gecko-dev/js/xpconnect/idl/nsIXPConnect.idl

694 lines
27 KiB
Plaintext
Raw Normal View History

/* -*- Mode: IDL; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4 -*-
*
2012-05-21 11:12:37 +00:00
* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
/* The core XPConnect public interfaces. */
#include "nsISupports.idl"
#include "nsIClassInfo.idl"
#include "xpccomponents.idl"
#include "xpcjsid.idl"
#include "xpcexception.idl"
#include "nsIInterfaceInfo.idl"
#include "nsIInterfaceInfoManager.idl"
#include "nsIExceptionService.idl"
#include "nsIVariant.idl"
#include "nsIObjectOutputStream.idl"
#include "nsIObjectInputStream.idl"
%{ C++
#include "jspubtd.h"
#include "xptinfo.h"
#include "nsAXPCNativeCallContext.h"
class nsWrapperCache;
%}
/***************************************************************************/
2010-07-15 06:19:36 +00:00
// NB: jsval and jsid are declared in nsIVariant.idl
[ptr] native JSContextPtr(JSContext);
[ptr] native JSClassPtr(JSClass);
[ptr] native JSFreeOpPtr(JSFreeOp);
[ptr] native JSObjectPtr(JSObject);
[ptr] native JSValPtr(jsval);
2010-07-15 06:19:36 +00:00
[ptr] native JSValConstPtr(const jsval);
native JSPropertyOp(JSPropertyOp);
native JSEqualityOp(JSEqualityOp);
[ptr] native JSScriptPtr(JSScript);
[ptr] native voidPtrPtr(void*);
[ptr] native nsScriptObjectTracerPtr(nsScriptObjectTracer);
[ref] native nsCCTraversalCallbackRef(nsCycleCollectionTraversalCallback);
[ptr] native nsAXPCNativeCallContextPtr(nsAXPCNativeCallContext);
[ptr] native nsWrapperCachePtr(nsWrapperCache);
/***************************************************************************/
// forward declarations...
interface nsIXPCScriptable;
interface nsIXPConnect;
interface nsIXPConnectWrappedNative;
interface nsIInterfaceInfo;
interface nsIXPCSecurityManager;
interface nsIPrincipal;
%{C++
class nsCycleCollectionTraversalCallback;
class nsScriptObjectTracer;
%}
/***************************************************************************/
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
[uuid(8916a320-d118-11d3-8f3a-0010a4e73d9a)]
interface nsIXPConnectJSObjectHolder : nsISupports
{
readonly attribute JSObjectPtr JSObject;
};
[uuid(f819a95a-6ab5-4a02-bda6-32861e859581)]
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
interface nsIXPConnectWrappedNative : nsIXPConnectJSObjectHolder
{
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/* attribute 'JSObject' inherited from nsIXPConnectJSObjectHolder */
readonly attribute nsISupports Native;
readonly attribute JSObjectPtr JSObjectPrototype;
/**
* These are here as an aid to nsIXPCScriptable implementors
*/
readonly attribute nsIXPConnect XPConnect;
2010-07-15 06:19:36 +00:00
nsIInterfaceInfo FindInterfaceWithMember(in jsid nameID);
nsIInterfaceInfo FindInterfaceWithName(in jsid nameID);
void debugDump(in short depth);
/*
* This finishes initializing a wrapped global, doing the parts that we
* couldn't do while the global and window were being simultaneously
* bootstrapped. This should be called exactly once, and only for wrapped
* globals.
*/
void finishInitForWrappedGlobal();
/*
* This returns a pointer into the instance and care should be taken
* to make sure the pointer is not kept past the life time of the
* object it points into.
*/
voidPtrPtr GetSecurityInfoAddress();
/*
* NOTE: Add new IDL methods _before_ the C++ block below if you
* add them. Otherwise the vtable won't be what xpidl thinks it
* is, since GetObjectPrincipal() is virtual.
*/
%{C++
/**
* Faster access to the native object from C++. Will never return null.
*/
nsISupports* Native() const { return mIdentity; }
protected:
nsISupports *mIdentity;
public:
%}
};
%{C++
#include "nsCOMPtr.h"
inline
const nsQueryInterface
do_QueryWrappedNative(nsIXPConnectWrappedNative *aWrappedNative)
{
return nsQueryInterface(aWrappedNative->Native());
}
inline
const nsQueryInterfaceWithError
do_QueryWrappedNative(nsIXPConnectWrappedNative *aWrappedNative,
nsresult *aError)
{
return nsQueryInterfaceWithError(aWrappedNative->Native(), aError);
}
%}
[uuid(BED52030-BCA6-11d2-BA79-00805F8A5DD7)]
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
interface nsIXPConnectWrappedJS : nsIXPConnectJSObjectHolder
{
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/* attribute 'JSObject' inherited from nsIXPConnectJSObjectHolder */
readonly attribute nsIInterfaceInfo InterfaceInfo;
readonly attribute nsIIDPtr InterfaceIID;
void debugDump(in short depth);
void aggregatedQueryInterface(in nsIIDRef uuid,
[iid_is(uuid),retval] out nsQIResult result);
};
/***************************************************************************/
/**
* This is a sort of a placeholder interface. It is not intended to be
* implemented. It exists to give the nsIXPCSecurityManager an iid on
* which to gate a specific activity in XPConnect.
*
* That activity is...
*
* When JavaScript code uses a component that is itself implemented in
* JavaScript then XPConnect will build a wrapper rather than directly
* expose the JSObject of the component. This allows components implemented
* in JavaScript to 'look' just like any other xpcom component (from the
* perspective of the JavaScript caller). This insulates the component from
* the caller and hides any properties or methods that are not part of the
* interface as declared in xpidl. Usually this is a good thing.
*
* However, in some cases it is useful to allow the JS caller access to the
* JS component's underlying implementation. In order to facilitate this
* XPConnect supports the 'wrappedJSObject' property. The caller code can do:
*
* // 'foo' is some xpcom component (that might be implemented in JS).
* try {
* var bar = foo.wrappedJSObject;
* if(bar) {
* // bar is the underlying JSObject. Do stuff with it here.
* }
* } catch(e) {
* // security exception?
* }
*
* Recall that 'foo' above is an XPConnect wrapper, not the underlying JS
* object. The property get "foo.wrappedJSObject" will only succeed if three
* conditions are met:
*
* 1) 'foo' really is an XPConnect wrapper around a JSObject.
* 2) The underlying JSObject actually implements a "wrappedJSObject"
* property that returns a JSObject. This is called by XPConnect. This
* restriction allows wrapped objects to only allow access to the underlying
* JSObject if they choose to do so. Ususally this just means that 'foo'
* would have a property tht looks like:
* this.wrappedJSObject = this.
* 3) The implemementation of nsIXPCSecurityManager (if installed) allows
* a property get on the interface below. Although the JSObject need not
* implement 'nsIXPCWrappedJSObjectGetter', XPConnect will ask the
* security manager if it is OK for the caller to access the only method
* in nsIXPCWrappedJSObjectGetter before allowing the activity. This fits
* in with the security manager paradigm and makes control over accessing
* the property on this interface the control factor for getting the
* underlying wrapped JSObject of a JS component from JS code.
*
* Notes:
*
* a) If 'foo' above were the underlying JSObject and not a wrapper at all,
* then this all just works and XPConnect is not part of the picture at all.
* b) One might ask why 'foo' should not just implement an interface through
* which callers might get at the underlying object. There are three reasons:
* i) XPConnect would still have to do magic since JSObject is not a
* scriptable type.
* ii) JS Components might use aggregation (like C++ objects) and have
* different JSObjects for different interfaces 'within' an aggregate
* object. But, using an additional interface only allows returning one
* underlying JSObject. However, this allows for the possibility that
* each of the aggregte JSObjects could return something different.
* Note that one might do: this.wrappedJSObject = someOtherObject;
* iii) Avoiding the explicit interface makes it easier for both the caller
* and the component.
*
* Anyway, some future implementation of nsIXPCSecurityManager might want
* do special processing on 'nsIXPCSecurityManager::CanGetProperty' when
* the interface id is that of nsIXPCWrappedJSObjectGetter.
*/
[scriptable, uuid(254bb2e0-6439-11d4-8fe0-0010a4e73d9a)]
interface nsIXPCWrappedJSObjectGetter : nsISupports
{
readonly attribute nsISupports neverCalled;
};
/***************************************************************************/
/*
* This interface is implemented by outside code and registered with xpconnect
* via nsIXPConnect::setFunctionThisTranslator.
*
* The reason this exists is to support calls to JavaScript event callbacks
* needed by the DOM via xpconnect from C++ code.
*
* We've added support for wrapping JS function objects as xpcom interfaces
* by declaring the given interface as a [function] interface. However, to
* support the requirements of JS event callbacks we need to call the JS
* function with the 'this' set as the JSObject for which the event is being
* fired; e.g. a form node.
*
* We've decided that for all cases we care about the appropriate 'this' object
* can be derived from the first param in the call to the callback. In the
* event handler case the first param is an event object.
*
* Though we can't change all the JS code so that it would setup its own 'this',
* we can add plugin 'helper' support to xpconnect. And that is what we have
* here.
*
* The idea is that at startup time some code that cares about this issue
* (e.g. the DOM helper code) can register a nsIXPCFunctionThisTranslator
* object with xpconnect to handle calls to [function] interfaces of a given
* iid. When xpconnect goes to invoke a method on a wrapped JSObject for
* an interface marked as [function], xpconnect will check if the first param
* of the method is an xpcom object pointer and if so it will check to see if a
* nsIXPCFunctionThisTranslator has been registered for the given iid of the
* interface being called. If so it will call the translator and get an
* interface pointer to use as the 'this' for the call. If the translator
* returns a non-null interface pointer (which it should then have addref'd
* since it is being returned as an out param), xpconnect will attempt to build
* a wrapper around the pointer and get a JSObject from that wrapper to use
* as the 'this' for the call.
*
* If a null interface pointer is returned then xpconnect will use the default
* 'this' - the same JSObject as the function object it is calling.
*
* The translator can also return a non-null aIIDOfResult to tell xpconnect what
* type of wrapper to build. If that is null then xpconnect will assume the
* wrapper should be for nsISupports. For objects that support flattening -
* i.e. expose nsIClassInfo and that interface's getInterfaces method - then
* a flattened wrapper will be created and no iid was really necessary.
*
* XXX aHideFirstParamFromJS is intended to allow the trimming of that first
* param (used to indicate 'this') from the actual call to the JS code. The JS
* DOM does not require this functionality and it is **NOT YET IMPLEMENTED**
*
*/
[uuid(039ef260-2a0d-11d5-90a7-0010a4e73d9a)]
interface nsIXPCFunctionThisTranslator : nsISupports
{
nsISupports TranslateThis(in nsISupports aInitialThis,
in nsIInterfaceInfo aInterfaceInfo,
in uint16_t aMethodIndex,
out boolean aHideFirstParamFromJS,
out nsIIDPtr aIIDOfResult);
};
/***************************************************************************/
%{ C++
// For use with the service manager
// {CB6593E0-F9B2-11d2-BDD6-000064657374}
#define NS_XPCONNECT_CID \
{ 0xcb6593e0, 0xf9b2, 0x11d2, \
{ 0xbd, 0xd6, 0x0, 0x0, 0x64, 0x65, 0x73, 0x74 } }
%}
[uuid(bd300b18-1c34-4589-8285-23a12cc580ea)]
interface nsIXPConnect : nsISupports
{
%{ C++
NS_DEFINE_STATIC_CID_ACCESSOR(NS_XPCONNECT_CID)
%}
/**
* Initializes classes on a global object that has already been created.
*/
void
initClasses(in JSContextPtr aJSContext,
in JSObjectPtr aGlobalJSObj);
/**
* Creates a new global object using the given aCOMObj as the global
* object. The object will be set up according to the flags (defined
* below). If you do not pass INIT_JS_STANDARD_CLASSES, then aCOMObj
* must implement nsIXPCScriptable so it can resolve the standard
* classes when asked by the JS engine.
*
* @param aJSContext the context to use while creating the global object.
* @param aCOMObj the native object that represents the global object.
* @param aPrincipal the principal of the code that will run in this
* compartment. Can be null if not on the main thread.
* @param aFlags one of the flags below specifying what options this
* global object wants.
*/
nsIXPConnectJSObjectHolder
initClassesWithNewWrappedGlobal(
in JSContextPtr aJSContext,
in nsISupports aCOMObj,
in nsIPrincipal aPrincipal,
in uint32_t aFlags);
const uint32_t INIT_JS_STANDARD_CLASSES = 1 << 0;
// Free bit here!
const uint32_t OMIT_COMPONENTS_OBJECT = 1 << 2;
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/**
* wrapNative will create a new JSObject or return an existing one.
*
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* The JSObject is returned inside a refcounted nsIXPConnectJSObjectHolder.
* As long as this holder is held the JSObject will be protected from
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* collection by JavaScript's garbage collector. It is a good idea to
* transfer the JSObject to some equally protected place before releasing
* the holder (i.e. use JS_SetProperty to make this object a property of
* some other JSObject).
*
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* This method now correctly deals with cases where the passed in xpcom
* object already has an associated JSObject for the cases:
* 1) The xpcom object has already been wrapped for use in the same scope
* as an nsIXPConnectWrappedNative.
* 2) The xpcom object is in fact a nsIXPConnectWrappedJS and thus already
* has an underlying JSObject.
*
* It *might* be possible to QueryInterface the nsIXPConnectJSObjectHolder
* returned by the method into a nsIXPConnectWrappedNative or a
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* nsIXPConnectWrappedJS.
*
* This method will never wrap the JSObject involved in an
* XPCNativeWrapper before returning.
*
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* Returns:
* success:
* NS_OK
* failure:
* NS_ERROR_XPC_BAD_CONVERT_NATIVE
* NS_ERROR_XPC_CANT_GET_JSOBJECT_OF_DOM_OBJECT
* NS_ERROR_FAILURE
*/
nsIXPConnectJSObjectHolder
wrapNative(in JSContextPtr aJSContext,
in JSObjectPtr aScope,
in nsISupports aCOMObj,
in nsIIDRef aIID);
/**
* Same as wrapNative, but also returns the JSObject in aVal. C++ callers
* can pass in null for the aHolder argument, but in that case they must
* ensure that aVal is rooted.
* aIID may be null, it means the same as passing in
* &NS_GET_IID(nsISupports) but when passing in null certain shortcuts
* can be taken because we know without comparing IIDs that the caller is
* asking for an nsISupports wrapper.
* If aAllowWrapper, then the returned value will be wrapped in the proper
* type of security wrapper on top of the XPCWrappedNative (if needed).
* This method doesn't push aJSContext on the context stack, so the caller
* is required to push it if the top of the context stack is not equal to
* aJSContext.
*/
void
wrapNativeToJSVal(in JSContextPtr aJSContext,
in JSObjectPtr aScope,
in nsISupports aCOMObj,
in nsWrapperCachePtr aCache,
in nsIIDPtr aIID,
in boolean aAllowWrapper,
out jsval aVal,
out nsIXPConnectJSObjectHolder aHolder);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/**
* wrapJS will yield a new or previously existing xpcom interface pointer
* to represent the JSObject passed in.
*
* This method now correctly deals with cases where the passed in JSObject
* already has an associated xpcom interface for the cases:
* 1) The JSObject has already been wrapped as a nsIXPConnectWrappedJS.
* 2) The JSObject is in fact a nsIXPConnectWrappedNative and thus already
* has an underlying xpcom object.
* 3) The JSObject is of a jsclass which supports getting the nsISupports
* from the JSObject directly. This is used for idlc style objects
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
* (e.g. DOM objects).
*
* It *might* be possible to QueryInterface the resulting interface pointer
* to nsIXPConnectWrappedJS.
*
* Returns:
* success:
* NS_OK
* failure:
* NS_ERROR_XPC_BAD_CONVERT_JS
* NS_ERROR_FAILURE
*/
void
wrapJS(in JSContextPtr aJSContext,
in JSObjectPtr aJSObj,
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
in nsIIDRef aIID,
[iid_is(aIID),retval] out nsQIResult result);
/**
* Wraps the given jsval in a nsIVariant and returns the new variant.
*/
nsIVariant
jSValToVariant(in JSContextPtr cx, in JSValPtr aJSVal);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/**
* This only succeeds if the JSObject is a nsIXPConnectWrappedNative.
* A new wrapper is *never* constructed.
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
*/
nsIXPConnectWrappedNative
getWrappedNativeOfJSObject(in JSContextPtr aJSContext,
in JSObjectPtr aJSObj);
[noscript, notxpcom] nsISupports
getNativeOfWrapper(in JSContextPtr aJSContext,
in JSObjectPtr aJSObj);
JSObjectPtr
getJSObjectOfWrapper(in JSContextPtr aJSContext,
in JSObjectPtr aJSObj);
void setSecurityManagerForJSContext(in JSContextPtr aJSContext,
in nsIXPCSecurityManager aManager,
in uint16_t flags);
void getSecurityManagerForJSContext(in JSContextPtr aJSContext,
out nsIXPCSecurityManager aManager,
out uint16_t flags);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/**
* The security manager to use when the current JSContext has no security
* manager.
*/
void setDefaultSecurityManager(in nsIXPCSecurityManager aManager,
in uint16_t flags);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
void getDefaultSecurityManager(out nsIXPCSecurityManager aManager,
out uint16_t flags);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
nsIStackFrame
createStackFrameLocation(in uint32_t aLanguage,
in string aFilename,
in string aFunctionName,
in int32_t aLineNumber,
in nsIStackFrame aCaller);
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
/**
* @deprecated do-nothing function.
Lots of xpconnect bug fixes... - fix bug 12954 "should throw when setting non-settable props". - fix bug 13418 "xpconnect needs to be threadsafe". I think I filled in the cracks. Tests would be nice :) - fix bug 22802 "[MLK] XPConnect Leaks". - fix bug 24119 "[MLK] Reminder about cleaning up maps". - fix bug 24453 "xpconnect needs default security manager". I also changed the code in DOM and caps to just install a default secman and not install a secman for each JSContext. - fix bug 24687 "xpconect should avoid resolve performance suckage". Added (modified) patch from shaver to create my JSObjects with the global object as the temporary proto to avoid losing lookup. - hack for bug 24688 "runtime errors in wrapped JS are not made obvious" Added a debug only printf. We still need a JSErrorConsole service for this. - fix bug 16130 "createInstanace and getService can create wrappers around wrappers" Fixing this one really entailed changing the semantics of nsIXPConnect::wrapNative and nsIXPConnect::wrapJS to use common code in xpcconvert that deals with existing wrappers and DOM objects (with their own schemes for wrapping and unwrapping). So, I changed the callers because the params changed slightly and some callers were doing more work than necessary given the new semantics. - Continued in the crusade to replace manaual refcounting with nsCOMPtrs whenever touching old code. - Added myself as first contributor to xpconnect files (vanity prevails!) - Added new copyright header on some files that were missing it. - Added some API comments. - Converted nsXPCWrappedJS to implement nsIXPConnectWrappedJS via MI rather than the old loser scheme of the nsIXPConnectWrappedJSMethods tearoff object. - added DumpJSStack as globals to xpconnect and DOM dlls to be callable from debuggers. I have ideas on how to improve and expand this support soon. r=mccabe
2000-01-26 08:38:10 +00:00
*/
[deprecated] void syncJSContexts();
readonly attribute nsIStackFrame CurrentJSStack;
readonly attribute nsAXPCNativeCallContextPtr CurrentNativeCallContext;
void debugDump(in short depth);
void debugDumpObject(in nsISupports aCOMObj, in short depth);
void debugDumpJSStack(in boolean showArgs,
in boolean showLocals,
in boolean showThisProps);
void debugDumpEvalInJSStackFrame(in uint32_t aFrameNumber,
in string aSourceText);
/**
* wrapJSAggregatedToNative is just like wrapJS except it is used in cases
* where the JSObject is also aggregated to some native xpcom Object.
* At present XBL is the only system that might want to do this.
*
* XXX write more!
*
* Returns:
* success:
* NS_OK
* failure:
* NS_ERROR_XPC_BAD_CONVERT_JS
* NS_ERROR_FAILURE
*/
void
wrapJSAggregatedToNative(in nsISupports aOuter,
in JSContextPtr aJSContext,
in JSObjectPtr aJSObj,
in nsIIDRef aIID,
[iid_is(aIID),retval] out nsQIResult result);
// Methods added since mozilla 0.6....
/**
* This only succeeds if the native object is already wrapped by xpconnect.
* A new wrapper is *never* constructed.
*/
nsIXPConnectWrappedNative
getWrappedNativeOfNativeObject(in JSContextPtr aJSContext,
in JSObjectPtr aScope,
in nsISupports aCOMObj,
in nsIIDRef aIID);
nsIXPCFunctionThisTranslator
getFunctionThisTranslator(in nsIIDRef aIID);
nsIXPCFunctionThisTranslator
setFunctionThisTranslator(in nsIIDRef aIID,
in nsIXPCFunctionThisTranslator aTranslator);
nsIXPConnectJSObjectHolder
reparentWrappedNativeIfFound(in JSContextPtr aJSContext,
in JSObjectPtr aScope,
in JSObjectPtr aNewParent,
in nsISupports aCOMObj);
void
moveWrappers(in JSContextPtr aJSContext,
in JSObjectPtr aOldScope,
in JSObjectPtr aNewScope);
void clearAllWrappedNativeSecurityPolicies();
nsIXPConnectJSObjectHolder
getWrappedNativePrototype(in JSContextPtr aJSContext,
in JSObjectPtr aScope,
in nsIClassInfo aClassInfo);
void releaseJSContext(in JSContextPtr aJSContext, in boolean noGC);
jsval variantToJS(in JSContextPtr ctx, in JSObjectPtr scope, in nsIVariant value);
nsIVariant JSToVariant(in JSContextPtr ctx, in jsval value);
/**
* Create a sandbox for evaluating code in isolation using
* evalInSandboxObject().
*
* @param cx A context to use when creating the sandbox object.
* @param principal The principal (or NULL to use the null principal)
* to use when evaluating code in this sandbox.
*/
[noscript] nsIXPConnectJSObjectHolder createSandbox(in JSContextPtr cx,
in nsIPrincipal principal);
/**
* Evaluate script in a sandbox, completely isolated from all
* other running scripts.
*
* @param source The source of the script to evaluate.
* @param cx The context to use when setting up the evaluation of
* the script. The actual evaluation will happen on a new
* temporary context.
* @param sandbox The sandbox object to evaluate the script in.
* @param returnStringOnly The only results to come out of the
* computation (including exceptions) will
* be coerced into strings created in the
* sandbox.
* @return The result of the evaluation as a jsval. If the caller
* intends to use the return value from this call the caller
* is responsible for rooting the jsval before making a call
* to this method.
*/
[noscript] jsval evalInSandboxObject(in AString source, in JSContextPtr cx,
in nsIXPConnectJSObjectHolder sandbox,
in boolean returnStringOnly);
/**
* Root JS objects held by aHolder.
* @param aHolder The object that hold the JS objects that should be rooted.
* @param aTrace The tracer for aHolder.
*/
[noscript] void addJSHolder(in voidPtr aHolder,
in nsScriptObjectTracerPtr aTracer);
/**
* Stop rooting the JS objects held by aHolder.
* @param aHolder The object that hold the rooted JS objects.
*/
[noscript] void removeJSHolder(in voidPtr aHolder);
2012-07-06 17:28:51 +00:00
/**
* Test to see if a JS holder is in our hashtable.
* Only available in debug builds.
* @param aHolder The object to test for.
*/
[noscript] bool testJSHolder(in voidPtr aHolder);
/**
* Note aJSContext as a child to the cycle collector.
* @param aJSContext The JSContext to note.
* @param aCb The cycle collection traversal callback.
*/
[noscript,notxpcom] void noteJSContext(in JSContextPtr aJSContext,
in nsCCTraversalCallbackRef aCb);
/**
* Whether or not XPConnect should report all JS exceptions when returning
* from JS into C++. False by default, although any value set in the
* MOZ_REPORT_ALL_JS_EXCEPTIONS environment variable will override the value
* passed here.
*/
void setReportAllJSExceptions(in boolean reportAllJSExceptions);
/**
* Trigger a JS garbage collection.
* Use a js::gcreason::Reason from jsfriendapi.h for the kind.
*/
void GarbageCollect(in uint32_t reason);
/**
* Signals a good place to do an incremental GC slice, because the
* browser is drawing a frame.
*/
void NotifyDidPaint();
%{C++
/**
* Get the object principal for this wrapper. Note that this may well end
* up being null; in that case one should seek principals elsewhere. Null
* here does NOT indicate system principal or no principals at all, just
* that this wrapper doesn't have an intrinsic one.
*/
virtual nsIPrincipal* GetPrincipal(JSObject* obj,
bool allowShortCircuit) const = 0;
virtual char* DebugPrintJSStack(bool showArgs,
bool showLocals,
bool showThisProps) = 0;
%}
/**
* Creates a JS object holder around aObject that will hold the object
* alive for as long as the holder stays alive.
*/
nsIXPConnectJSObjectHolder holdObject(in JSContextPtr aJSContext,
in JSObjectPtr aObject);
/**
* Return the caller object of the current call from JS.
*/
[noscript,notxpcom] void getCaller(out JSContextPtr aJSContext,
out JSObjectPtr aObject);
/**
* When we place the browser in JS debug mode, there can't be any
* JS on the stack. This is because we currently activate debugMode
* on all scripts in the JSRuntime when the debugger is activated.
* This method will turn debug mode on or off when the context
* stack reaches zero length.
*/
[noscript] void setDebugModeWhenPossible(in boolean mode,
in boolean allowSyncDisable);
[noscript] void writeScript(in nsIObjectOutputStream aStream,
in JSContextPtr aJSContext,
in JSScriptPtr aJSScript);
[noscript] JSScriptPtr readScript(in nsIObjectInputStream aStream,
in JSContextPtr aJSContext);
[noscript] void writeFunction(in nsIObjectOutputStream aStream,
in JSContextPtr aJSContext,
in JSObjectPtr aJSObject);
[noscript] JSObjectPtr readFunction(in nsIObjectInputStream aStream,
in JSContextPtr aJSContext);
};