RFR: 8300819: -Dfile.encoding=Cp943C option does not work as expected since jdk18 [v2]

[Ichiroh Takiguchi]

On Tue, 24 Jan 2023 12:51:31 GMT, Alan Bateman <alanb at openjdk.org> wrote:

> Do you know if there is any configuration on AIX that would derive Cp943C as the default charset? That is, are they running with -Dfile.encoding=Cp943C on the AIX systems or is it chosen by default. This goes to the question as to whether they just moving these applications to Linux and expecting the default charset to be the same.

In my understanding, my client uses `-Dfile.encoding=Cp943C` option on Japanese IBM-943 locale on AIX.
Default charset on Japanese IBM-943 locale with IBM Java8 and OpenJDK JDK11+ is x-IBM943C(Cp943C).
(We need to use `-Dfile.encoding=Cp943C` for OpenJDK JDK8.)
We never thought we could just move to Linux because of JEP-400. But we don't move the apps all at once to Linux.
We expected that we could change default charset by `-Dfile.encoding=Cp943C`, at least until Java8 EOS.

> Do you know which APIs they are using? We filled in the gaps many releases ago so that all APIs that do encoding/decoding allow the charset to be specified and I'm wondering why they don't use those.

We checked String.getByte()/new String(...)/Reader/Writer/ByteArrayOutputStream.toString()...
Is there good way to pick up which parts need to be fixed ?

-------------

PR: https://git.openjdk.org/jdk/pull/12132