315061 - Donat Polygons in Mozilla SVG

Status: RESOLVED FIXED

(Core :: SVG, defect)

Description

[Ruth Lang]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.8) Gecko/20051025 Firefox/1.5
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.8) Gecko/20051025 Firefox/1.5

Attached a svg file which shows that donat polygons are not displayed correctly. Please have a look at the same file with the Internet Explorer and the Adobe SVG Viewer. There the yellow polygon covers the white "donut"
Thank you for your help
Ruth

<?xml version="1.0" encoding="iso-8859-1" standalone="no"?>
<svg  id="map" x="0" y="0" width="1000" height="888"  viewBox="234.003 147.272 356.055 316.428" preserveAspectRatio="xMinYMin" xmlns="http://www.w3.org/2000/svg"  version="1.1" baseProfile="full">
<path id="th0rec55" fill="blue" d="M463.013 281.513l3.383 -1.354l3.617 6.354l-0.500 6.500l-2.000 5.500l-4.000 12.500l-0.500 1.000l-2.000 5.500l-1.000 4.000l-1.000 3.500l-0.500 5.000l0.500 6.000l2.000 5.500l5.000 4.500l6.500 4.000l8.500 6.500l10.000 5.000l7.000 4.500l4.000 3.000l2.500 1.500l5.500 3.000h5.000l4.500 -1.500l3.000 -2.500l3.500 -4.000l6.500 2.000l-18.000 23.500l-2.000 -3.500l-3.000 -6.000l-5.500 -4.000l-6.500 -4.500l-6.000 -4.000l-1.390 -1.057l-14.054 -8.362l-10.572 -7.058l-19.945 -13.316l-20.353 -8.362l-9.273 -2.406l-36.575 -17.239l-0.328 -0.223l-35.713 -17.052l-14.797 -5.425l-2.246 -0.658l-5.238 -1.535l-3.167 -0.273l-4.763 0.966l-10.147 7.560l-15.231 16.045l-3.421 3.193l0.271 0.125l-1.254 1.298l-3.808 3.942
l-15.049 15.600l-1.005 0.341l-5.910 5.503l1.968 -3.607l-2.674 -1.337l14.795 -14.499l8.721 -9.180l4.088 -4.303l11.123 -12.471l4.361 -3.894l5.077 -4.533l6.067 -4.045l6.404 -1.348l5.056 -0.337l6.067 1.685l16.815 6.540l3.100 1.722l7.500 3.000l31.259 17.104l20.885 10.070l13.856 4.826l8.000 1.000l8.000 -2.500l6.000 -4.500l5.000 -7.000l9.000 -11.500l7.500 -15.500 z
m-3.125 25.874l-4.935 7.403l-6.267 7.585l-4.947 4.617l-8.245 3.298l5.277 1.979l6.926 3.298l5.280 3.443l-0.464 -6.497l0.500 -6.000l1.500 -7.500l4.500 -9.500 z" />
<path id="th0rec63" style="fill:rgb(255,215,0)" d="M459.888 307.387l-0.875 2.126l-4.500 9.500l-1.500 7.500l-0.500 6.000l0.464 6.497l-5.280 -3.443l-6.926 -3.298l-5.277 -1.979l8.245 -3.298l4.947 -4.617l6.267 -7.585 z" />
</svg>


Reproducible: Always

Comment 1

[PikeUK]

->Core:SVG

Comment 2

[Jonathan Watt [:jwatt]]

Attachment: ruth's file

Comment 3

[Ben Fowler]

The render in netscape looks the same as Firefox. Do you have a png of the
correct render?

Comment 4

[Ben Fowler]

(In reply to comment #3)
> The render in netscape looks the same as Firefox. Do you have a png of the
> correct render?

That should be INKSCAPE http://www.inkscape.org/ ! Inkscape! I don't use
Netscape, sorry.

Comment 5

[Jonathan Watt [:jwatt]]

Ben, if you look at Ruth's testcase with latest branch (try RC1) you'll see what's wrong.

Comment 6

[Ben Fowler]

(In reply to comment #5)
> Ben, if you look at Ruth's testcase with latest branch (try RC1) you'll see
> what's wrong.

Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.9a1) Gecko/20051104 Firefox/1.6a1

Fair do's I am using a trunk build only a few minutes old. Are you suggesting that
that trunk is OK, but the branch is not ...

Comment 7

[Ben Fowler]

That's it. The yellow segment is about 15 pixels to the left and above where 
it should be.

Comment 8

[Jonathan Watt [:jwatt]]

Yup. They've diverged considerably, and in the case of SVG, the branch uses cairo whereas the trunk still uses GDI+. Matching your build with the Version field of bugs is much more important than being up to date to the minute. ;-)

Comment 9

[:Gavin Sharp [email: [e-mail]]]

(In reply to comment #8)
> Yup. They've diverged considerably, and in the case of SVG, the branch uses
> cairo whereas the trunk still uses GDI+.

Is that still true, given bug 310957? He's also using a Mac build, so that shouldn't matter, right?

Comment 10

[tor]

Attachment: nsSVGCairoPathBuilder::ClosePath wasn't returning new point

Comment 11

[Jonathan Watt [:jwatt]]

(In reply to comment #9)
> (In reply to comment #8)
> > Yup. They've diverged considerably, and in the case of SVG, the branch uses
> > cairo whereas the trunk still uses GDI+.
> 
> Is that still true, given bug 310957? He's also using a Mac build, so that
> shouldn't matter, right?

To be clearer about the state of GDI+ vs cairo: cairo is the default on all platforms, but some of the w32 trunk tinderbox still build with GDI+ because their version of cairo hasn't been updated yet. Anyway, my point (badly made I admit) was that trunk and branch are different, and testing should be against the bug's Version. I didn't mean to say cairo was or wasn't the source, even though that seems to be what tor has found. Sorry about that.

Comment 12

[tor]

Checked in on trunk.