Code written in Dev-C++ won't compile on linux

The teacher gave us (in the course) the choice of either C or C++. I started with C and my friend with C++. I'm trying to help this friend with this code because she said it is not formatting okay. The students all do the exercises in Dev-C++ on windows. I'm the only one doing it on linux and vim. The problem is that when I try to compile the c++ code I get some errors. It runs fine in Dev-C++.

$ g++ luciana_matriz.cpp -o luciana_matriz
luciana_matriz.cpp: In function ‘int main()’:
luciana_matriz.cpp:105:13: error: ‘else’ without a previous ‘if’
luciana_matriz.cpp:142:12: error: ‘stdin’ was not declared in this scope
luciana_matriz.cpp:142:17: error: ‘fflush’ was not declared in this scope
luciana_matriz.cpp:143:13: error: ‘getchar’ was not declared in this scope

Here's the code:

Code:

#include <iostream>
#include <cstdlib>
/* #include <cstring> */
/* #include <unistd.h> */
/* #include <cstdio.h> */
/* #include <stdio.h> */


using namespace std;

main() {
    // exercicio 1:
    int l, c, TL, TC, num;
    cout<< "Informe o numero de linhas: \n";
    cin >> TL;
    cout<< "Informe o numero de de colunas: \n";
    cin>> TC;
    int M[TL][TC], N[TL][TC], Mm[TL][TC], S[TL][TC], A[TL][TC], M1[TL][TC], D[5][5], Id[4][4];
    srand(time(NULL));

    cout<< "\n M: \n" ;
    for(l=0;l<TL; l++) {
        for(c=0; c<TC; c++)
        {
            M[l][c]=rand()%500;
            cout<< M[l][c] <<"\t";
        }  
        cout <<"\n";
    }

    cout<< "\n N: \n" ;
    for(l=0;l<TL; l++) {

        for(c=0; c<TC; c++)
        {    
            N[l][c]=rand()%500;
            cout<< N[l][c] <<"\t";

        }
        cout <<"\n";
    }

    // exercicio 1 a:

    cout<< "\n\n a) Escolha um numero para multiplicar a matriz M:\t";
    cin>> num;

    cout<< "\n   M: \t\t\t\t   Mm: \n" ;
    for(l=0;l<TL; l++) {
        for(c=0; c<TC; c++) {
            cout << M[l][c]<<"\t"; 
        }
        cout<<"\t\t";
        for(c=0; c<TC; c++) {
            Mm[l][c]= num * M[l][c];    
            cout << Mm[l][c]<<"\t"; 
        }
        cout <<"\n\n";

    }    

    //exercicio 1 b:
    cout<< "\n\n b) Gerar Matriz S sendo a soma de M e N. \n" <<" Matriz S: \n";

    for(l=0;l<TL; l++) {
        for(c=0; c<TC; c++)
        {
            S[l][c]= M[l][c] + N[l][c];
            cout<< S[l][c]<< "\t";
        }
        cout <<"\n";
    }  

    // exercício 1 c:
    cout<< "\n\n c) Gerar Matriz A de qualquer dimensao formada pela lei A[L,C]=(3*L-2*C). \n" <<"Matriz A: \n";

    cout<< "\n A: \n" ;
    for(l=0;l<TL; l++) {
        for(c=0; c<TC; c++) {
            A[TL][TC]=(3*l - 2*c);
            cout<< A[TL][TC] <<"\t";
        }  
        cout <<"\n";
    } 
    //exercício 1 d
    cout<< "\n\n d) Mostrar a matriz lida e criar matriz M1,trasnposta da matriz M. \n" <<"Matriz M1: \n";
    cout<< "\n M1: \n" ;
    for(l=0;l<TL; l++) {
        for(c=0; c<TC; c++) {
            M1[l][c]= M[c][l];
            cout<< M1[l][c] <<"\t";
        }  
        cout <<"\n";
    }
    //exercicio 1 e  
        cout<< "\n\n d) Montar uma matriz diagonal. \n" <<"Matriz D: \n";
    cout<< "\n D: \n" ;

    for(l=0; l<5; l++) {
        for(c=0; c<5; c++) {
            if(l==c)
                D[l][c]= '1';
            cout<<"\t %l" << D[l][c]<<"\n";

            else 

                D[l][c]= '0';
            cout<< "\t %c" << D[l][c];
        }
    }   

    for(l=0; l<5; l++)
    {
        for(c=0; c<5; c++)
        {
            cout<<D[l][c]<< " \t ";
        }
        cout<< "\n";
    }                
    // exercício 1 f
        cout<< "\n\n f) Montar uma Matriz identidade. \n" <<"Matriz Id: \n";

    cout<< "\n Id: \n" ;

    for(l=0;l<4; l++)
    {
        for(c=0; c<4; c++) {
            if(l==c)
                Id[l][c]=1;

            else
                // if(l!=c)
                Id[l][c]= 0;  
            cout<< "\t%c" <<'%' << Id[l][c] << "\t";
        }
        cout << "\n";
    }
    cout<< "\n";    



    fflush(stdin);
    getchar();
}

Any help would be appreciated.

Are you sure you didn't just copied the code in a wrong way?, because I doubt Dev-C++( using GCC ) compiles just fine with these diamonds! ( C++ is strongly-typed, meaning it isn't as torelant as C )

EDIT: Seeing by the comments, obviously you didn't include the necessary headers.

Originally Posted by GReaper

Are you sure you didn't just copied the code in a wrong way?, because I doubt Dev-C++( using GCC ) compiles just fine with these diamonds! ( C++ is strongly-typed, meaning it isn't as torelant as C )

She sent the source file to me by email. I guess she saved it from Dev-C++ to the hard disc before
sending it to me. (I didn't copy).

What do you mean by "these diamonds"?

Originally Posted by FernandoBasso

What do you mean by "these diamonds"?

Well, obviously there is an error, as it says, that an else is used without an if. That happens as I see because there's a stray line in between. ( If you want to execute multiple lines in if statements you must use curly braces )

All the others are because C++ is picky about declaration-prior-use. Include <cstdio> and you'll be fine. ( Although, I wouldn't recommend mixing C and C++ Input/Output )

Originally Posted by GReaper

Well, obviously there is an error, as it says, that an else is used without an if. That happens as I see because there's a stray line in between. ( If you want to execute multiple lines in if statements you must use curly braces )

All right. I hadn't noticed that. The teacher doesn't teaches us the language. She only "teaches" algorithms. And yes, she instructed the students that "we didn't need curly braces for if else if we wrote only one line after if or else. I got used to always use them, no matter how much lines I write (and I use autoclose.vim so that I don't forget to close a curly brace or a parenthesis or even a ".

Originally Posted by GReaper

All the others are because C++ is picky about declaration-prior-use. Include <cstdio> and you'll be fine. ( Although, I wouldn't recommend mixing C and C++ Input/Output )

Well, I was not aware of that (and I bet the "teacher" doesn't either). I write a little bit php/sql already but found C much more interesting and enjoyable). I see the students doing all kinds of wrong things, like mixing indenting styles, different styles of where to put the opening curly brace after if or for. Well, bad practices. So, when the code just runs, even if it throws a lot of warnings, it is okay to the teacher and the students.

I search for solutions in this forum, books and google. One of my goals is really learn good standards.

Thanks for your help. Solved.

By the way, can't we mark threads as solved in this forum?

Read FAQ: Why fflush(stdin) is wrong.

Originally Posted by GReaper

Are you sure you didn't just copied the code in a wrong way?, because I doubt Dev-C++( using GCC ) compiles just fine with these diamonds! ( C++ is strongly-typed, meaning it isn't as torelant as C )

EDIT: Seeing by the comments, obviously you didn't include the necessary headers.

Be careful here with your wording. In programming, strongly-typed has a very defined meaning and that isn't it. Strongly type means that the languages variables need to be assigned a type, either directly like when you state "int myvar;“ or inferred like when you say "var myVar=42;" as opposed to weak or duck typed languages ( as in, looks like a duck, quacks like a duck, its a duck )

What I believe you meant to say in C++ sources are sensitive to character encoding.


Sadly character encoding is platform dependant, although most will support UtF8 in addition to ascii

Strongly type means that the languages variables need to be assigned a type, either directly like when you state "int myvar;“ or inferred like when you say "var myVar=42;" as opposed to weak or duck typed languages ( as in, looks like a duck, quacks like a duck, its a duck )

(It seems to me..)Your definition does not fit into C++'s new features like auto, decltype etc.

Anyways, I said it because he didn't include the prototypes of the functions, and a strongly-typed language demands, among other, to declare everything before using them. So essentially you said the same thing but in a different way!

Not true, auto is still strongly typed, its type is simply inferred by the rvalue. I would need to google to verify, as I've not used auto yet, but I assume it will throw an error if you violate the initial type it infers. So for example

auto meaningOfLife=42.0f; // type inferred a float
meaningOfLife = "Brian"; // blammo!

In a weak typed language, the above would be perfectly fine, but in a strongly typed language like C# or C++, once the type is assign via inference, it cannot be changed.

Now C#`s dynamic muddies the waters of strongly/weakly typed language divide, although at its core I believe dynamic still forces a type on the far, even if a obscenely base one like object.

Oddly I can't edit my post using the mobile interface.


A key thing I should point out, and the biggest difference between dynamic and strongly typed languages is *when* type is determined.

In strongly typed language, the inference happens as part of the compilation process. So if you assign x type a value it cannot support once x has been inferred by the compier, it will through an error. Behind the scenes, x is still assigned a type, its just the compiler that determines at, as I said earlier, generally using the rvalue.

Dynamically typed languages however are resolved at run time, generally allowing variables to be mutable, something untrue of strongly typed languages. This a double edged sword, as it is extremely flexible and can result in some remarkably concise code, it also is slower and potentially dangerous as hell.

Originally Posted by Serapth

Not true, auto is still strongly typed, its type is simply inferred by the rvalue. I would need to google to verify, as I've not used auto yet, but I assume it will throw an error if you violate the initial type it infers. So for example

auto meaningOfLife=42.0f; // type inferred a float
meaningOfLife = "Brian"; // blammo!

In a weak typed language, the above would be perfectly fine, but in a strongly typed language like C# or C++, once the type is assign via inference, it cannot be changed.

Agreed..
I meant your "( as in, looks like a duck, quacks like a duck, its a duck )" analogy.
When you use auto... if the rvalue quacks like a duck.. the newly declared one will be a duck!

Ah, good point, it does somewhat invalid the expression.

I suppose "once a duck, always a duck" may take its place!

Originally Posted by Serapth

I suppose "once a duck, always a duck" may take its place!

Generally...
But in python (which claims to be strongly typed) ... you can delete the duck and make a rabbit with the same name inside the same scope.

Originally Posted by manasij7479

Generally...
But in python (which claims to be strongly typed) ... you can delete the duck and make a rabbit with the same name inside the same scope.

Well Python takes the screwy cake prize for typing. I by no means pretend to be an expert on the subject, but I believe it goes something like this...


someVar = 42
someVar = "42"


What happens here ( I believe ), is that a completely new string object is created and the label "someVar" is rebound to it, effectively replacing someVar's completely. Then the int version of someVar is released if its reference count is 0. However:

someVar = 42
someOtherVar = someVar;
someVar = "42"
someVar = 42

Will cause the same action to change again, in that new int object is created and the "someVar" label or name is applied to the new object, and if the old string version has no external references, it is purged. In this case however, since someOtherVar refers to it, it is kept around, but no longer accessable as "someVar" or as anything other than the reference someOtherVar.

To use your example, if you turned a variable "animal" from a duck object to a rabbit object, you actually are creating a new rabbit object and assigning the label "animal" to it. The type never actually changed behind the scenes, just the object referenced did.

So, really, the type doesn't actually changed, thus Python remains a "strongly typed" langauge. And a screwy language.


There is a reason I don't really seem to mesh well with Python after all. You are right though, it really does muddy the metaphor up even worse.