Hi Peter,
"Peter" <[EMAIL PROTECTED]> wrote on 10/25/2006 05:23:04 PM:
> I got a few more 1.3 VM crashes and therefore commented out a few tests
(as below)
>
> <!-- <test id="samples/tests/spec/fonts/fontArabic.svg" /> -->
> <!-- <test id="samples/tests/spec/fonts/fontGlyphsBoth.svg" />-->
> <!-- <test id="samples/tests/spec/fonts/fontGlyphsChildSVG.svg" />
-->
> <!-- <test id="samples/tests/spec/fonts/fontGlyphsD.svg" /> -->
My guess is a corrupt font on your system. For these
tests Batik probably starts trolling all of the installed
fonts looking for one what supports Arabic and/or other
strange chars.
> After that I still see 33 differences
>
> 1. samples/textRotate.svg
> 2. samples/textRotateShadows.svg
> 3. samples/tests/spec12/text/flowBidi.svg
> 4. samples/tests/spec12/text/flowText.svg
> 5. samples/tests/spec12/text/flowText2.svg
> 6. samples/tests/spec12/text/flowText3.svg
> 7. samples/tests/spec12/text/flowText4.svg
> 8. samples/tests/spec12/text/flowText5.svg
> 9. samples/extensions/flowText.svg
> 10. samples/tests/spec/text/verticalText.svg
> 11. samples/tests/spec/scripting/textAllProperties.svg
> 12. samples/tests/spec/scripting/textpathProperties.svg
> 13. samples/tests/spec/scripting/text_content.svg
> 14. shapes-rect-BE-01.svg
> 15. text-i18n-BE-09.svg
This one requires a simple change to the test in the
test-suite.
> Most of these are rendering differences, except
It is common to see rendering differences in text rendering
on different systems (slightly different fonts/font metrics etc).
Just take a close look at the difference images in regard
and make sure that the differences look reasonable in this
context.
> - text-i18n-BE-09.svg which fais with
>
> The content of element type "g" must match
This is a bug in the SVG test IIRC.
> - DoubleStringPerformanceTest, PerformanceTestValidator
> ==> (unexpected.performance.improvement)
This test fails quite often. I've a mind to simply
pull it since it also takes a while to run...
> -
org.apache.batik.test.svg.SelfContainedSVGOnLoadTestValidator$ProcessingError
> SelfContainedSVGOnLoadTestValidator.entry.error.code
>
SelfContainedSVGOnLoadTest.error.unexpected.number.of.test.result.elements
> SelfContainedSVGOnLoadTestValidator.entry.expected.error.code
> SelfContainedSVGOnLoadTest.error.while.processing.svg.document
>
> - bridge/error/ellipse-missing-rx, bridge/error/ellipse-missing-ry,
> bridge/error/rect-missing-height, bridge/error/rect-missing-width
>
> DO not return the expected exception
These are a bit worrying, I wonder if this is some subtle (or
not so subtle) error in class path.
> - TranscoderInput fails with an NPE
> java.lang.NullPointerException at org.apache.batik.dom.util.
> SAXDocumentFactory.startElement(SAXDocumentFactory.java:562) at
org.apache.
> xerces.parsers.AbstractSAXParser.startElement(Unknown Source) at
org.apache.
> xerces.impl.dtd.XMLDTDValidator.startElement(Unknown Source) at
This might be a bug in Batik (but probably not your stuff).
The failure is in the code:
try {
isStandalone = parser.getFeature
("http://xml.org/sax/features/is-standalone";);
} catch (SAXNotRecognizedException ex) {
}
try {
xmlVersion = (String) parser.getProperty
("http://xml.org/sax/properties/document-xml-version";);
} catch (SAXNotRecognizedException ex) {
}
Cameron what is the purpose of this code? Can we check if parser
is null and provide some default? Can you guess why parser would be
null? This isn't the first time the code has failed here.
> Anyone any thoughts? Could any of these be related to changes I made
(minor
> changes wrt color handling)
The only one's that might be caused by your changes are the rendering
ones, but I suspect those are font related.
> > -----Original Message-----
> > From: Peter [mailto:[EMAIL PROTECTED]
> > Sent: Wednesday, October 25, 2006 10:02 PM
> > To: '[email protected]'
> > Subject: RE: Running regard
> >
> > When I try to run regard with jdk1.3 on XP
> >
> > java version "1.3.1_19"
> > Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.1_19-b03)
> > Java HotSpot(TM) Client VM (build 1.3.1_19-b03, mixed mode)
> >
> >
> > I get a vm crash
> >
> > ...
> > Running samples/tests/spec/fonts/fontArabic.svg
> >
> > An unexpected exception has been detected in native code outside the
VM.
> > Unexpected Signal : EXCEPTION_ACCESS_VIOLATION occurred at
PC=0x6D1E8E26
> > Function name=(N/A)
> > Library=C:\Program Files\Java\jdk1.3.1_19\jre\bin\fontmanager.dll
> >
> > NOTE: We are unable to locate the function name symbol for the error
> > just occurred. Please refer to release documentation for
possible
> > reason and solutions.
> >
> >
> > Anybody any ideas?
> >
> > Thanks,
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
> > > Sent: Tuesday, October 24, 2006 2:57 PM
> > > To: [email protected]
> > > Cc: [email protected]
> > > Subject: RE: Running regard
> > >
> > > Hi Peter,
> > >
> > > "Peter" <[EMAIL PROTECTED]> wrote on 10/19/2006 04:37:41
PM:
> > >
> > > > So the regard run eventually finished and a *lot* of the tests
fail
> > with
> > >
> > >
> > > > I have used 1.4 for the test run but at the same time the message
is
> > > (very)
> > > > old so I am not sure how relevant it still is.
> > >
> > > It is still relevant. Batik still targets JDK 1.3 for support.
So
> > our
> > > testing is primarily done with that JDK. It is possible to
establish a
> > > baseline
> > > rendering on JDK 1.4 and use that for your testing if you want to.
> > >
> > > > Could someone clarify or should I start worrying I have introduced
a
> > > regression?
> > >
> > > Most likely you did not, but it would be best to either test
under
> > JDK
> > > 1.3
> > > (my preference) or at least establish a baseline JDK 1.4 rendering
and
> > > then
> > > test your changes against that baseline (still won't pick up JDK
> > 1.4isms).
> > >
> > > >
> > > > Thanks,
> > > >
> > > > Peter
> > > >
> > > > > -----Original Message-----
> > > > > From: Peter [mailto:[EMAIL PROTECTED]
> > > > > Sent: Thursday, October 19, 2006 9:22 PM
> > > > > To: '[email protected]'
> > > > > Subject: RE: Running regard
> > > > >
> > > > > Cameron,
> > > > >
> > > > > Your tip did it. I have not intentionally added these
dependencies,
> > > but I
> > > > > did rebuild the pdf transcoder jar file and I probably added too
> > much
> > > in
> > > > > there while I was fighting with the fop/batik dependency.
> > > > >
> > > > > Btw, I did find a remark in the pdf transcoder readme file in
the
> > > batik
> > > > > trunk saying that there is a dependency on avaolon
> > > > >
> > > > > <quote>
> > > > > The pdf-transcoder.jar file from the Apache Fop project
> > > > > (http://xmlgraphics.apache.org/fop), this is only needed if you
want
> > > to
> > > > > transcode to PDF, otherwise it can be removed from
distributions.
> > > > >
> > > > > This jar file is slightly modified from the standard
> > > > > pdf-transcoder.jar to also includes two required classes from
the
> > > > > Apache Avalon project (http://avalon.apache.org):
> > > > > org.apache.avalon.framework.Enum and
> > > > > org.apache.avalon.framework.ValuedEnum.
> > > > > </quote>
> > > > >
> > > > > Anyway, tests are running now (the fan on my laptop says so :))
> > > > >
> > > > > Thanks!
> > > > >
> > > > > Peter
> > > > >
> > > > > > -----Original Message-----
> > > > > > From: Cameron McCormack [mailto:[EMAIL PROTECTED]
> > > > > > Sent: Thursday, October 19, 2006 12:47 AM
> > > > > > To: [email protected]
> > > > > > Subject: Re: Running regard
> > > > > >
> > > > > > Peter:
> > > > > > > If I do that I get something else (see lower) Have not
looked
> > into
> > > it
> > > > > > > so perhaps the solution is simple...but it is late here :)
and
> > > someone
> > > > > > > else might just know the solution without having to look for
> > > it....
> > > > > > >
> > > > > > > Note I am running the trunk with a few changes.
> > > > > >
> > > > > > Do your changes use any classes from Avalon? Batik doesn't
use
> > it,
> > > so
> > > > > > the classpath that's set up by the ant 'regard' target
wouldn't
> > > include
> > > > > > it. If you need to add it, stick its jar in the lib
directory,
> > and
> > > > > > it'll be picked up automatically.
> > > > > >
> > > > > > --
> > > > > > Cameron McCormack, http://mcc.id.au/
> > > > > > xmpp:[EMAIL PROTECTED] ▪ ICQ 26955922 ▪ MSN
[EMAIL PROTECTED]
> > > > > >
> > > > > >
> > >
---------------------------------------------------------------------
> > > > > > To unsubscribe, e-mail: batik-dev-
> > [EMAIL PROTECTED]
> > > > > > For additional commands, e-mail:
> > > [EMAIL PROTECTED]
> > > >
> > > >
> > > >
---------------------------------------------------------------------
> > > > To unsubscribe, e-mail:
[EMAIL PROTECTED]
> > > > For additional commands, e-mail:
[EMAIL PROTECTED]
> > > >
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
