I know that passing arguments by const instead of value is more efficient and allows us to avoid allocating a temporary local variable of the argument type. But is this always true? Or are there some cases when calling functions with constant arguments should be avoided? If so, is passing by pointer the most efficient way?
View 6 RepliesHaving a little trouble passing arguments to functions.
I wrote this simple program to get the hang of it but I'm quite stuck. I'm sure you will be able to get at what i want the program to do...
Code:
#include <conio.h>
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#include <windows.h>
void menu(int HP,int Gold,int Armour);
[Code] ....
I have a 1wire program from maxim running in visual studio. There is this argument in the main function that requires the com port to be specified the command line. If I do pass it as "COM1" the program works as expected.
I don't want to depend on having to pass "COM1" in the command line and into main. I've tried creating a string for COM1 and passing it right into the if function but it doesn't work.
Code:
int main(int argc, char **argv) {
int len, addr, page, answer, i;
int done = FALSE;
SMALLINT bank = 1;
uchar data[552];
[Code] .....
I have a program that is working very well when I pass C++ vectors as arguments to my functions by reference, but I get some compilation errors when try to make a modification. I am also posting the entire program and its output below. so that you can see what is going on. I have commented out the line that causes an error.(Some of the indentation that got corrupted when I copied the code to the browser.)
This program basically calculates the coefficients of a least square polynomial and then evaluates this polynomial at artificial data points and verifies that this actually reproduces the original data within reasonable floating point error.
The function that computes the coefficients of the least square polynomial is Code: vector<double> LSPVecValued_GSL( const int, const vector<float> &, const vector<float> &); and as you can see it returns a vector by value, and this vector contains the coefficients of the least square polynomial.
There is also a function that evaluates this polynomial by accepting a vector argument by reference : Code: float evaluate_polynomial(double, vector<double>& ) ; I have also created another version of the evaluation function which accepts the same vector argument by value: Code: float evaluate_polynomial_ByValue(double t, vector<double> vec_a) ; In the program I call the first evaluation function (whose vector argument is passed by reference) by first using an intermediate vector variable containing the coefficients, and then I pass this vector as an argument to the evaluation function, as follows:
Code:
vec_a = LSPVecValued_GSL( deg, vec_x , vec_y);
for(int j=0; j< n ; j=j+20 ) {
cout<<"x["<<j<<"] = " << vec_x[j] << " ,y["<<j<<"] = " << vec_y[j] <<" , p(x["<<j<<"]) ( EVALUATED FROM REFERENCE) = "
<< evaluate_polynomial( vec_x[j], vec_a) << endl; // This version works without error
[Code] .....
As you can see above, I am also able to call the second evaluation function (the one whose vector argument is passed by value) directly by plugging in the function LSPVecValued_GSL"(...)" and this works without error, and this is a one step process, only one line of code is involved.
However, I get a compilation error (line number 12 that I have commented out above) if I try to plug in the function "LSPVecValued_GSL(...)" into the first evaluation function that expects a vector argument by reference. I tried to put a "&" in front ofLSPVecValued_GSL but this did not fix the bug.
What syntax is appropriate to use the first evaluation function (which accepts a vector argument by reference) if I want to plug in the vector-valued function LeastSquarePolynomial_GSL directly in the the first version of the evaluation function which expects a vector argument by reference?
I am trying to make quicksort and binary search and I get error when I am passing dynamic array to argument. It also says error during initialization of the dynamic array.
//.h file
#ifndef SortableArray_h
#define SortableArray_h
#include <iostream>
#include <string>
[Code] ....
Code:
void Class1::Func(shared_ptr<type1> parameter)
{
}
or
void Class1::Func(const shared_ptr<type1>& parameter)
{
}
or
Should I ever pass arguments/parameters to other objects using shared_ptr's or raw pointers?
I am using visual studio 2012 and i pass three command line arguments as 10 20 30 and when i m compile the program get error.....
Code:
#include<stdio.h>
#include<conio.h>
#include<stdlib.h>
main(int n,char **p) {
int sum=0,i;
if(n>=2)
[Code] .....
Suppose if i write a test program like
Code:
void function1(unsigned int var1);
int main(void) {
function1(-3);
}
void function1(unsigned int var1) {
printf("%d", var1);
}
The output is -3. how it happens the argument is unsigned but iam passing signed but still prints the signed value. My bigger question is how the arguments are handled if the passing parameters are different types compared to declaration.
#include <stdio.h>
#include <stdlib.h>
int size_b_row1;
int size_b_col1;
int main(void) {
double C[4][4] = {{1,3,5,2},{7,6,2,2},{1,2,7,3},{2,3,5,3}};
double TC[4][4];
transpose(C, TC);
[Code] ......
Are there other ways of calling a const/non-const override? I want to defined some functions in terms of others, particularly accessors which might or might not require constness- in order to not copy & paste code. This is my current solution:
struct dumbArray {
dumbArray(unsigned int size):
m_array(new int[size]){
}
~dumbArray(){
delete m_array;
[Code] .....
difference between const and static const, more effectively. I know the basic concept of const and static but I need clear explanation of declaring "const" and "static const"
Ex:
const int a;
static const int a;
Is there any way to cast a non-const variable to const one?
I want to read variable n from file and then use it to declare array "int arr[n]", but because n is non-const, the compiler doesn't allow me to do that.
So I'm making setTimeout and setInterval functions.
I have this remember function (that is part of Timing class) which takes a function pointer and a void pointer, which are remembered in that object.
Another (timing) function of that object is called in every loop of the program and when specific time passes that function calls the remembered function whit the remembered void pointer as argument.
The problem is that the functions that need to be called require unknown multiple parameters, so what I need to do is make a new class that will store the needed arguments. I make the function that needs to be called and that storage object and pass pointers to them to my remember function, when the remembered function is called it stores the data from storage object in new variables and dose it's thing.
So I start thinking about what's the difference between this 2 code
map<const int, float> map_data;
map<int, float> map_data;
But it seems I can't find the difference, is there any difference between this 2 code ?
#include <iostream>
#include <iomanip>
using namespace std;
// cin.get() <-------------- used to let the user read the screen
// Function prototypes
void calcSales(const int [], const double [], double [], int);
void showOrder(const double [], const int [], int);
[Code] ....
How can i change the "const int NUM_PRODS = 12;" from saying id 1, id 2, id 3, etc. to custom product numbers?
I want to use a const char* as a buffer. I am reading values from a file and adding them to a buffer. How to extract the values is simple enough. I am reading through a filestream, reading each character into a char pointer and progressing that char pointer every time. I have another char pointer marking the start positon
eg.
char *mychar = new char;
char *char1 = new char;
char *char2 = new char;
const char *constchar ;
char2 = char1;
while(filestream.read(mychar,1) {
*char1 = *mychar;
++char1;
}
Then I get this problem: constchar = mychar; // const char* = char*.
Constchar does not catch all the data in other words. At some stage some data is lost due to zeros in the data.. How can I put values into a const char and get around this problem? The const char* will //only record everything up until the first zero.
What is the best way to define const strings when there are separate header and source files?
For example, I have a header that only declare some enums. In that same header I would like to add string representations of those enums so that I can print them easily i.e string_representation[my_enum] for debug and error printing and so on.
If I define them in the header file, I will get a linker error for multiple definitions. If I remove the definition, then I can not define it in the source file.
Now I am stuck with getting const char* array from function to main.
Code:
const char* values[3];
strings_to_array()
printf("%s
", values[1]);
printf("%s
", values[2]); function: Code:
const char* strings_to_array()
}
[code]....
First, I cant get strings printed out in main.
Second, is here way to get number of such elements which array contains like higher languages have "count" or "ubound" or such?
Third, next function which need that array assumes that array of const chars* is 1 based. Can that be set in C or here are allways zero based arrays?
I'm currently finishing up an assignment that was half written by my professor. Below in the testGrades section of code there are two errors both are the same message.
Error: no matching function for call to Grades:: Grades(const char [15])
Test Grades
//Purpose: Test program for the class Grades
// Create stu1 Grades object
// Add 5 grades to stu1 - only 3 can be stored in stu1 - other 2 discarded
// Create stu2 Grades object
// Add only 2 grades
#include <iostream>
#include <string>
#include <iomanip>
#include "Grades.h"
using namespace std;
[Code] .....
I've been having a problem concerning the initialization of const static integral members with floating point calculations. I'll let his sample program do the explaining:
class Foo {
public :
Foo() {}
const static int samplerate = 44100;
const static unsigned short tempo = 120;
[Code].....
I know you can't initialize const static non-integral types on the same line on which they're declared, but I don't see why even an implicit cast to an integral type should be disallowed. I make my calculations using doubles, so I'm surprised that even though it should degenerate into an integer - it's still a problem for the compiler.
What I need to know is how I can pass an argument to the Book constructor so I can change the const data member Category (with cascading capacity if possible. I also posted some of my set functions for further comprehension.
class Book
{
friend void CompPrice(Book &,Book&); //friend function that has access to the member functions of this class
//The arguments sent to it are by address, and of type the class Book, so that it can have access to its member functions
private: //private data members
[Code].....
i am trying to describe the unusual situation where you declare a class member function with this format:
bool class::function_name(void) const
Specifically where the 'const' follows the parameter list. It is my understanding this is a very useful way of ensuring that whatever code you put in the function definition cannot change any data members of its class.
However I have recently read that this form of declaration should not be used as it leads to less optimized and slower code. Is this correct?
i found a lot about how to convert int to const char * but non of them explain correctly or sometimes dont even work.
Here is my code:
#include <stdio.h>
#include <cstdlib>
using namespace std;
int main () {
int i =0;
char c1 =0;
char c2 =0;
[code]....
Why I'm getting an error here. I've dumbed down my code for simplicity and removed irrelevant code.
PHP Code:
class Foo {
public:
bool IsNull() const;
private:
std::map<int, int*> test;
[Code] ....
I'm getting a "passing...discards qualifiers" error on my if statement and not sure why because I'm not changing anything. I know removing const or making test mutable fixes the issue. I've been taught to always make a function const if it doesn't change anything, in which case, have I finally come across an acceptable time to use mutable?
i have this rather large class, which (in a way) somehow resembles a custom dialog control). This control is supposed to display data, which it does just fine. To do so, it maintains a
byte settings[10];
array, which holds information on how to display the data.
There are multiple ways to represent this custom set of data.In order to remain flexible in representing it, i thought of implementing some sort of DisplayProvider, which can be registered to the base class and provides that settings byte array.
Preferably, i would now have a set of static const instances of this provider.Using a struct would work nicely here:
PHP Code:
struct DisplayProvider
{
int settings[10];
}
static const DisplayProvider prov1 = {1,2,3,4,5,6,7,8,9,0};
The problem: The DisplayProvider would have to do some pre-processing, before handing over control to the base class, which then does the main work.I would end up with something like this:
PHP Code:
class DispalyProvider
{
baseclass* owner;
int settings[10];
void PreProcessing(...);//ends up calling the owner.Processing(...) function
};
The main thing here is, that i dont really see a way to create a stock of default "static const DisplayProvder = {...}"s, as i could when using a struct.
I hope I got all the jargon correct. I have something like this:
Code:
const Fl_Color my_fl_dark_gray=fl_color_cube(64*(FL_NUM_RED-1)/255, 64*(FL_NUM_GREEN-1)/255, 64*(FL_NUM_BLUE-1)/255);
in a header file and the header file is included in several C files.
Questions:
At run time,
Is there just one copy of the const variable my_fl_dark_gray or are there multiple copies for the multiple C files?If a function uses the const variable, does the initialization statement "my_fl_dark_gray=fl_color_cube(...);" run every time the function is called or does it just run once and then when the function is called it just uses the value stored in memory?