Hello,
I've complitely reworked the patch according to investigation in [1].
1) usage of JavaLangAccess allows to rid reflection completely and access
package-private String.isLatin1 via SharedSecrets only,
so no black wizardry any more
2) now we have only a few cases with noticeable regeressio
Oh, I was unaware of it! Thanks!
13.02.2020, 21:49, "Brent Christian" :
> As a point of interest, some investigation of updating StringJoiner for
> CompactStrings was done a while back.
>
> See https://bugs.openjdk.java.net/browse/JDK-8148937
>
> -Brent
>
> On 2/3/20 2:38 PM, Сергей Цыпанов wrote:
As a point of interest, some investigation of updating StringJoiner for
CompactStrings was done a while back.
See https://bugs.openjdk.java.net/browse/JDK-8148937
-Brent
On 2/3/20 2:38 PM, Сергей Цыпанов wrote:
Hello,
as of JDK14 java.util.StringJoiner still uses char[] as a storage of glued
-latin. This likely to reduce
>> regression related to usage of reflection.
>>
>> Regards,
>> Sergey Tsypanov
>>
>> 05.02.2020, 23:21, "fo...@univ-mlv.fr" :
>> > - Mail original -
>> >> De: "Сергей Цыпанов"
n case at least one
>> of them is non-latin. This likely to reduce
>> regression related to usage of reflection.
>>
>> Regards,
>> Sergey Tsypanov
>>
>> 05.02.2020, 23:21, "fo...@univ-mlv.fr" :
>> > - Mail original -
>>
deal with EnumSet?
>> >
>> > yes !
>> > crossing package boundary in a non public way is not free,
>> > but given that StringJoiner is used quite often (directly or indirectly
>> using Collectors.joining()), it may worth the cost.
>> >
>>
-- Mail original -
> >> De: "Сергей Цыпанов"
> >> À: "Remi Forax" , "core-libs-dev"
> >>
> >> Envoyé: Mercredi 5 Février 2020 22:12:34
> >> Objet: Re: [PATCH] Enhancement proposal for java.util.StringJoiner
>
rax" , "core-libs-dev"
>>
>> Envoyé: Mercredi 5 Février 2020 22:12:34
>> Objet: Re: [PATCH] Enhancement proposal for java.util.StringJoiner
>
>> Hello,
>>
>>> If you want to optimize StringJoiner, the best way to do it is to use the
>
- Mail original -
> De: "Сергей Цыпанов"
> À: "Remi Forax" , "core-libs-dev"
>
> Envoyé: Mercredi 5 Février 2020 22:12:34
> Objet: Re: [PATCH] Enhancement proposal for java.util.StringJoiner
> Hello,
>
>> If you want to o
y
04.02.2020, 12:12, "Remi Forax" :
> - Mail original -
>> De: "Сергей Цыпанов"
>> À: "jonathan gibbons" , "core-libs-dev"
>>
>> Envoyé: Mardi 4 Février 2020 08:53:31
>> Objet: Re: [PATCH] Enhancement proposal
- Mail original -
> De: "Сергей Цыпанов"
> À: "jonathan gibbons" , "core-libs-dev"
>
> Envoyé: Mardi 4 Février 2020 08:53:31
> Objet: Re: [PATCH] Enhancement proposal for java.util.StringJoiner
> Hello,
Hi Sergey,
>
> I'd
Hello,
I'd probably agree about a new class in java.lang, but what is wrong about
exposing package-private method
which doesn't modify the state of the object and has no side effects?
04.02.2020, 00:58, "Jonathan Gibbons" :
> Sergey,
>
> It is equally bad to create a new class in the java.lang p
Sergey,
It is equally bad to create a new class in the java.lang package as it
is to add a new public method to java.lang.String.
-- Jon
On 2/3/20 2:38 PM, Сергей Цыпанов wrote:
Hello,
as of JDK14 java.util.StringJoiner still uses char[] as a storage of glued
Strings.
This applies for the
13 matches
Mail list logo
