[OpenJDK 2D-Dev] Fix proposal for 9001948: UnixPrintServiceLookup not returning consistent values

[Patrick Reinhart]

Hi Phil,

On Tue, 2013-04-23 at 09:56 -0700, Jennifer Godinez wrote:
> Hi Phil,
> 
> I haven't confirmed this.  I was about to ask for the program but now I 
> see it is included in the webrev.  I'll let you know.
> 
> Jennifer
> 
> On 4/23/2013 9:27 AM, Phil Race wrote:
> > That's a Java incident number, not a JDK bug number.
> > JDK numbers begin with "JDK-8", incident numbers begin with "JI-9"
> > I imagine jcheck will reject use of that number - it should !

This is not the case as I did run the test before and after my fix to
prove that it works. Thanks anyway for the hint.

> > I asked Jennifer to look into this right after you first posted.
> > Jennifer ?
> >
> > Off hand I'm not sure I like the proposed patch since it seems to ignore
> > the point of the existing comment at lines 359-362 ..

Good point! I missed that by the excitement about my fix ;-)  I did take
this in account now and updated my webrev accordingly.

I introduced a createPrintService() method that replaces 3 identical
code parts within the existing UnixPrintServiceLookup and used this
method now on the existing getNamedPrinterNameBSD() and
getNamedPrinterNameSysV() methods.

Cheers Patrick