Subj : Re: JS_EvaluateScriptForPrincipals() vs JS_EvaluateScript()
To   : Oystein Hallaraker
From : Brendan Eich
Date : Wed Jun 30 2004 08:26 pm

Oystein Hallaraker wrote:

> that's a possibility... but I really want to understand what is going on!

http://lxr.mozilla.org/mozilla/ is your friend.

> This is what I believe is going on between spidermonkey and the dom, 
> please correct me where I am wrong (I assume that will be alot...):
> 
> The dom will first initialize a runtime,

Actually, xpconnect does that: 
http://lxr.mozilla.org/mozilla/source/js/src/xpconnect/src/xpcruntimesvc.cpp#106.

> the global object and the 
> standard ecma objects (date,boolean etc). But when are all the dom 
> objects initialized in spiderMonkey as JSObjects and XPCWrappedNative's? 
> Does this happen in the initialization process or when a script tries to 
> access a property that has not been initialized as a JSObject?

See 
http://lxr.mozilla.org/mozilla/source/dom/src/base/nsDOMClassInfo.cpp#4160 
among other hits for 
http://lxr.mozilla.org/mozilla/search?string=JS_ResolveStandardClass.

> As far as 
> I have understood  this, all dom objects will be registered as JSObjects 
> and XPCWrappedNative objects in spidermonkey,

You're mixing things.  JSObject is SpiderMonkey's type. 
XPCWrappedNative is XPConnect's type for a C++ type that has a JS object 
wrapping it.  They are not both registered "in spidermonkey".

> with callbacks back to the DOM.
> 
> I can see that the Principals are used in the DOM, but I cannot see 
> where the principals are used in spidermonkey... Maybe it's just me, but 
> I have been looking at that for weeks...

Did you look at JSPRINCIPALS_HOLD as I suggested? 
http://lxr.mozilla.org/mozilla/search?string=JSPRINCIPALS_HOLD, in 
particular http://lxr.mozilla.org/mozilla/source/js/src/jsscript.c#1104.

/be

.