How can I concatenate 2 arrays by using reference operator '&'?

Question

2.33/5 (3 votes)

See more:

Hi!
I want to concatenate 2 array as follows:

C++

double little_arrayA[] = {1, 2, 3, 4, 5};
double little_arrayB[] = {6, 7, 8, 9, 10};

double *BIG_array = new double [2*5];

//START "BLOCK 1" (concatenate by using reference operator '&')
&BIG_array[0] = little_arrayA;
&BIG_array[5] = little_arrayB;
//END "BLOCK 1"

Obviously I can not compile this code but I would like to understand how to implement the function in "BLOCK 1" in the most efficient way. For efficiency, using cycle seems unsuitable, but rather I wanted to use not working reference operator '&'.
Do you have any idea?

Thank you!

Sergio

EDIT:
I thank you for all the information received! Hoping to do a bit of clarity, I present below some constraints to be considered:
1) I have to concatenate 2 array of double in a single array of double.
2) By the word "efficiency" I mean perform the function in the shortest possible time.
3) Each array is large/very large.
4) I don't know array size in compile time.

Posted 18-Apr-12 13:47pm

cammaratasergio

Updated 19-Apr-12 2:42am

v2

Add a Solution

Comments

Philippe Mori 18-Apr-12 22:54pm

That question does not make much sense... & is the address-of operator when used that way. Thus, on the left-hand side, you have a pointer... but it cannot be modified (think of it as a constant pointer). And operator= is a member function when overloaded so it not possible to overload it as the left-hand side is a pointer.

You should learn to use STL and avoid low-level stuff like that. Uses classes like std::vector<double> instead.

[no name] 18-Apr-12 23:24pm

If you had offered this as a solution I would have voted for it.

cammaratasergio 19-Apr-12 6:39am

I agree with you:). In fact, I wrote that "OBVIOUSLY I can not compile this code but I would like to understand how to implement the function in the most efficient way."

Philippe Mori 19-Apr-12 8:08am

Well efficiency in that academic context is not that much useful. In a real application, the most efficient way would depend a lot on what you are doing.

For small arrays like that it won't matter much and for larger arrays or operation done in large loops, the performance will generally depends much more on what is around that code than the code itself.

cammaratasergio 19-Apr-12 10:51am

Thanks to all!
At the end, given that I use not overlapping memory blocks, I prefer memcpy.
Moreover, it would seem that memcpy is more efficient than std::copy(). (http://stackoverflow.com/questions/4707012/c-memcpy-vs-stdcopy)

5 solutions

Solution 4

Hm... :) :

C++

{
    int arSmallOwnerA[] = { 1, 2, 3, 4, 5, 6, 7, 8 };
    int arSmallOwnerB[] = { 1, 2, 3 };
    int* piConcatenation[_countof(arSmallOwnerA) + _countof(arSmallOwnerB) + 1] = { NULL };
 
    int i = 0;
    for( ;i < _countof(arSmallOwnerA); i++ )
    {
        piConcatenation[i] = &arSmallOwnerA[i];
    }
    for( int j =0; i < _countof(arSmallOwnerA) + _countof(arSmallOwnerB); i++, j++ )
    {
        piConcatenation[i] = &arSmallOwnerB[j];
    }
 
    for( int i = 0; i < _countof( piConcatenation ) - 1; i++ )
    {
	cout<< *(piConcatenation[i]) << endl;
    }
  }

Posted 18-Apr-12 19:55pm

Eugen Podsypalnikov

Updated 18-Apr-12 20:38pm

v3

Comments

Resmi Anna 19-Apr-12 2:40am

"Code Edition" not so bad I guess.....

Eugen Podsypalnikov 19-Apr-12 2:45am

Thanks, 5-ed :)

Resmi Anna 19-Apr-12 2:53am

:)

cammaratasergio 19-Apr-12 6:45am

Thank you for your code! But I think that using cycles is a suboptimal solution.
Rather, I thought to implement the function by explicitly managing memory.

[no name] 19-Apr-12 8:20am

No it isn't. It depends on the size of the arrays. You have posted a very generic question. For small arrays this method is fine. For large arrays memcpy is the only way. Provide specific information in your question and you will get better answers.

Philippe Mori 19-Apr-12 8:01am

There is an extra * in the declaration of piConcatenation. Thus the code is wrong.

Solution 2

Check out here:

http://stackoverflow.com/questions/1700182/how-do-i-merge-two-arrays-having-different-values-into-one-array[^]

http://stackoverflow.com/questions/1696074/how-can-i-concatenate-two-arrays-in-c[^]

Posted 18-Apr-12 16:07pm

Lakamraju Raghuram

Add a Solution

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Eugen Podsypalnikov · Accepted Answer · 2012-04-19T03:26:00

Solution 6

// EDIT
How about it ? :) :

C++

template <typename T>
class Concatenation {
  T* m_pDataA;
  int m_iCountA;
  
  T* m_pDataB;
  int m_iCountB;

public:
  Concatenation(T* pDataA, int iCountA, T* pDataB, int iCountB)
  : m_pDataA(pDataA), m_iCountA(iCountA)
  , m_pDataB(pDataB), m_iCountB(iCountB)
  {
    // some assertions... :)
  }

  T* GetAt(int iIdx)
  {
    // some assertions... :)

    T* pResult(NULL);

    if (0 <= iIdx && iIdx < (m_iCountA + m_iCountB)) {
      if (iIdx < m_iCountA) {
        pResult = &m_pDataA[iIdx];
      } else {
        pResult = &m_pDataB[iIdx - m_iCountA];
      }
    }

    return pResult;
  }
};

void Test(int iCountA, int iCountB)
{
  // some assertions... :)

  int* pIntA(new int[iCountA]);
  int* pIntB(new int[iCountB]);

  Concatenation<int> cCat(pIntA, iCountA, pIntB, iCountB);
  *cCat.GetAt((iCountA + iCountB) /2) = 123;
  ASSERT(123 == *cCat.GetAt((iCountA + iCountB) /2));

  delete[] pIntA;
  delete[] pIntB; 
}

Posted 19-Apr-12 3:26am

Eugen Podsypalnikov

Updated 19-Apr-12 12:15pm

v6

Comments

enhzflep 19-Apr-12 9:58am

Good generalized answer.
Poor solution where speed is paramount.

If type was char, there would be a copy operation for each element in each array. Using memcpy, copying may be optimized(by the memcpy function) such 8 or more bytes are copied per cpu instruction - an 8 times improvement in speed of the basic copying, and without the overhead of the class.

Eugen Podsypalnikov 19-Apr-12 10:05am

Thank you,
but I can not see any copy operation here, for <char> as well... :)

enhzflep 19-Apr-12 10:10am

Oh my goodness. I'm more tired than I had realized. I'd misread the first if statement as a for-loop.

Despite my overwhelming urge to delete my initial comment and hide my embarrassment, I will leave it in place as a testament to my stupidity.

You sir, clearly have the fastest method - also the one that satisfies the question as posed. Would you accept a 5 from a fool? :slinks-away:

Eugen Podsypalnikov 19-Apr-12 10:25am

Thank you ! :)

[no name] 19-Apr-12 10:26am

Very neat a 5 for that anyway.
I think this test is wrong - just a typo:
if (0 <= iIdx && iIdx < (m_iCountA + m_iCountA)) {
This does not however create a new array and that may be a requirement.

Eugen Podsypalnikov 19-Apr-12 10:35am

Thanks,
I agree with you :)

Niklas L · Accepted Answer · 2012-04-18T22:10:00

Solution 5

If you're interested in a C++ solution, have a look at std::copy()[^]. It will get you away from all the looping misery.

If you want a C version, take a look at pwassers solution.

Posted 18-Apr-12 22:10pm

Niklas L

User 59241 · Accepted Answer · 2012-04-18T16:08:00

Solution 3

All these arrays are in different and not contiguous locations in memory. To get the contents all in one place efficiently this is one option.

C#

double little_arrayA[] = {1, 2, 3, 4, 5};
double little_arrayB[] = {6, 7, 8, 9, 10};


double *BIG_array = (double *) new char [sizeof(little_arrayA) + sizeof(little_arrayB)];


memcpy(BIG_array, little_arrayA, sizeof(little_arrayA));
memcpy(BIG_array + sizeof(little_arrayA), little_arrayB, sizeof(little_arrayB));

Posted 18-Apr-12 16:08pm

User 59241

Comments

Philippe Mori 19-Apr-12 8:03am

If the size is known at compile time, it can even be more efficient by declaring the resulting array on the stack.

enhzflep 19-Apr-12 9:54am

Can't possibly get any quicker than this approach using memcpy.
Given that the array sizes aren't known at compile time, the sizeof operator won't work correctly, but it doesn't matter, since you already know sizeof(double) and number of elements in each array. (the product of which, of course equals the desired value)
My 5.

[no name] 19-Apr-12 10:10am

Thanks. In my example the array size is known at compile. If not as you say the size must be got another way at runtime. Not really a big deal.

Chuck O'Toole 19-Apr-12 10:11am

Of course, this works and is efficient if memory size is no object. The original poster said the arrays were "large / very large" which makes for an interesting intellectual exercise. What if you cannot contain both arrays *and* the concatenated array in memory at the same time? The "sample" code posted showed statically allocated initial arrays so it's easy to ignore the "size" problem. The "large / very large" arrays are probably not entered as constants into the compiler. If we knew where those came from, it might generate a different solution.

How can I concatenate 2 arrays by using reference operator '&'?

5 solutions

Solution 6

Solution 5

Solution 3

Solution 4

Solution 2

Add your solution here

Preview 0

How can I concatenate 2 arrays by using reference operator '&'?

5 solutions

Solution 6

Solution 5

Solution 3

Solution 4

Solution 2

Add your solution here

Preview 0

Existing Members

...or Join us