[rfc][icedtea-web] Console Output Encoding Fix

[Jie Kang]

Hello,

Thanks for the review!

Just an update and also addressing some issues you guys pointed out. First of all, I am working on improving the TeeOutputStream and also providing unit tests. Attached is what it looks like so far. Really just need to add unit tests and it will be submitted for review again;

I have added responses in-line.

>+    private void appendString(String s) {
>+        string.append(s);
>+        flushLog();
>+    }
>
>Why are you flushing there? I thing the flushing is done right. Have you encountered some  case when 
>some logs were missing?

The reason I chose to flush in write(byte, int, int) was to follow PrintStream's spec for automatic-line flushing. Flushing on new line for write(int) only occurs if automatic-line flushing is enabled and for write(byte, int, int), if automatic-line flushing is enabled, it is always flushed . So if we assume TeeOutputStream has 'automatic-line flushing' then it follows PrintStream's spec.

However, in our situation it's probably best to flush on new lines for both write(byte, int, int) and write(int) which is what I have chosen to do.

> >>> See my much longer reply :) -
> >>> http://mail.openjdk.java.net/pipermail/distro-pkg-dev/2014-July/028521.html
> >>>
> >>> Maybe you have arguments why write(int b) should not be fixed to.
> >>
> >> I have not said a word about write(int b). ;-) Please read carefully.
> >
> > I know. But I did (and was not sure bout it)
> >>
> >> As matter of fact, I cannot see much that needs fixing in write(int). It
> >> flush()es after every '\n'
> >> as write(byte[],int,int) does. Well, you could do String.valueOf(b) before
> >> appending though. ;-)
> >
> > ValueOf will not help when just one byte of multibyte char arrives (???).

Correct, good catch though I think this is a really really weird use case if someone ever decides to write multibyte characters byte by byte to the Stream when they can write the multibyte character all at once. Then again, if you read PrintStream's spec it says it will write the byte as is, so we should follow that and not convert it so early.

> 
> Oh now I understand what you mean. Good point! Yeah, then using ByteBuffer
> first
> and converting it to a String on a flush() is most probably the best thing to
> do. Same applies for write(byte[],int,int).

ByteBuffer requires a capacity on construction :\ I have chosen to use the ByteArrayOutputStream instead.


> 
> Jacob
> 

-- 

Jie Kang
-------------- next part --------------
A non-text attachment was scrubbed...
Name: console-1858-2.patch
Type: text/x-patch
Size: 2999 bytes
Desc: not available
URL: <http://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20140715/1c48e538/console-1858-2-0001.patch>