RFR: 8153213: Jar manifest attribute "Multi-Release" accepts any value

[Paul Sandoz]

> On 8 Apr 2016, at 18:54, Steve Drach <steve.drach at oracle.com> wrote:
> 
>>> I’ve put up a new webrev that addresses the issue of having spaces before (and after) the value “true” in the Multi-Release attribute.
>>> 
>> 
>> Is some or all of that really necessary? since the we can specify domain of values.
> 
> I think it is.  The spec states that one can have an arbitrary amount of leading/trailing spaces in the value part of the attribute.  Apparently attribute values could also have “hidden” characters such as vertical tab or nak for example.  I wish the spec was tighter here.  I’m merely stating the the domain can be
> “ *TRUE *” for upper/lower case letters.
> 

AFAICT the so called “spec" says nothing about trimming white space from values and although not explicitly called out the actual value has to be what constitutes the character sequence for the “otherchar" definition (otherwise the continuation space and newline characters would also be part of the actual value and that makes no sense).

So it seems you may have quite a bit of wiggle room here :-) and it’s up to each attribute definition to state what constitutes its domain of possible valid values based on “otherchar”.


>> 
>> For example, the Sealed attribute can take on a value of “true” or “false” (case ignored), but there is no white space trimming.
> 
> Well then “Sealed:   true” won’t work, but the spec says it should work.
> 

Where does the “spec” state that it should?

The value is literally the string “ true”, at least in one interpretation of the “spec” :-)

I suspect it’s ok to go with finding “Multi-Release: true” (lower/upper case) as in your first patch, then check if it is terminated with a newline (and ignore the continuation case)

Paul.


> I am fine without doing this nonsense, but I think we need to change the spec to make it correct.  We could change the definition of "otherchar” to something like this
> 
> otherchar:=ALPHA EXTALPHANUM*
> ALPHA:=[A-Z | a-z | _]
> EXTALPHANUM:=[ALPHA | 0-9 | SPACE]
> 
> Even this will still allow trailing spaces, so after matching TRUE we’d need to make sure the next char is SPACE, \r, or \n.
> 
> Other ideas?
> 
>> 
>> Paul.
>> 
>>>> Hi,
>>>> Please review this simple fix to require that the jar manifest Multi-Release attribute has a value of “true" in order to be effective, i.e. to assert the jar is a multi-release jar.
>>>> issue: https://bugs.openjdk.java.net/browse/JDK-8153213 <https://bugs.openjdk.java.net/browse/JDK-8153213>
>>>> webrev: http://cr.openjdk.java.net/~sdrach/8153213/webrev/index.html <http://cr.openjdk.java.net/~sdrach/8153213/webrev/index.html>
>>>> Thanks
>>>> Steve
>> 
>