When we reach this code everything in the job is already
configured by a
combination of initial settings and user updates and and we just
need to read
the settings and pass it on to the native NSPrintInfo.
So surely switching the order should not matter unless one of these
is using the 'wrong' PageFormat ?
-----------------
the body of the method called here :-
javaPrinterJobToNSPrintInfo(env, jthis, pageable, printInfo);
gets its PageFormat as follows :-
static JNF_MEMBER_CACHE(jm_getPageFormat, sjc_CPrinterJob,
"getPageFormatFromAttributes", "()Ljava/awt/print/PageFormat;");
....
jobject page = JNFCallObjectMethod(env, srcPrinterJob,
jm_getPageFormat);
if (page != NULL) {
javaPageFormatToNSPrintInfo(env, NULL, page, dst);
}
So this uses the result of making a call to
RasterPrinterJob.getPageFormatFromAttributes()
protected PageFormat getPageFormatFromAttributes() {
if (attributes == null) {
return null;
}
return attributeToPageFormat(getPrintService(),
this.attributes);
}
-----------------------------
whereas
javaPageFormatToNSPrintInfo(env, jthis, page, printInfo);
is using a PageFormat obtained as follows :-
static JNF_MEMBER_CACHE(jm_getPageFormat, sjc_CPrinterJob,
"getPageFormat", "(I)Ljava/awt/print/PageFormat;");
jobject page = JNFCallObjectMethod(env, jthis,
jm_getPageFormat, 0);
This is CPrinterJob.getPageFormat() { .. }
Although its not easily apparent what the returned values are in
each of these cases
it does seem they must be different.
The latter one appears 'correct' in this case since applying it
second
fixes the output but I don't have enough information to know why the
values differ.
Looking at the fix for 8011069, it avoided an NPE by always
creating an 'attributes' map, albeit an empty one.
This can change the result of calling getPageFormatFromAttributes
from
'null' to a PageFormat built from an empty attribute set.
If the no-args native printDialog() and the no-args print() call
is used this will be empty.
So the method will indeed build - at that moment - a page format
built from
default values.
Now. If we *do* use the printDialog(PrintRequestAttributeSet) and
print(PrintRequestAttributeSet) methods, then it may well be that
this
method is the one that should be called.
And I think we were previously only in this block of code if that
were the case
by virtue of the block being guarded by "if (page != NULL)", which
means
there is an attributeset, which previously meant one of those
"with args"
methods had been used.
So I wonder/suspect if the switching of the order will introduce
the equivalent
problem in that 'with args' case.
As you can tell just looking at the webrev its nigh on impossible
to tell
for sure and you'd probably need to play around with testing changing
paper size and orientation in native and cross-platform dialogs to
test it.
You could start by seeing if the test 'passes' simply by switching to
'with args' before & after your fix - ensuring that the same paper
sizes
are being used. I am not sure what the default settings were that
were
created for the empty attribute set vs the ones that are used when
you
fixed this. You'll have to tell me that.
Perhaps what is needed is a unified call to get the PageFormat which
can figure out whether to use the attributes or not. And you could
check if the call to CPrinterJob.getPageFormat() already performs
that ..
-phil.
On 10/28/2014 01:03 PM, Alexander Scherbatiy wrote:
Hello,
Could you review the fix?
Thanks,
Alexandr.
On 7/15/2014 3:28 PM, Alexander Scherbatiy wrote:
Hello,
Could you review the fix:
bug: https://bugs.openjdk.java.net/browse/JDK-8044444
webrev: http://cr.openjdk.java.net/~alexsch/8044444/webrev.00
Native method printLoop from CPrinterJob calls
javaPrinterJobToNSPrintInfo(...) method after
javaPageFormatToNSPrintInfo(...).
Both methods set the page size. The initial page size is set
in defaultPage(PageFormat) method.
After the fix 8011069 the printDialog() initializes attributes
which leads that new page size is created in the
attributeToPageFormat(PrintService, PrintRequestAttributeSet)
method.
The fix changes order of the javaPrinterJobToNSPrintInfo(...)
javaPageFormatToNSPrintInfo(...) call so initial page size is
set at the end.
There is the JCK test that covers the issue.
Thanks,
Alexandr.
