[whatwg] Canvas 2d: proposal to add path(DOMString)

[Juriy Zaytsev]

I'd love to see something like this as well.

In my experience this takes more than 80 lines. In Fabric.js [1], for
example, `fabric.Path#render` has ~180 LOC chunk [2] dedicated to doing
exactly this — drawing translated-from-SVG path using a combination of
`lineTo`, `moveTo`, `bezierCurveTo`, etc. And that's not counting another
~100 lines of arc normalization [3] where arc command is translated to a
sequence of `bezierCurveTo` commands (due to SVG's a/A algorithms being so
different from canvas' `arc` one).

`HTMLRenderingContext::path` seems sufficient since it's always possible to
end it with `fill` or `stroke` (the same way you would do it with
combination of path commands followed by `fill` or `stroke`).

[1] http://kangax.github.com/fabric.js/demos/kitchensink/
[2]
https://github.com/kangax/fabric.js/blob/gh-pages/src/path.class.js#L247-433
[3]
https://github.com/kangax/fabric.js/blob/gh-pages/src/path.class.js#L17-121

-- 
kangax


On Fri, Jul 29, 2011 at 4:44 PM, Charles Pritchard <chuck at jumis.com> wrote:

> Having spoken to several developers, I think that we're clear to add a new
> method to the canvas 2d api;
> though there is some question about how it should be added.
>
> The method would take a DOMString using SVG path semantics.
>
> I'm not certain which route is better:
> ctx.path('M....z')
> or
> ctx.fillPath('M ..z'), ctx.strokePath('M...z');
>
> This usually takes ~80 lines of JavaScript to support,
> we're eager to see it added to native implementations.
>
> All browser vendors have SVG support now, and could easily re-use
> code from their SVG implementations to implement this canvas 2d
> supplemental.
>
>
> -Charles
>