Help with a small problem (beginner)

Hello,

I just recently started to learn C and i have a small question.
I wanted to use Flags to send to a function, but so far i havent been able to get them to work.

I have created a small command program in VS 2008. The only thing it prints out is Red, what am i missing?

Code:

---

#include "stdafx.h"

#define RED                0001
#define BLUE                0010
#define GREEN                0011
#define BLACK                0000
#define YELLOW                0100

void PrintData(char);

int _tmain(int argc, _TCHAR* argv[])
{
        int res;

        PrintData(RED | BLUE);


        scanf("%d", &res);
        return 0;
}

void PrintData(char mask)
{
        if ((RED & mask) == 1)
                printf("Red is used.\n");
        if ((BLUE & mask) == 1)
                printf("Blue is used.\n");
        if ((GREEN & mask) == 1)
                printf("Green is used.\n");
        if ((BLACK & mask) == 1)
                printf("Black is used.\n");
        if ((YELLOW & mask) == 1)
                printf("Yellow is used.\n");
}

---

Code:

---

        (RED & mask)

---

What is that supposed to result in?
And what about

Code:

---

        (BLUE & mask)

---

Do you really want Yellow to be the value 64 decimal, or do you believe that your colour selections are in BINARY and you actually got 4 decimal?

--
Mats

Quote:

---

I just recently started to learn C and i have a small question.

---

This is the C++ programming forum. Shall I move this thread to the C programming forum?

Quote:

---

This is the C++ programming forum. Shall I move this thread to the C programming forum?

---

-Yes

Quote:

---

Do you really want Yellow to be the value 64 decimal, or do you believe that your colour selections are in BINARY and you actually got 4 decimal?

---

-I want to make it into binary or hex, so the Decimal value would be 4.

Quote:

---

Originally Posted by piffo

-Yes


-I want to make it into binary or hex, so the Decimal value would be 4.

---

In which case you can use one of the following:

Code:

---

#define YELLOW 4
#define YELLOW (1 << 2)
#define YELLOW 04
#define YELLOW 0x04

---

What you have with

Code:

---

#define YELLOW 0100

---

is an octal number - base 8 - so you get 1 * 8 * 8 = 64 decimal.

--
Mats

Quote:

---

Originally Posted by matsp

In which case you can use one of the following:

Code:

---

#define YELLOW 4
#define YELLOW (1 << 2)
#define YELLOW 04
#define YELLOW 0x04

---

What you have with

Code:

---

#define YELLOW 0100

---

is an octal number - base 8 - so you get 1 * 8 * 8 = 64 decimal.

--
Mats

---

But i still cant manage to compare it against the defined variables and use it as flags

Code:

---

#define RED                        0x01
#define BLUE                0x02
#define GREEN                0x03
#define BLACK                0x00
#define YELLOW                0x05

void PrintData(char);

int _tmain(int argc, _TCHAR* argv[])
{
        int res;

        PrintData(RED | BLUE);


        scanf("%d", &res);
        return 0;
}

void PrintData(char mask)
{
        if ((RED & mask) == 1)
                printf("Red is used.\n");
        if ((BLUE & mask) == 1)
                printf("Blue is used.\n");
        if ((GREEN & mask) == 1)
                printf("Green is used.\n");
        if ((BLACK & mask) == 1)
                printf("Black is used.\n");
        if ((YELLOW & mask) == 1)
                printf("Yellow is used.\n");
}

---

I want to check using binary, ex. 00000101. Then i have red declared as Decimal 1, if it excists inside the binary stirng, i want it to print, red is used. Then it should check if the 2nd digit in the binary string is used, if so, print Blue is used. etc.

Code:

---

if ((RED & mask) == 1)
if ((BLUE & mask) == 1)
if ((GREEN & mask) == 1)
if ((BLACK & mask) == 1)
if ((YELLOW & mask) == 1)

---

Only two of those comparisons will ever possibly evaluate to true: RED and GREEN. If you pass RED into the function, I'll bet it prints RED and GREEN. You need to reconsider your comparison strategy there (after you fix the accidental octal notation you used).

What you are doing (in binary):

Code:

---

    0001 (red)
OR  0010 (blue)
    ----
    0011 (green)

so 'mask' == 0011

    0011 (mask)
AND 0011 (green)
    ----
    0011

so (GREEN & mask) == 0011

What are you comparing it against....??

---

Your binary values, when used in the context you are describing, needs to have unique bit combinations, such as:

Code:

---

#define RED  0x01
#define GREEN 0x02
#define BLUE  0x04

---

As it stands, your GREEN & RED != 0.

And to use that in the way you do now, you can either use:

Code:

---

if (val & BLUE)

---

or

Code:

---

if (!!(val & BLUE) == 1)

---

The if(!!x == 1) is a simple way of writing ((x != 0) == 1).

--
Mats

Quote:

---

Originally Posted by arpsmack

Code:

---

if ((RED & mask) == 1)
if ((BLUE & mask) == 1)
if ((GREEN & mask) == 1)
if ((BLACK & mask) == 1)
if ((YELLOW & mask) == 1)

---

Only two of those comparisons will ever possibly evaluate to true: RED and GREEN. If you pass RED into the function, I'll bet it prints RED and GREEN. You need to reconsider your comparison strategy there (after you fix the accidental octal notation you used).

What you are doing (in binary):

Code:

---

    0001 (red)
OR  0010 (blue)
    ----
    0011 (green)

so 'mask' == 0011

    0011 (mask)
AND 0011 (green)
    ----
    0011

so (GREEN & mask) == 0011

What are you comparing it against....??

---

---

Ahh i see what you mean. What i tried to accomplish was, when using your code:

Code:

---

PrintData(RED | BLUE);
This would return  "0011"


    0011 (flag)
&  0001 (red)
    ----
    0001

or ((flag & RED) == 1)
To see if the red bit in the flag is set to 1

simliar with

    0011 (flag)
&  0010 (blue)
    ----
    0010

or ((flag & BLUE) == 1)
To see if the blue bit in the flag is set to 1

---

You may benefit from doing this:

Code:

---

#define PRINT(x, y) printf(#x " ((%d) & " #y " (%d) == 1) = %d\n", x, y, x & y, (x & y) == 1)
PRINT(RED|BLUE, BLUE);
PRINT(RED, RED);

---

--
Mats

Quote:

---

Originally Posted by matsp

You may benefit from doing this:

Code:

---

#define PRINT(x, y) printf(#x " ((%d) & " #y " (%d) == 1) = %d\n", x, y, x & y, (x & y) == 1)
PRINT(RED|BLUE, BLUE);
PRINT(RED, RED);

---

--
Mats

---

Problem is, it dosent solve what i am trying to accomplish.
It dosent matter to me if the decimal value is 3 or 5, i am only after if the bit in that location is turned on or off.

0101
what i want to check is, if bit "1" is set to 1, print Red is used. If bit "2" is set to 1, print Blue is used. If bit "3" is set to 1, print Green is used. If bit "4" is set to 1, print Yellow is used.

So from the above binary string i would get the following output:
Red is used.
Green is used.

That way, i can send in alot of constants to the program and i only have to have one parameter that accepts it. So instead of writing a function:

void func1 (bool RED, bool BLUE, bool GREEN etc);

i can write:

void func1 (char colorstouse);

Then just call it with the flags that i need:

func1(RED | BLUE | GREEN | YELLOW);

instead of

func1(true, false, true, false, false etc);

Quote:

---

Originally Posted by piffo

Problem is, it dosent solve what i am trying to accomplish.
It dosent matter to me if the decimal value is 3 or 5, i am only after if the bit in that location is turned on or off.

0101
what i want to check is, if bit "1" is set to 1, print Red is used. If bit "2" is set to 1, print Blue is used. If bit "3" is set to 1, print Green is used. If bit "4" is set to 1, print Yellow is used.

So from the above binary string i would get the following output:
Red is used.
Green is used.

That way, i can send in alot of constants to the program and i only have to have one parameter that accepts it. So instead of writing a function:

void func1 (bool RED, bool BLUE, bool GREEN etc);

i can write:

void func1 (char colorstouse);

Then just call it with the flags that i need:

func1(RED | BLUE | GREEN | YELLOW);

instead of

func1(true, false, true, false, false etc);

---

You can't test that the result of the & equals one, since it might equal two, or four, or eight, or whichever bit you're looking at. You just need to test that it's nonzero: if (colorstouse & RED). And if you plan to do that, you need to make sure that there are no overlaps in the bits, which is why you use powers-of-two.

You are trying to use the individual bits of a parameter to determine what the information passed is.
But you must also know then, that we're working on a binary level, not octal or decimal or hexdecimal.
You should know that the binary OR operator adds together bits and that the binary AND operator selects out bits.
For example: 0000 1111 | 1111 0000 = 1111 1111
0000 1111 & 0000 1000 = 0000 1000
0000 1111 & 0001 0000 = 0000 0000

Since we can't directly write numbers in binary format, you have to convert them to decimal or hexdecimal. Use a calculator for that (Windows calc supports this in scientific mode).
I quickly find out that:
0000 0001 = 0x1
0000 0010 = 0x2
0000 0100 = 0x4
0000 1000 = 0x8
0001 0000 = 0x10
0010 0000 = 0x20
0100 0000 = 0x40
1000 0000 = 0x80

If, for whatever reason, you cannot figure out how to do it from this information, then you need to re-read how binary operators works.

You may also want to use an enum if you only need ONE flag at a time.

And in post #8, I _DID_ suggest a solution to that problem (in fact two different solutions). Perhaps re-reading that post, and asking further questions if unclear, is a good idea.

--
Mats