I'm using protobuf inside nif function (erlang nif) and need to create resource of protobuf message type. I wrote something like this:
ERL_NIF_TERM create_resource(ErlNifEnv *env, const MyClass &msg)
{
size_t size = sizeof(MyClass);
MyClass *class = (MyClass *)enif_alloc_resource(MY_CLASS, size);
memcpy(class, &msg, size);
// class->CopyFrom(&msg);
ERL_NIF_TERM term = enif_make_resource(env, class);
enif_release_resource(class);
return term;
}
The question is.. is it legal for the protobuf message to be copied like this and in cleanup just release it with:
delete pointer
?
Seems that everything is right here, but I'm not shure, cause the constructor of the copied object was not invoked and may be there is some magick with static vars and etc...
Also.. do I need to call CopyFrom after memcpy?
upd: MyClass is C++ class not C
enif_alloc_resource, enif_release_resource, and enif_make_resource do all the memory management for you. You can make it somewhat easier by making your resource type a pointer, in which case you call delete from your defined resource destructor (the function pointer you pass when calling enif_open_resource_type).
As far as what you're doing with memcpy, it's not safe for complex objects. For instance, if one of your class members is a pointer to a dynamically allocated resource which it destroys in its destructor, and you memcpy it, two objects are now sharing that same resource. When one of the objects is destroyed (falling out of scope, delete operator), the other object is left with a pointer to freed memory.
This is why you define copy and assignment constructors if you have a complex class. I'm guessing CopyFrom should, in fact, be both your assignment and copy constructor.
Related
I am writing some functions for a C extension module for python and need to import a module I wrote directly in python for access to a custom python type. I use PyImport_ImportModule() in the body of my C function, then PyObject_GetAttrString() on the module to get the custom python type. This executes every time the C function is called and seems like it's not very efficient and may not be best practice. I'm looking for a way to have access to the python custom type as a PyObject* or PyTypeObject* in my source code for efficiency and I may need the type in more than one C function also.
Right now the function looks something like
static PyObject* foo(PyObject* self, PyObject* args)
{
PyObject* myPythonModule = PyImport_ImportModule("my.python.module");
if (!myPythonModule)
return NULL;
PyObject* myPythonType = PyObject_GetAttrString(myPythonModule, "MyPythonType");
if (!myPythonType) {
Py_DECREF(myPythonModule);
return NULL;
}
/* more code to create and return a MyPythonType instance */
}
To avoid retrieving myPythonType every function call I tried adding a global variable to hold the object at the top of my C file
static PyObject* myPythonType;
and initialized it in the module init function similar to the old function body
PyMODINIT_FUNC
PyInit_mymodule(void)
{
/* more initializing here */
PyObject* myPythonModule = PyImport_ImportModule("my.python.module");
if (!myPythonModule) {
/* clean-up code here */
return NULL;
}
// set the static global variable here
myPythonType = PyObject_GetAttrString(myPythonModule, "MyPythonType");
Py_DECREF(myPythonModule);
if (!myPythonType) {
/* clean-up code here */
return NULL;
/* finish initializing module */
}
which worked, however I am unsure how to Py_DECREF the global variable whenever the module is finished being used. Is there a way to do that or even a better way to solve this whole problem I am overlooking?
First, just calling import each time probably isn't as bad as you think - Python does internally keep a list of imported modules, so the second time you call it on the same module the cost is much lower. So this might be an acceptable solution.
Second, the global variable approach should work, but you're right that it doesn't get cleaned up. This is rarely a problem because modules are rarely unloaded (and most extension modules don't really support it), but it isn't great. It also won't work with isolated sub-interpreters (which isn't much of a concern now, but may become more more popular in future).
The most robust way to do it needs multi-phase initialization of your module. To quickly summarise what you should do:
You should define a module state struct containing this type of information,
Your module spec should contain the size of the module state struct,
You need to initialize this struct within the Py_mod_exec slot.
You need to create an m_free function (and ideally the other GC functions) to correctly decref your state during de-initialization.
Within a global module function, self will be your module object, and so you can get the state with PyModule_GetState(self)
I'm making a project where there are separate source files/modules that add functions to a single Dictionary contained in a higher level file. However, I find that nothing in these source files evaluates on its own, even functions that take no arguments/code that isn't even inside a function.
As a result nothing is being added to the Dictionary. Is there a way to forcibly evaluate complete function calls in a module automatically? I'll give an example of what I'm trying to do:
Registry.fs:
let private functions = Dictionary<string, MessageHandler>()
let add type handler =
functions.Add(type, handler)
Handler1.fs:
Registry.add "type1" (fun m -> ....
)
Handler2.fs:
Registry.add "type2" (fun m -> ....
)
I believe you need to see this relevant topic. Loose method calls would get compiled as method calls inside of a static constructor for the enclosing type/module, when the F# code gets compiled to IL. This would roughly be equivalent to the following C# code, just to see the picture:
static class Handler1 {
static Handler1() {
// this is the static constructor
Registry.add "type1" ....
}
}
In .NET static constructors are not eagerly initialized1. This means, if you want to cause the .NET runtime to call the Handler1 static constructor, you need to access a static member of the type Handler1.
An example of using the type in a static context would be to
Expose a sufficiently accessible static member/method:
module Handler1 =
[<Literal>]
let Name = "Handler1"
Access that static member from your code, such as the main method:
[<EntryPoint>]
let main args =
printf Handler1.Name
The above line will force the .NET runtime to load the Handler1 type's static context, which will result in invoking the static constructor if the type is encoutered by your code for the first time. If your code never encounters a given type's static context (any static member or method), then it will never be initialized -- the static constructors will never get called.
This behaviour is by design of the .NET framework (and that is regardless of the chosen language -- C#, F#, VB, others -- they all compile to similar IL). The point is to not allocate unnecessary resources by types that are never actually used.
1 Until .NET 4, static type context was initialized when the given type was first encountered by the executing code, regardless if the user code is interacting with instace or static members of that type. After .NET 4, this slightly changed -- the static context is initialized only when the user code interacts with static members of the type.
I have the following C# method:
private static bool IsLink(string shortcutFilename)
{
var pathOnly = Path.GetDirectoryName(shortcutFilename);
var filenameOnly = Path.GetFileName(shortcutFilename);
var shell = new Shell32.Shell();
var folder = shell.NameSpace(pathOnly);
var folderItem = folder.ParseName(filenameOnly);
return folderItem != null && folderItem.IsLink;
}
I have tried converting this to F# as:
let private isLink filename =
let pathOnly = Path.GetDirectoryName(filename)
let filenameOnly = Path.GetFileName(filename)
let shell = new Shell32.Shell()
let folder = shell.NameSpace(pathOnly)
let folderItem = folder.ParseName(filenameOnly)
folderItem <> null && folderItem.IsLink
It however reports an error for the let shell = new Shell32.Shell() line, saying that new cannot be used on interface types.
Have I just made a silly syntactic mistake, or is there extra work needed to access COM from F#?
I don't know enough about the F# compiler but your comments makes it obvious enough. The C# and VB.NET compilers have a fair amount of explicit support for COM built-in. Note that your statement uses the new operator on an interface type, Shell32.Shell in the interop library looks like this:
[ComImport]
[Guid("286E6F1B-7113-4355-9562-96B7E9D64C54")]
[CoClass(typeof(ShellClass))]
public interface Shell : IShellDispatch6 {}
IShellDispatch6 is the real interface type, you can also see the IShellDispatch through IShellDispatch5 interfaces. That's versioning across the past 20 years at work, COM interface definitions are immutable since changing them almost always causes an undiagnosable hard crash at runtime.
The [CoClass] attribute is the important one for this story, that's what the C# compiler goes looking for you use new on a [ComImport] interface type. Tells it to create the object by creating an instance of Shell32.ShellClass instance and obtain the Shell interface. What the F# compiler doesn't do.
ShellClass is a fake class, it is auto-generated by the type library importer. COM never exposes concrete classes, it uses a hyper-pure interface-based programming paradigm. Objects are always created by an object factory, CoCreateInstance() is the workhorse for that. Itself a convenience function, the real work is done by the universal IClassFactory interface, hyper-pure style. Every COM coclass implements its CreateInstance() method.
The type library importer makes ShellClass look like this:
[ComImport]
[TypeLibType(TypeLibTypeFlags.FCanCreate)]
[ClassInterface(ClassInterfaceType.None)]
[Guid("13709620-C279-11CE-A49E-444553540000")]
public class ShellClass : IShellDispatch6, Shell {
// Methods
[MethodImpl(MethodImplOptions.InternalCall, MethodCodeType=MethodCodeType.Runtime), DispId(0x60040000)]
public virtual extern void AddToRecent([In, MarshalAs(UnmanagedType.Struct)] object varFile, [In, Optional, MarshalAs(UnmanagedType.BStr)] string bstrCategory);
// Etc, many more methods...
}
Lots of fire and movement, none of it should ever be used. The only thing that really matters is the [Guid] attribute, that provides the CLSID that CoCreateInstance() needs. It also needs the IID, the [Guid] of the interface, provided by the interface declaration.
So the workaround in F# is to create the Shell32.ShellClass object, just like the C# compiler does implicitly. While technically you can keep the reference in a ShellClass variable, you should strongly favor the interface type instead. The COM way, the pure way, it avoids this kind of problem. Ultimately it is the CLR that gets the job done, it recognizes the [ClassInterface] attribute on the ShellClass class declaration in its new operator implementation. The more explicit way in .NET is to use Type.GetTypeFromCLSID() and Activator.CreateInstance(), handy when you only have the Guid of the coclass.
In Delphi, suppose I have a method with a (much simplified) signature like this:
procedure abc( const prop1:string; const arg1:TValue; const prop2:string;
out arg2:TValue );
I'm building a TList<PPropValPair> of records like this using the parameters provided:
type
TPVPType = (ptIn, ptOut);
PPropValPair = ^TPropValPair;
TPropValPair = record
io : TPVPType;
prop : string; // property name
iVal : TValue; // input value
oVar : Variant; // <-- how to save for later use??? Variant? TValue?
end;
(On the face of it, this example looks silly. Again, it's quite simplified just to communicate the problem.)
At run-time, I want to stuff all of the input values ival (where io=ptIn) into each public property 'prop' in a class, call a class method, then extract the values of all public properites 'prop' (where io=ptOut) into oVar.
The input side is working fine using RTTI.
However, I need to somehow save a REFERENCE to the output vars in oVar so I can save the value of the associated properties after the class method has been called.
I'm not assigning anything to arg2 directly. Rather, I'm saving a reference to arg2 and assigning the value indirectly later on.
The trick is ... I don't want to have to do any additional annotations of the output parameters in abc(...).
In C++, you can declare a parameter as a 'reference' by prepending it with '&'. So in C++ terms this might be defined roughly as:
procedure abc( arg1 : TValue; &arg2 : TValue );
Later, you can refer to &arg2 and it's using a POINTER to the object. But in calling the function, you just say:
abc( somevar1, somevar2 );
somevar1 is passed by value, and somevar2 is passed by reference. Inside the function, I can save somevar2 in a reference var, then later on assign a value to it via the pointer (if it's a string) by saying &arg2ref = 'abc'.
I'm guessing there's a way to do this in Delphi, either with a Variant as the oVar type, or using RTTI, or something else. I just haven't figured out the magic combination of pieces yet. (I just don't use pointers very often in Delphi.)
Maybe I need to save a raw pointer in oVar along with the type (say, oType), and cast a value through the pointer to save the property's value?
I'm hoping someone here might have some clear ideas.
BTW, I'm using Delphi XE3.
Use a pointer. It doesn't have to (and indeed shouldn't) be a "raw" pointer. Use a typed pointer, PValue. Pass in a PValue to your function, and then store that in oVal, which you should also declare a a PValue. Use # to create a pointer, and ^ to dereference.
I would not recommend passing arg2 by reference. Although you can still use # on it to get a pointer to the original variable passed to abc, the reference parameter disguises the fact that the variable needs to remain available indefinitely. Instead, declare arg2 as PValue so it's more obvious to the caller that indirection is involved.
// declaration
procedure abc(...; arg2: PValue);
// call
abc(..., #somevar2);
There is something I don't get, please enlighten me.
Is there a difference between the following (client side code)?
1) blah = (const char *)"dummy";
2) blah = CORBA::string_dup("dummy");
... just googling a bit I see string_dup() returns a char * so the 2 may be equivalent.
I was thinking 2) does 2 deep copies and not 1.
I'm firing the question anyway now, please briefly confirm.
Thanks!
const char* blah = "dummy";
The C++ compiler generates a constant array of characters, null-terminated, somewhere in a data section of your executable. blah gets a pointer to it.
char* blah = CORBA::string_dup("dummy");
The function string_dup() is called with an argument that is a pointer to that constant array of characters. string_dup() then allocates memory from the free store and copies the string data into the free-store-allocated memory. The pointer to the free-store memory is returned to the caller. It is the caller's job to dispose of the memory when finished with CORBA::string_free(). Technically the ORB implementation is allowed to use some special free-store, but most likely it is just using the standard heap / free-store that the rest of your application is using.
It is often much better to do this:
CORBA::String_var s = CORBA::string_dup("dummy");
The String_var's destructor will automatically call string_free() when s goes out of scope.