[13] RFR: 8220037: Inconsistencies of generated timezone files between Windows and Linux

[naoto.sato at oracle.com]

Thanks, Roger. Webrev is modified accordingly:

http://cr.openjdk.java.net/~naoto/8220037/webrev.01/

Naoto

On 5/6/19 11:59 AM, Roger Riggs wrote:
> Hi Naoto,
> 
> CLDRConverter:
>   358-271: Is a lambda viable here?  Not significantly different, just a 
> more contemporary style.
>      (o1, o2) -> {...}.
> 
> CLDRDisplayNamesTest.java:
>   136-137, 155, 158:  Another option is to have the stream return the count
> of the number of errors and use peek to print the markers.
> 
> errors += List.of(Locale.ROOT,
>          Locale.CHINA,
>          Locale.GERMANY,
>          Locale.JAPAN,
>          Locale.UK,
>          Locale.US,
>          Locale.forLanguageTag("hi-IN"),
>          Locale.forLanguageTag("es-419")).stream()
>          .peek(System.out::println)
>          .map(l -> DateFormatSymbols.getInstance(l).getZoneStrings())
>          .flatMap(zoneStrings -> Arrays.stream(zoneStrings))
>          .filter(namesArray -> Arrays.stream(namesArray)
>                  .anyMatch(aName -> aName.equals(NO_INHERITANCE_MARKER)))
>          .peek(marker -> {
>              System.err.println("No inheritance marker detected with 
> tzid: " + marker[0]);
>          })
>          .count();
> 
> 
> 151:  findAny() short circuits the count and stream so no more than 1 
> error would be counted.
> 
> $.02, Roger
> 
> 
> 
> On 05/06/2019 01:24 PM, naoto.sato at oracle.com wrote:
>> Hello,
>>
>> Please review the fix to the following issue:
>>
>> https://bugs.openjdk.java.net/browse/JDK-8220037
>>
>> The proposed fix is located at:
>>
>> http://cr.openjdk.java.net/~naoto/8220037/webrev.00/
>>
>> The inconsistency comes from the different enumeration order of CLDR 
>> source files on each platform. Fix is to sort its order uniquely.
>>
>> Naoto
>