1039167 - SVG path partially clipped unnecessarily

Status: RESOLVED INVALID

(Core :: SVG, defect)

Description

[Giuseppe Bilotta]

User Agent: Opera/9.80 (X11; Linux x86_64) Presto/2.12.388 Version/12.16

Steps to reproduce:

View http://wok.oblomov.eu/ars/elian/#compositionandshaping and scroll down to find the image captioned “Oblomov in Elian, try #4: kerning #2”. Equivalently, just open the SVG file http://wok.oblomov.eu/ars/elian/oblomov-kern-2.svg

(This shows that the misrendering happens both if the SVG is viewed stand-alone and if it's viewed as included as an <img> in a web page.)


Actual results:

The SVG is misrendered: the horizontal bar of the bottom right path is not rendered, for no apparent reason


Expected results:

The path is correctly rendered in Opera/Presto as well as by rsvg-view-3 and by ImageMagick's display command.

Comment 1

[Jonathan Watt [:jwatt]]

Presto is dead, so I'm not sure we care to base any decision on that. Chrome and Firefox seem to render the SVG in the same way. Is there some spec. based justification for arguing that Firefox/Chrome are rendering it incorrectly? (I've not looked at the file's source FWIW.)

Comment 2

[Robert Longson [:longsonr]]

Attachment: reduced testcase

Comment 3

[Robert Longson [:longsonr]]

The key thing is that it's <use> pointing to a <symbol> which means that the <symbol> is replaced by an <svg> element. The content overflows that <svg> element and we clip it to the generated <svg>'s area.

If you put overflow="visible" on the symbol then it displays.

The final paragraph of http://www.w3.org/TR/SVG/masking.html#OverflowProperty says that visibility:hidden is the default and this bug is therefore invalid.

Comment 4

[Giuseppe Bilotta]

(In reply to Robert Longson from comment #3)
> The key thing is that it's <use> pointing to a <symbol> which means that the
> <symbol> is replaced by an <svg> element. The content overflows that <svg>
> element and we clip it to the generated <svg>'s area.
> 
> If you put overflow="visible" on the symbol then it displays.
> 
> The final paragraph of
> http://www.w3.org/TR/SVG/masking.html#OverflowProperty says that
> visibility:hidden is the default and this bug is therefore invalid.

Ah, thanks, I had missed that. Looks like the discrepancy is actually a bug is in librsvg then instead. I'll report it there.