<i18n dev> Review request : 7180362: RFE: Implement date cutover functionality for currency.properties file

[Seán Coffey]

Hi Naoto,

thanks for reviewing. Comments below.

On 29/08/2012 19:55, Naoto Sato wrote:
> Hi Sean,
>
> It looks good overall. There are some minor comments that I have 
> below. All in Currency.java
>
> Line 666: Comments refer to "GMT", while the class description uses 
> "UTC." I'd use "UTC" here as well (and in the actual code on line 728).
Agreed. I've only used "GMT" now in the datestamp string examples. I 
think that's the normal case ?
>
> Line 681, "countOccurrences(curdata, ',') == 3": Should "==" be ">="?
True - good to catch/flag any format issue.
>
> Line 727: For locale invariant operation, I'd use "Locale.ROOT".
Done.

> Line 731-734: These can be simplified to "return 
> System.currentTimeMillis() >= time;"
Done. much cleaner!
new webrev : http://cr.openjdk.java.net/~coffeys/webrev.7180362.2.jdk8/ 
<http://cr.openjdk.java.net/%7Ecoffeys/webrev.7180362.2.jdk8/>

regards,
Sean.

>
> Naoto
>
> On 8/28/12 6:48 AM, Seán Coffey wrote:
>>
>> 7180362 deals with an enhancement to allow the JRE specify cutover dates
>> when currency.properties file is provided. I've added the required
>> syntax to the new javadoc comments in Currency class.
>>
>> bug report :http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7180362
>> webrev : http://cr.openjdk.java.net/~coffeys/webrev.7180362.1.jdk8/
>> <http://cr.openjdk.java.net/%7Ecoffeys/webrev.7180362.1.jdk8/>
>>
>> I hope to port an almost identical change to 7u shortly. (minus API
>> javadoc comments)
>>
>> I've kept the testcase logic different to the JRE logic around how the
>> new property is parsed to better
>> test the golden result expectations. (both approaches should be equal)
>>
>> Regards,
>> Sean.
>