Bug 12272

Summary: SVG picture is clipped (edges not visible)
Product: WebKit Reporter: Christopher Ramsey <christopher.ramsey>
Component: SVGAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Major Keywords: NeedsReduction
Priority: P2    
Version: 419.x   
Hardware: Mac   
OS: OS X 10.4   
Attachments:
Description Flags
SVG graph with X-axis label which is clipped
none
First attempt eric: review+

Christopher Ramsey
Reported 2007-01-14 15:05:47 PST
A page with an SVG image of defined size seems to be presented in a smaller viewport so that the edges of the image are clipped.
Attachments
SVG graph with X-axis label which is clipped (21.78 KB, image/svg+xml)
2007-01-14 15:08 PST, Christopher Ramsey 		no flags
Details
First attempt (33.13 KB, patch)
2007-01-15 11:55 PST, Rob Buis 		eric: review+
DetailsFormatted DiffDiff
View All Add attachment proposed patch, testcase, etc.
Christopher Ramsey
Comment 1 2007-01-14 15:08:18 PST
Created attachment 12434 [details] SVG graph with X-axis label which is clipped This image views fine in: Firefox Opera Adobe SVG viewer Adobe Illustrator In Web-Kit the edges are clipped. This seems to be true for all similar examples I have.
Eric Seidel (no email)
Comment 2 2007-01-14 21:07:31 PST
We'd have to reduce this. <svg> has implicit overflow:hidden according to the SVG spec. It's possible that no one else is respecting that overflow:hidden.
Eric Seidel (no email)
Comment 3 2007-01-15 00:24:04 PST
This seems to be a problem in our handling of viewbox when the SVG is given a size in pts instead of px. Not entirely sure yet.
Eric Seidel (no email)
Comment 4 2007-01-15 01:44:17 PST
rwlbuis says the problem here is that SVGLength uses 90 instead of 96 for px-per-inch
Rob Buis
Comment 5 2007-01-15 11:55:43 PST
Created attachment 12464 [details] First attempt Use the css dpi constant in svg, now we match Opera. Another testcase showing the problem: http://www.w3.org/TR/SVG11/images/script/script01.svg Cheers, Rob.
Eric Seidel (no email)
Comment 6 2007-01-15 23:50:40 PST
Comment on attachment 12464 [details] First attempt Your new test is kinda strange, and probably not needed given the other test changes. Looks like a great fix!
Rob Buis
Comment 7 2007-01-16 00:08:10 PST
Landed in r18882
Note You need to log in before you can comment on or make changes to this bug.