blob: 7d6eb44db73a947306a0d7b31776d54b0d24d4c6 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
|
<svg id="svg-root" width="100%" height="100%"
viewBox="0 0 480 360" xmlns="http://www.w3.org/2000/svg"
xmlns:xlink="http://www.w3.org/1999/xlink">
<!-- The test checks if edgeMode wrapping wraps around the input surface bounds,
or the input surface bounds clipped to the primitive subregion.
The spec doesn't specify anything, Chromium uses the latter, and Firefox uses the former.
Since both are wrapping the input surface clipped to the filter effects region,
wrapping while obeying clipping to the primitive subregion seems like the better behavior,
so that's what librsvg does. -->
<filter id="primitive-bounds-narrower">
<feConvolveMatrix kernelUnitLength="1" x="0" y="0" width="25" height="25" edgeMode="wrap" order="5" targetX="0" targetY="0" kernelMatrix="1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0"/>
</filter>
<!-- Same as above, but the primitive subregion is now larger than the input surface.
Both Chromium and Firefox don't tile the input surface in this case. -->
<filter id="primitive-bounds-larger">
<feFlood x="0" y="0" width="30" height="30" flood-color="red" result="f1"/>
<feFlood x="5" y="5" width="20" height="20" flood-color="green" result="f2"/>
<feMerge>
<feMergeNode in="f1"/>
<feMergeNode in="f2"/>
</feMerge>
<feConvolveMatrix kernelUnitLength="1" x="0" y="0" width="50" height="50" edgeMode="wrap" order="5" targetX="0" targetY="0" kernelMatrix="1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0"/>
</filter>
<g transform="translate(30 20)" filter="url(#primitive-bounds-narrower)">
<rect fill="red" x="1" y="1" width="50" height="50"/>
<rect fill="green" x="6" y="6" width="40" height="40"/>
</g>
<g transform="translate(90 20)" filter="url(#primitive-bounds-larger)">
<rect x="1" y="1" width="50" height="50"/>
</g>
</svg>
|
