Thread: streaming errors

can someone tell me what's wrong with my code? i'm using msvc++. the errors i get are that i can't access private data types (i think the code's right) and that the << operator is ambiguous (?).

i can't even get it to compile but i've gone through my code a dozen+ times.

thanks,
barneygumble742

main.cpp

Code:

#include <iostream>
#include "file.h"
using namespace std;

int main()
{
	BigFloat x;
	cout <<	x	<< endl;
	x = "-123456.9832155";

return 0;
}

file.h

Code:

#ifndef BigFloat_h
#define BigFloat_h

#include <iostream>
#include <string>
using namespace std;

class BigFloat
{
public:	BigFloat();		// default constructor
	~BigFloat();		// destructor
	friend ostream& operator << (ostream &ostr, const BigFloat &BF);
	void operator = (const string &s);

private:string digit;
	int DecimalPlace;
	bool sign;
};

	ostream& operator << (ostream &ostr, const BigFloat &BF);

#endif

file.cpp

Code:

#include "file.h"

BigFloat::BigFloat()		// initialize data members
{
	digit="00";
	DecimalPlace = 1;
	sign = true;		// sign initially set to positive
}

// need destructor

ostream& operator << (ostream &ostr, const BigFloat &BF)
{
	ostr	<<	"digit   = "	<< BF.digit		<< endl;
	ostr	<<	"decimal = "	<< BF.DecimalPlace	<< endl;
	ostr	<<	"sign    = "	<< BF.sign		<< endl;
	return ostr;
}

void BigFloat::operator =(const string &s)
{
	sign = true;
	string ss = s;
	if(ss[0] == '-')
	{
		sign = false;
		ss = ss.substr(1, ss.length());
	}
	else
	{
		if(ss[0] == '+')
		{
			ss = ss.substr(1,ss.length());
		}
	}
	int dp = ss.find('.');
	dp = DecimalPlace;
	DecimalPlace = ss.length() - dp;
	digit = digit + ss.substr(dp+1, DecimalPlace);
}

You should tell us the exact error message that you get.
I have tried your code (with g++ ) and it compiles and runs ok. It is propably not doing what you expect but that is another issue.
Kurt

here's the error message i get. i've seen this same exact code run on msvc++ and that ran ok. i have a feeling that i'm missing something very small somewhere.

thanks,
barneygumble742

--------------------Configuration: main - Win32 Debug--------------------
Compiling...
BigFloat.cpp
c:\file.cpp(14) : error C2248: 'digit' : cannot access private member declared in class 'BigFloat'
c:\file.h(15) : see declaration of 'digit'
c:\file.cpp(15) : error C2248: 'DecimalPlace' : cannot access private member declared in class 'BigFloat'
c:\file.h(16) : see declaration of 'DecimalPlace'
c:\file.cpp(16) : error C2248: 'sign' : cannot access private member declared in class 'BigFloat'
c:\file.h(17) : see declaration of 'sign'
main.cpp
c:\main.cpp(8) : error C2593: 'operator <<' is ambiguous
Error executing cl.exe.

main.exe - 4 error(s), 0 warning(s)

Ok. I pasted all your code into one file and tested that.
You could try to put the implementation of the << operator into file.h.
there seems to be an issue with friend. maybe that helps.

Code:

inline ostream& operator << (ostream &ostr, const BigFloat &BF)
{
	ostr	<<	"digit   = "	<< BF.digit		<< endl;
	ostr	<<	"decimal = "	<< BF.DecimalPlace	<< endl;
	ostr	<<	"sign    = "	<< BF.sign		<< endl;
	return ostr;
}

Kurt

i pasted all the code into one file and tested it and i got the same error messages as before.

i don't think any of the header files should have any implementation code; only the class.

i think the problem has to be somewhere in th << operator implementation. once i fix the ambiguity error, then i think the other "private data" errors will be fixed.

I have split your code the same way as you did ( changed a few things ).
compiles and runs.
BTW the destructor was not implemented.
main.cpp

Code:

#include <iostream>
#include "bf.h"
using namespace std;

int main() {
   BigFloat x;
   cout <<	x	<< endl;
   x = "-123456.9832155";
   cout <<	x	<< endl;
   return 0;
}

bf.h

Code:

#ifndef bigfloat_included
#define bigfloat_included

#include <string>
#include <iostream>

//using namespace std; // never do that in a header

class BigFloat
{
public:	BigFloat();	// default constructor
	~BigFloat(){}	// destructor not doing anything
	void operator = (const std::string &s);
	
	friend std::ostream& operator << (std::ostream &ostr, const BigFloat &BF);

private:
        std::string digit;
	int DecimalPlace;
	bool sign;
};

#endif

bf.cpp

Code:

#include <iostream>
#include "bf.h"
using namespace std;

void BigFloat::operator =(const string &s)
{
	sign = true;
	string ss = s;
	if(ss[0] == '-')
	{
		sign = false;
		ss = ss.substr(1, ss.length());
	}
	else
	{
		if(ss[0] == '+')
		{
			ss = ss.substr(1,ss.length());
		}
	}
	int dp = ss.find('.');
	//DecimalPlace=dp; // don't think this is intended
	DecimalPlace = ss.length() - dp;
	digit = digit + ss.substr(dp+1, DecimalPlace);
}

BigFloat::BigFloat()		// initialize data members
{
	digit="00";
	DecimalPlace = 1;
	sign = true;		// sign initially set to positive
}

std::ostream& operator << (std::ostream &ostr, const BigFloat &BF) {
	ostr	<<	"digit   = "	<< BF.digit		<< std::endl;
	ostr	<<	"decimal = "	<< BF.DecimalPlace	<< std::endl;
	ostr	<<	"sign    = "	<< BF.sign		<< std::endl;
	return ostr;
}

Kurt

hi kurt...thanks for all the help. but wouldn't

Code:

using namespace std;

remove the necessity for using

Code:

std::

all over the place?

the way kurt fixed it works however i've seen it done without all the extra std::

here's my current code:

main.cpp:

Code:

#include <iostream>
#include "BigFloat.h"
using namespace std;

int main()
{
	BigFloat x;
	x = "-12345.67890";
	cout <<	x	<< endl;

return 0;
}

bigfloat.cpp:

Code:

#include "BigFloat.h"

BigFloat::BigFloat()		// initialize data members
{
	digit="00";
	DecimalPlace = 1;
	sign = true;		// sign initially set to positive
}

ostream& operator << (ostream &ostr, const BigFloat &BF)
{
	ostr	<<	"digit   = "	<< BF.digit			<< endl;
	ostr	<<	"decimal = "	<< BF.DecimalPlace	<< endl;
	ostr	<<	"sign    = "	<< BF.sign			<< endl;
	return ostr;
}

void BigFloat::operator = (const string &s)
{
	sign = true;
	string ss = s;
	if(ss[0] == '-')
	{
		sign = false;
		ss = ss.substr(1, ss.length());
	}
	else
	{
		if(ss[0] == '+')
		{
			ss = ss.substr(1,ss.length());
		}
	}
	int dp = ss.find('.');
	dp = DecimalPlace;
	DecimalPlace = ss.length() - dp;
	digit = digit + ss.substr(dp+1, DecimalPlace);
}

bigfloat.h:

Code:

#ifndef BigFloat_h
#define BigFloat_h

#include <iostream>
#include <string>
using namespace std;

class BigFloat
{
public:	BigFloat();		// default constructor
		~BigFloat(){};		// destructor
		friend ostream& operator << (ostream &ostr, const BigFloat &BF);
		void operator = (const string &s);

private:string digit;
		int DecimalPlace;
		bool sign;
};


#endif

error messages:
bigfloat.cpp(12) : error C2248: 'digit' : cannot access private member declared in class 'BigFloat'
bigfloat.h(15) : see declaration of 'digit'
bigfloat.cpp(13) : error C2248: 'DecimalPlace' : cannot access private member declared in class 'BigFloat'
bigfloat.h(16) : see declaration of 'DecimalPlace'
bigfloat.cpp(14) : error C2248: 'sign' : cannot access private member declared in class 'BigFloat'
bigfloat.h(17) : see declaration of 'sign'
main.cpp
Error executing cl.exe.

The code should work. Change the compiler.
Kurt

yes of course. i don't know why but g++ does the trick.

thanks kurt and to everyone else. i though sp6 for msvc++ covered all the bugs.

i just installed sp6 for msvs and that took care of the problem. the last set of code that i posted runs perfectly on msvc++.

thanks again everyone.