattern 'c' and 'W'  were previously allowed to have 'zero padding' 
>>> which
>>> was not explicitly mentioned in CLDR
>>> (http://unicode.org/reports/tr35/tr35-dates.html).
>>> To allow 'c' and 'W' to take part in adjacent value  parsing ( at 
>>> the same
>>> time, 2 digits are  not required for these patterns), restricted the 
>>> max
>>> width of these patterns to 1.
>>>
>>> Special thanks to Stephen for the help.
>>>
>>> -- 
>>> Thanks and Regards,
>>> Nadeesh TV
>>>
>



More information about the core-libs-dev mailing list