[cairo] SVG test suite improvements

Carl Worth cworth at cworth.org
Wed Feb 20 02:17:42 PST 2008


I've mentioned recently that I've been pretty discouraged about the
sad shape of our test suite results.

For example, for a long time the SVG backend was always passing the
test suite perfectly, (librsvg does a better job of pushing as much as
possible into cairo so it runs into fewer false negatives than poppler
for example). But recently the SVG backend has had more than a dozen
failing test cases.

Well, I just noticed that the bulk of those were failing only in the
-similar tests, (which are relatively new), and it was precisely the
set of tests that have svg-specific reference images. What was
happening was that the SVG backend was rasterizing the results in each
of these cases and just embedding an image. Fixing this was extremely
trivial thanks to the FLATTEN_TRANSPARENCY support that Adrian had
done---I simply needed to turn it on within cairo-svg-surface.c.

So I've done that now[*], (just before the snapshot which I'm really
about to make---honest), and now the SVG backend is only failing 3
tests, (hurrah!). If someone wants to look closer at these three
failures with cairo-svg that would be quite appreciated:

	a8-mask
	extend-reflect-similar
	mask-alpha

Besides cleaning up the test suite, this should also help avoid
excessive rasterization in the SVG backend, so this is a very good
change.

Cairo 1.6 is feeling closer all the time. And I'm quite happy about
that.

I hope you're all having as much fun as I am.

-Carl

[*] Locally, so far, but I haven't quite pushed just yet. I will
before I go to sleep tonight.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.cairographics.org/archives/cairo/attachments/20080220/3067fddf/attachment.pgp 


More information about the cairo mailing list