Introduction

This is an attempt to list some common programming features and how they are implemented in Fortran and in C and C++. The main goal is to help anyone who has to write programs in both languages and needs a quick way to compare/translate some function or the other. This note also covers topics related to writing mixed language programs. This is not an attempt to list the relative merits of either language. This is also not very complete, just the things that I found interesting or convenient to document.

It is important to bear in mind that C and C++ are quite different from each other, though they share some common syntax. I have used them interchangeably for this note, but I have tried to add a comment if something will work in one and not the other.

The Fortran features discussed below are for Fortran-77, and newer Fortrans such as Fortran-90 may do some things differently. Some of the examples/syntax described below may be specific to Digital Fortran. I have shown Fortran examples in uppercase characters, but they work just as well in lowercase. The mixed-language examples are for the Microsoft Visual C++ compiler and Digital Visual Fortran. They should work in other environments with minimal changes.

If you notice any error or omission please send me an email.

General Observations

(1) Most things in C/C++ are case-sensitive, in Fortran they are not. For example, the variables named SUM and sum are different in C, but the same in Fortran.

(2) Fortran-77 (and earlier) will ignore anything after the 72'nd column of any record. In addition, the first 5 columns are reserved for either a statement label or comment identifier. Column-6 has a special meaning, and is used to indicate a continuation of the previous line. By contrast, C and C++ (and Fortran-90) are much more free-format.

Back to Top

Arithmetic operations/functions

.	Fortran	Example	C/C++	Example	Comment
Add	+	A=B+C	+	a=b+c;	.
Subtract	-	A=B-C	-	a=b-c;	.
Multiply	*	A=B*C	*	a=b*c;	.
Divide	/	A=B/C	/	a=b/c;	.
Modulus	MOD	A=MOD(B,C)	%	a=b%c;	ints only in C, reals possible in Fortran
Power	**	A=B**C	pow()	a=pow(b,c);	C/C++ function not intrinsic, provided via math.h

Back to Top

Relational and Logical Operators

.	Fortran	Example	C/C++	Example	Comment
Equal to	.EQ.	IF(A.EQ.B)...	==	if(a==b)...	.
Not Equal to	.NE.	IF(A.NE.B)...	!=	if(a!=b)...	.
Less Than	.LT.	IF(A.LT.B)...	<	if(a<b)...	.
Greater Than	.GT.	IF(A.GT.B)...	>	if(a>b)...	.
Less Than or Equal to	.LE.	IF(A.LE.B)...	<=	if(a<=b)...	.
Greater Than or Equal to	.GE.	IF(A.GE.B)...	>=	if(a>=b)...	.
Logical Not	.NOT.	IF(.NOT.A)...	!	if(!a)...	.
Logical AND	.AND.	IF(A.AND.B)...	&&	if(a&&b)...	.
Logical OR	.OR.	IF(A.OR.B)...	||	if(a||b)...	.

Back to Top

Bitwise Operators

.	Fortran	Example	C/C++	Example	Comment
Bitwise AND	IAND	IAND(N,M)	&	n&m	.
Bitwise OR	IOR	IOR(N,M)	|	n|m	.
Bitwise Exclusive OR	IEOR	IEOR(N,M)	^	n^m	.
Bitwise 1's Complement	INOT	INOT(N)	~	~n	.
Bitwise Left Shift	ISHFT	ISHFT(N,M) (M > 0)	<<	n<<m	n shifted left by m bits
Bitwise Right Shift	ISHFT	ISHFT(N,M) (M < 0)	>>	n>>m	n shifted right by m bits

Back to Top

Mathematical Functions

Note: In C/C++, must include the header file "math.h" to use these functions. Angles must be specified in radians for these functions. "n/a" means "not available".

.	Fortran	Example	C/C++	Example	Comment
Sine	SIN	SIN(R)	n/a	n/a	Single Precision
Sine	DSIN	DSIN(R)	sin	sin(r)	Double Precision
Cosine	COS	COS(R)	n/a	n/a	Single Precision
Cosine	DCOS	DCOS(R)	cos	cos(r)	Double Precision
Tangent	TAN	TAN(R)	n/a	n/a	Single Precision
Tangent	DTAN	DTAN(R)	tan	tan(r)	Double Precision
Arc Sine	ASIN	ASIN(R)	n/a	n/a	Single Precision
Arc Sine	DASIN	DASIN(R)	asin	asin(r)	Double Precision
Arc Cosine	ACOS	ACOS(R)	n/a	n/a	Single Precision
Arc Cosine	DACOS	DACOS(R)	acos	acos(r)	Double Precision
Arc Tangent	ATAN	ATAN(R)	n/a	n/a	Single Precision
Arc Tangent	DATAN	DATAN(R)	atan	atan(r)	Double Precision
Hyperbolic Sine	SINH	SINH(R)	n/a	n/a	Single Precision
Hyperbolic Sine	DSINH	DSINH(R)	sinh	sinh(r)	Double Precision
Hyperbolic Cosine	COSH	COSH(R)	n/a	n/a	Single Precision
Hyperbolic Cosine	DCOSH	DCOSH(R)	cosh	cosh(r)	Double Precision
Hyperbolic Tangent	TANH	TANH(R)	n/a	n/a	Single Precision
Hyperbolic Tangent	DTANH	DTANH(R)	tanh	tanh(r)	Double Precision

Back to Top

Calling Fortran routines from C++

Example-1: Calling routines and functions

The following sample shows how Fortran routines and functions can be called from a C++ program.

 (1) The C++ file:

  //  This illustrates how a Fortran routine and function may be
  //  called from a main program in C++
  #include <iostream.h>
  extern "C" 
  {
      void __stdcall FR1(int*,int *);
      int __stdcall FF1(int *);
  }
  int main()
  {
      int n=10,nSq,nCube;	
      FR1(&n,&nSq);
      cout << "The square is:" << nSq << endl;
      nCube=FF1(&n);	
      cout << "The Cube is:" << nCube << endl;
      return 0;
  }
  
 (2) The Fortran File:

        SUBROUTINE FR1(N,M)
  C     COMPUTES THE SQUARE OF N, RETURNS IN M     
        M=N*N
        RETURN
        END
  C
        INTEGER FUNCTION FF1(N)
  C     COMPUTES THE CUBE OF N 
        FF1=N*N*N
        RETURN
        END

 (1) The C++ file:

  //  This illustrates how a Fortran routine may be
  //  called from a main program in C++, and a char[] string passed to it
  #include <iostream.h>
  #include <string.h>
  extern "C" 
  {
      void __stdcall FR1(int *,int *,char *);
  }
  int main()
  {
    	int n=10,nSq;
	char szCtest[20];
	strcpy(szCtest,"teststring");
	FR1(&n,&nSq,szCtest);
	cout << "The square is:" << nSq << endl;
        return 0;
  }
  
 (2) The Fortran File:

        SUBROUTINE FR1(N,M,CSTR)
	INTEGER*4 CSTR(1)
  C     HERE WE RECEIVE THE C CHAR STRING IN AN INTEGER ARRAY
  C     COULD ALSO HAVE USED A BYTE ARRAY
	M=N*N
	WRITE(6,20) (CSTR(L),L=1,3)
    20  FORMAT(' CSTR=',3A4)
        WRITE(6,*) 'DONE'
	RETURN
	END

 (1) The C++ file:

   // Illustrate passing integer and floating point arrays
   // from C++ to Fortran
   #include <iostream.h>
   extern "C" 
   {
       int __stdcall SUMIT(int *,int*);
       float __stdcall MEAN(float*,int*);
   }
   int main()
   {
       int iA[]={3,5,6,7,2,3,4,5,11,7},iN=10,iSum;
       float fpA[]={1.2f,3.f,44.f,2.5f,-1.3f,33.44f,5.f,0.3f,-3.6f,24.1f},fpMean;
       iSum=SUMIT(iA,&iN);	
       cout << "The Sum of iA is:" << iSum << endl;
       fpMean=MEAN(fpA,&iN);	
       cout << "The Mean of fpA is:" << fpMean << endl;
       return 0;
   }
  
 (2) The Fortran File:

      INTEGER FUNCTION SUMIT(IA,N)
      INTEGER IA(1)
      ISUM=0
      DO 50 J=1,N
  50  ISUM=ISUM+IA(J)
      SUMIT=ISUM
      RETURN
      END
C
      REAL FUNCTION MEAN(RA,N)
      REAL RA(1)
      SUM=0.
      DO 50 J=1,N
  50  SUM=SUM+RA(J)
      IF(N.GT.0) MEAN=SUM/FLOAT(N)
      RETURN
      END 

 (1) The Fortran file:

      INTEGER CR2
      N=10
      CALL CR1(N,M)
      WRITE(6,20) N,M
  20  FORMAT(' The square of',I3,' is',I4)
      K=CR2(N)
      WRITE(6,30) N,K
  30  FORMAT(' The cube of',I3,' is',I15)
      CALL EXIT
      END

 (2) The C++ files:

      extern "C" 
      {
	void __stdcall CR1(int *,int *);
	int __stdcall CR2(int *);
      }
      void __stdcall CR1(int *n, int *m)
      {
	// Compute the square of n, return in m
	int k;
	k=*n;
	*m=k*k;
	return;
      }
      int __stdcall CR2(int *n)
      //  Compute the cube of n
      {
        int m,k;
	k=*n;
        m=k*k*k;
	return m;
      }