How to pass array of integers from Tcl to C++? [closed] - c++

Question

Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
Questions concerning problems with code you've written must describe the specific problem — and include valid code to reproduce it — in the question itself. See SSCCE.org for guidance.
Closed 9 years ago.
Improve this question
How do I send Tcl array to C++? I have written the following code:
Tcl:
set ns [new Simulator]
set n [$ns node]
$n set X_ 100
$n set Y_ 30
$n set Z_ 0
set x [$n set X_]
set y [$n set Y_]
set z [$n set Z_]
#after 2000
set b {12 2 3 4 5}
set aa [new "Application/Trust/ITLeach"]
$aa set bufer_ 1
$aa set allnode_ $n
$aa set X_ $x
$aa set Y_ $y
$aa set Z_ $z
$aa set ClausterHeadID_ [array get b] **#send array to c++**
$ns at 0.0 "$aa start"
puts $b
$ns run
ITLEACH.h:
#ifndef ns_ITLeach_h
#define ns_ITLeach_h
#include "app.h"
#include "node.h"
#include "tcl.h"
#include "mobilenode.h"
#include <iostream>
#include <fstream>
class ITLeach;
#define TCL_OK 0
class ITLeach : public Application {
public:
ITLeach();
virtual int command(int argc, const char*const* argv);
protected:
// need to define recv and timeout
void start();
int Buffer;
MobileNode * node ;
ofstream nodeSetting;
double XPos ;
double YPos ;
double ZPos ;
int ClausterHeadID [] ; //int array that passed from tcl file
int ClausterID [] ;
int id_node;
};
#endif
ITLEACH.cc:
/*
* ITLeach.cc
*
* Created on: Oct 29, 2013
* Author: root
*/
#include "ITLeach.h"
static class ITLeachClass : public TclClass {
public:
ITLeachClass() : TclClass("Application/Trust/ITLeach") {}
TclObject* create(int, const char*const*) {
return (new ITLeach());
}
} class_app_ITLeach;
ITLeach::ITLeach() : Application() {
Tcl_Obj *baObj = Tcl_NewObj();
bind("bufer_",&Buffer);
bind("allnode_",&node);
bind("X_",&XPos);
bind("Y_",&YPos);
bind("Z_",&ZPos);
bind("ClausterHeadID_",(int *) &ClausterHeadID); // call array from tcl
bind("ClausterID_",ClausterID);
bind("id_",&id_node);
}
int ITLeach::command(int argc, const char*const* argv) {
if (strcmp(argv[1], "start") == 0) {
ITLeach::start();
return(TCL_OK);
}
return(ITLeach::command(argc, argv));
}
void ITLeach::start()
{
//double x=0, y =0 , z =0;
nodeSetting.open("./leachnode.txt",fstream::app);
//node = (MobileNode*)Node::get_node_by_address(i);
//node->location()->getLocation(x,y,z);
//node->getLoc(&x,&y,&z);
nodeSetting << "id " << id_node << " x "<< XPos << " y " << YPos << " z " << ZPos <<"\n";
nodeSetting.close();
printf(" claster head number : %d \n" ,ClausterHeadID[1]);
printf("node number is : %d \n",Buffer);
}
I send array from Tcl with this code:
$aa set ClausterHeadID_ [array get b] **#send array to c++**
and receive array from C++ with this code:
bind("ClausterHeadID_",(int *) &ClausterHeadID); // call array from tcl
But it doesn't work, please help me.

Answer 1

If you've got that command bound to the string interface (i.e., the arguments arrive via int argc, char **argv) then you use Tcl_SplitList() to take apart the relevant argument (which might be argv[argc-1], i.e., the last argument) and then Tcl_GetInt() to retrieve an integer from each of those values. Those integers are the members of that Tcl list.
int listc;
char **listv;
if (Tcl_SplitList(interp, argv[argc-1], &listc, &listv) != TCL_OK) {
// wasn't a valid list!
return TCL_ERROR;
}
std::vector<int> theArray(listc, 0);
for (int i=0 ; i<listc ; i++) {
if (Tcl_GetInt(interp, listv[i], &theArray[i]) != TCL_OK) {
// wasn't an int in the list!
return TCL_ERROR;
}
}
This isn't very fast! For a faster way, you need to use the Tcl_Obj-based API (the Tcl_Obj is the fundamental Tcl first-class value type), starting with registering your implementation function correctly. After that, it's fairly easy to convert the above code:
int listc;
Tcl_Obj **listv;
if (Tcl_ListObjGetElements(interp, argv[argc-1], &listc, &listv) != TCL_OK) {
// wasn't a valid list!
return TCL_ERROR;
}
std::vector<int> theArray(listc, 0);
for (int i=0 ; i<listc ; i++) {
if (Tcl_GetIntFromObj(interp, listv[i], &theArray[i]) != TCL_OK) {
// wasn't an int in the list!
return TCL_ERROR;
}
}
The big difference? A Tcl_Obj knows whether it is holding a string or an integer (or a float or any number of other things) and so the Tcl runtime doesn't normally need to reparse or type-convert values, whereas if everything is a string, you do a lot of conversions. (It's common to say “Everything is a string” in Tcl, but that's inaccurate; the correct version is “Everything has a perfect string serialization, or is a named entity” but that's rather more verbose.)