nuxeo ECM, and where is Context.compressReader() in openJDK?

[Andrew Haley]

Robert P. J. Day wrote:
> On Fri, 13 Nov 2009, Andrew Haley wrote:
> 
>> I think it was a nonstandard extension to Rhino to read compressed
>> javascript.
> 
>   so sun's java is using that non-standard extension while openjdk
> isn't?  so what's the story in terms of adherence to formal java
> standards?

Rhino is javascript, not java.  It's not part of the Java API.

> i'm not a java guru, but this tells me that openjdk
> adheres to the "standard" (whatever that is) more strictly than sun.
> is that the conclusion i should be drawing here?

No.  Rhino is not part of Java: it is present in openjdk, but it
isn't strictly part of Java.  We need it for LiveConnect.

>>>   someone from nuxeo just pointed out that they don't guarantee a
>>> successful build under openJDK, only under *sun's* java.  so does
>>> that mean sun's java *would* have that class method?  or am i
>>> misinterpreting this?  just for the entertainment value, it would
>>> be nice to finish that build with openJDK.  does anyone here know
>>> what the story is for Context.compressReader()?  am i reading
>>> correctly that it simply doesn't exist in openJDK?
>> Yes.  All you need to do is replace the call to
>> Context.compressReader() with a dynamic lookup.  There's an example
>> at
>>
>> http://fisheye5.cenqua.com/browse/dwr/java/org/directwebremoting/impl/ShrinkSafeCompressor.java?r1=1.1&r2=1.2&u=-1&ignore=&k=o
>>
>> which shows exactly how to use Context.class.getMethod("compressReader",
>> and
>>
>> compressReaderMethod.invoke
> 
>   yes, i'd already found that page and realized what was happening
> there.  all i really want to confirm is that the code above that's
> failing is failing because it's badly-behaved and calling a
> non-standard extension that will cause it to break under openjdk.

That's right.  I wouldn't call it badly behaved, but it is depending on
a nonstandard extension to Rhino.

Andrew.