Mack Stevenson wrote:
> Would it be possible for anyone to shortly describe the mechanism(s) by
> which gecko can be "activated"/"called" by an application?
http://www.mozilla.org/projects/embedding/embeddingOverview.html
> The use of
> gtk/MFC in the demos ends up obscuring the underlying mechanism a little; I
> am still wondering how I can pass gecko a pointer to a mem area containing
> the HTML code it should display,
You can't pass a raw pointer like that. if your html code is flat (ie. no
embedded content). you can use nsIDocShell (QIable off of nsIWebBrowser) to do a
LoadStream() after you wrap your content in a stream interface.
> not to mention how one can "tell" gecko
> what actual lower-level drawing/text-rendering functions to use; does one
> just pass function pointers to gecko?
we don't support this.
> My personal naive representation of gecko's interface is something like this
>
> geckoInit(int (*fpPutPixel)(int x, int y, struct_color color, alpha), int
> (*fpRenderText) (const char *text, int x, int y, const char *font, int
> size));
> geckoRenderPage(const char *HTML, .. /*pointers to CSS/images files*/);
> geckoClose();
>
> I would expect that, by writing the PutPixel and RenderText functions and
> passing them as pointers to gecko I would be giving it the necessary tools
> to render a page when I invoked geckoRenderPage(); am I completely off
> target?
You're too deep here. We don't go this low in the embedding context.
Jud
