Please could someone assist me with double digits

I wrote this code to evaluate a postfx expression but it works for only single digit. I need to make it work for double digit and floats. i.e making it work for something like 1315.5*8+ which is for the infix 13*15.7+8. I don't have problem converting from infix to postfix.

Code:

---

#include<stdio.h>
#include<stdio.lib>
#define MAXCOLS 80
main(){
      char instring[MAXCOLS], postring[MAXCOLS];
      int position = 0;
      float eval();
      float result;
      printf("%s", "Enter postfix expression: ");
      while ((postring[position++] = getchar()) != '\n');
      postring[--position] = '\0';
      printf("%s%s", "postfix expression is", postring);
      getchar();
      result=eval(postring);
      printf("%s%f\n", "value is", result);
      getchar();
      }
struct stack{
      int top;
      float items[MAXCOLS];
      };
      float eval(expr)
      char expr[];
      {
            int c, position;
            float op1, op2, value;
            float oper(), pop();
            struct stack opndstk;
            opndstk.top = -1;
            for (position =0; (c=expr[position]) != '\0'; position++)
            if(isdigit(c))
            push (&opndstk, (float)(c-'0'));
            else {
                op2 = pop(&opndstk);
                op1 = pop(&opndstk);
                value = oper(c, op1, op2);
                push (&opndstk, value);
                }
                return (pop(&opndstk));
                }
float oper(symb, operand1, operand2)
int symb;
float operand1, operand2;
{
      switch(symb){
              case '+': return (operand1 + operand2);
              case '-': return (operand1 - operand2);
              case '*': return (operand1 * operand2);
              case '/': return (operand1 / operand2);
              default: printf("%s", "illegal operation");
              exit(1);
              }
              }
push(ps, x)
struct stack *ps;
float x;
{    if(ps->top == MAXCOLS-1){
    printf("%s", "stack overflow");
    exit(1);
    }
    else
    ps->items[++(ps->top)] =x;
    return;
}
float pop(ps)
struct stack *ps;
{int top;
 if (top==-1){
      printf("%s", "stack underflow");
      exit(1);
      }
      else
      return(ps->items[ps->top--]);
      }

---

Quote:

---

1315.5*8+ which is for the infix 13*15.7+8

---

That looks like bad postfix to me. What's to say it isn't 1 * 315.5 + 8 or 131 * 5.5 + 5 ? You need something to separate your numbers, like a space. So:

Code:

---

131 5.5*8+

---

Might be valid, but with just

Code:

---

1315.5*8+

---

We have no idea where to break up the number.

What is

Code:

---

#include<stdio.lib>

---

I would have thought that ".lib" files are no good to include, as they are generally binary files with rather confusing result to the compiler.

And this is definitely broken if you need to do multi-digit numbers:

Code:

---

            if(isdigit(c))
            push (&opndstk, (float)(c-'0'));

---

--
Mats

Somebody please tell me what I should do. to make it work for double digits.

I just pointed out at least part of what doesn't work with the current code. Someone else pointed out that you need to have a separator between your numerbs.

You don't REALLY expect US to write the code FOR YOU, do you?

--
Mats

Quote:

---

Originally Posted by ben2000

Somebody please tell me what I should do. to make it work for double digits.

---

You can start by responding to those who have already started trying to help you.

Quote:

---

Originally Posted by matsp

I just pointed out at least part of what doesn't work with the current code. Someone else pointed out that you need to have a separator between your numerbs.

You don't REALLY expect US to write the code FOR YOU, do you?

--
Mats

---

Remember what I posted is just part of my whole code, the #include<stdlib.h> was needed for some other parts of my program that's why you see it appear there.

Quote:

---

Originally Posted by Cactus_Hugger

That looks like bad postfix to me. What's to say it isn't 1 * 315.5 + 8 or 131 * 5.5 + 5 ? You need something to separate your numbers, like a space. So:

Code:

---

131 5.5*8+

---

Might be valid, but with just

Code:

---

1315.5*8+

---

We have no idea where to break up the number.

---

Should I seperate it at the infix level or after the convertion to postfix. Cos I'm wondering if I do it at the infix level whether the postfix will appear with spaces.

I don't think you understood the original comment abotut spaces. As I understand it (and the code appears to support this) your code READS postfix expressions. So if the expression is 1213+, how do you know which of the following options:
1 + 213
12 + 13
121 + 3
1213 + <nothing> [incorrect number of arguments for add, but still possibly what the user typed]

So the input format will have to require spaces (or some other separator, but space is the natural choice, I would think), so we type 12 13+.

Your conversion from "char to int" is currently only written to cope with one digit, so you'll definitely need to fix that - as I pointed out in a previous post. To support floating point, you'll also have to consider the decimal point and do the conversion of anything after the decimal point correctly, which makes for a slightly more complicated function, but there are some standard functions to help you.

Oh, and I was pointing out about "stdio.lib", not "stdlib.h" - the latter isn't in your code, but the former is. This may of course be a consequence of you only showing us part of your code and you may have edited it by hand when adding it to the forum - but as it stands, it would definitely not compile.

--
Mats

Quote:

---

Originally Posted by matsp

I don't think you understood the original comment abotut spaces. As I understand it (and the code appears to support this) your code READS postfix expressions. So if the expression is 1213+, how do you know which of the following options:
1 + 213
12 + 13
121 + 3
1213 + <nothing> [incorrect number of arguments for add, but still possibly what the user typed]

So the input format will have to require spaces (or some other separator, but space is the natural choice, I would think), so we type 12 13+.

Your conversion from "char to int" is currently only written to cope with one digit, so you'll definitely need to fix that - as I pointed out in a previous post. To support floating point, you'll also have to consider the decimal point and do the conversion of anything after the decimal point correctly, which makes for a slightly more complicated function, but there are some standard functions to help you.

Oh, and I was pointing out about "stdio.lib", not "stdlib.h" - the latter isn't in your code, but the former is. This may of course be a consequence of you only showing us part of your code and you may have edited it by hand when adding it to the forum - but as it stands, it would definitely not compile.

--
Mats

---

I have written a code convert from infix to postfix and when I input 12+13 it gives me the postfix as 1213+ and that's why I'm asking the particular point I need to do the tokenization so I have the postfix as 12 13+?

Quote:

---

Originally Posted by ben2000

I have written a code convert from infix to postfix and when I input 12+13 it gives me the postfix as 1213+ and that's why I'm asking the particular point I need to do the tokenization so I have the postfix as 12 13+?

---

But that's not the code you've presented in this thread - it reads a string containing postfix operations, so that's where the separators need to be (infix, you can use the operator as a separator, so you don't need to do anything there). If you compare it with a HP calculator, you have to hit enter between two numbers, then hit + to add them. I'm not sure if the "infix to postfix" is part of this same application (but not shown here), or if it's a separate application. Either way, your string that is input to your postfix parser is the part that needs the separation!

--
Mats

Quote:

---

Originally Posted by matsp

But that's not the code you've presented in this thread - it reads a string containing postfix operations, so that's where the separators need to be (infix, you can use the operator as a separator, so you don't need to do anything there). If you compare it with a HP calculator, you have to hit enter between two numbers, then hit + to add them. I'm not sure if the "infix to postfix" is part of this same application (but not shown here), or if it's a separate application. Either way, your string that is input to your postfix parser is the part that needs the separation!

--
Mats

---

ok I have succeded in making my infix code produce postfix delimited by spaces, so 13.5+25 will be 13.5 25 + Can you guide me or someone guide me on how to construct the loop cos I need a loop to read through the string and convert it to double before pushing it. This has been difficult for, please assist me.

Code:

---

#include<stdio.h>
#include<stdlib.h>
#define MAXCOLS 80
main(){
          char instring[MAXCOLS], postring[MAXCOLS];
          int position = 0;
          float eval();
          float result;
          printf("%s", "Enter postfix expression: ");
          while ((postring[position++] = getchar()) != '\n');
          postring[--position] = '\0';
          printf("%s%s", "postfix expression is", postring);
          getchar();
          result=eval(postring);
          printf("%s%f\n", "value is", result);
          getchar();
}
struct stack{
                  int top;
                  float items[MAXCOLS];
                  };
float eval(expr)
char expr[];
                    {
                    int c, position;
                    float op1, op2, value;
                    float oper(), pop();
                    struct stack opndstk;
                    opndstk.top = -1;
                    for (position =0; (c=expr[position]) != '\0'; position++)
                    if(isdigit(c))
/*I guess this is the point I need a loop to convert to double, if I have 44 55 +, I need to read in 44 convert it to double push it, skip the space and read 55, this code below should do the conversion but I don't know how to integrate it.
 
double val, power,res;
int sign;
sign = (expr[i] == '-') ? -1 : 1;
for (val = 0.0; isdigit(expr[i]); i++)
val = 10.0 * val + (expr[i] - '0');
if (expr[i] == '.')
{
i++;
}
else
{
}
for (power = 1.0; isdigit(expr[i]); i++)
      {
      val = 10.0 * val + (expr[i] - '0');
      power *= 10;
      }
      res= sign * val / power;
push(&opndstk, res)*/

            push (&opndstk, (float)(c-'0'));
else {
        op2 = pop(&opndstk);
        op1 = pop(&opndstk);
        value = oper(c, op1, op2);
        push (&opndstk, value);
        }
      return (pop(&opndstk));
}

float oper(symb, operand1, operand2)
int symb;
float operand1, operand2;
        {
          switch(symb){
                              case '+': return (operand1 + operand2);
                              case '-': return (operand1 - operand2);
                              case '*': return (operand1 * operand2);
                              case '/': return (operand1 / operand2);
                              default: printf("%s", "illegal operation");
                              exit(1);
                            }
        }

push(ps, x)
struct stack *ps;
float x;
    {
      if(ps->top == MAXCOLS-1)
            {
            printf("%s", "stack overflow");
            exit(1);
            }
else
      ps->items[++(ps->top)] =x;
      return;
      }

float pop(ps)
struct stack *ps;
    {
    int top;
if (top==-1)
          {
printf("%s", "stack underflow");
exit(1);
          }
else
      return(ps->items[ps->top--]);
    }

---

Yes, that looks about right.

--
Mats

I tried replacing the i's in the red loop with position and removed "push (&opndstk, (float)(c-'0'));" but it didn't run. Mat please help me and work me through. Remember the postfix was delimited by spaces.

You will need to debug the code - try adding a few printf's in the code that deals with a number - perhaps printing the part of the string being parsed, for example.

--
Mats

this was what I did but it didn't work. Can you help me and study it to see what I'm missing?

Code:

---

#include<stdio.h>
#include<stdlib.h>
#define MAXCOLS 80
main(){
          char instring[MAXCOLS], postring[MAXCOLS];
          int position = 0;
          float eval();
          float result;
          printf("&#37;s", "Enter postfix expression: ");
          while ((postring[position++] = getchar()) != '\n');
          postring[--position] = '\0';
          printf("%s%s", "postfix expression is", postring);
          getchar();
          result=eval(postring);
          printf("%s%f\n", "value is", result);
          getchar();
}
struct stack{
                  int top;
                  float items[MAXCOLS];
                  };
float eval(expr)
char expr[];
                    {
                    int c, position;
                    float op1, op2, value;
                    float oper(), pop();
                    struct stack opndstk;
                    opndstk.top = -1;
                    for (position =0; (c=expr[position]) != '\0'; position++)
                    if(c== '  '){
                position++;
                                }
                    if(isdigit(c)){
                              double val, power,res; int sign;
                    sign = (expr[position] == '-') ? -1 : 1;
                    if (expr[position] == '+' || expr[position] == '-')
                    position++;
                    for (val = 0.0; isdigit(expr[position]); position++)
                 
                    val = 10.0 * val + (expr[position] - '0');
                    if (expr[position] == '.')
                    {
                                position++;
                                }
                                else
                                {
                                }
                    for (power = 1.0; isdigit(expr[position]); position++)
                        {
                          val = 10.0 * val + (expr[position] - '0');
                          power *= 10;
                          }
                    res= sign * val / power;
                    push(&opndstk, res);
                  }
      else if ( c  == '*' || c  == '+' || c  == '/' || c  == '%' || c  == '-' )
        {
        op1 = pop ( );
        op2 = Pop ( );

        switch(c)
                {
                case '+' : value= op2+op1; break;
                case '-' : value= op2-op1; break;
                case '*' : value= op2*op1; break;
                case '/' : value= op2/op1; break;
                }
                push (&opndstk, value);
        }
result = pop ( );
return result;
}

---

Yes, so you need to debug it. It's no point in me debugging it, because you will just take the fix I've written and paste it into your code. Learning to debug your own (or someone elses!) code is definitely something you need if you plan to ever do any real programming.

There are a few main ways to debug code:
1. Read the code and step by step follow what happens. This can be time-consuming. This works well if you have a small piece of code where you KNOW it's broken in that piece.
2. Use a debugger and step through the code line-by-line and follow what happens. Again, best if you know which part of the code is broken.
3. Add debug-code to print (and perhaps verify assumptions) important data in a suspect piece of code. As long as the added printout isn't TOO lengthy, it's usually a very quick way to check out where things go wrong.
4. Ask someone else to do it for you. If you have a manager position or a big checkbook, this method works, but on a "hobby" basis, it's perhaps not such a good idea. The main drawback here is that you don't learn anything from your mistakes (if you have debugged a problem for 6 hours (or a few weeks), you WILL NOT make that mistake again anytime soon! - if someone else spent hours to figure out what is wrong, you will probably just apply the fix and go on making similar mistakes again).

--
Mats

when I enter the postfix 34 25 + it works, result 59.0000, but when I enter 12 2 * 12 + it gives 24.0000. It seems to stop after executing the first operator. Again if I have 2.5 2 * it doesn't work at all. which means the '.' is causing some problem but in the code I have provision for point. Just help me and figure out where the problem is.

One subtle error is the "if (c == ' ')" above - it has two spaces in it, which probably means that it's ALWAYS going to be false.

You may want to add a printf to see what digits you are processing in the "after point" code - are they valid digits, is it what you think it should be, etc, etc?

--
Mats

I've sorted the '.' stuff. The only problem I've got now is the loop, it stops after executing the first operator. Any suggestions?

Hmm, I'm actually surprised that you get as far as the first operation, as I don't see anything that would "repeat" aside from the loops to make a number.

--
Mats

I don't understand what you said, but I'm saying that if enter a "2 3 * 4 +" it gives me the result as 6 i.e after executing the fisrt operator, I guess I'm missing out something in the loop, but I can't figure out what it is. Or there a way I could reconstruct the loop?

Can someone help me with the loop. when I enter the postfix 2 3 * 4 +, it gives me result as 6, i.e after executing the first operator and stops. It doesn't get to 4 and +.

What I was trying to say is that I can't see anything that _WILL_ make your eval() function work to even do two operands and an operation. But maybe the code you've posted isn't acutally the code you're running.

--
Mats

The only thing missing in the code above is the pop() and push() which is on page 1 of this tread. You can run it and see, you know what I want to do, and how to implement it. Just run it and see, if you've got suggestions then you can give concerning the loop.

Here is the code put together again, please if anyone has got suggestions about the loop, please make. It works for 2 3 + but not more than that. say 2 3 * 4 +

Code:

---

#include<stdio.h>
#include<stdlib.h>
#define MAXCOLS 80
main(){
          char instring[MAXCOLS], postring[MAXCOLS];
          int position = 0;
          float eval();
          float result;
          printf("&#37;s", "Enter postfix expression with space as delimiter: ");
          while ((postring[position++] = getchar()) != '\n');
          postring[--position] = '\0';
          printf("%s%s", "postfix expression is", postring);
          getchar();
          result=eval(postring);
          printf("%s%f\n", "value is", result);
          getchar();
}
struct stack{
                  int top;
                  float items[MAXCOLS];
                  };
float eval(expr)
char expr[];
                    {
                    int c, position;
                    float op1, op2, value;
                    float oper(), pop();
                    struct stack opndstk;
                    opndstk.top = -1;
                    for (position =0; (c=expr[position]) != '\0'; position++)
                    if(isdigit(c)){
                                                double val, power,res; int sign;
                    sign = (expr[position] == '-') ? -1 : 1;
                    if (expr[position] == '+' || expr[position] == '-')
                    position++;
                    for (val = 0.0; isdigit(expr[position]); position++)
                 
                    val = 10.0 * val + (expr[position] - '0');
                    if (expr[position] == '.')
                    {
                                position++;
                                }
                                else
                                {
                                }
                    for (power = 1.0; isdigit(expr[position]); position++)
                        {
                          val = 10.0 * val + (expr[position] - '0');
                          power *= 10;
                          }
                    res= sign * val / power;


push(&opndstk, res); 

  }      //  push (&opndstk, (float)(c-'0'));
else {
        op2 = pop(&opndstk);
        op1 = pop(&opndstk);
        value = oper(c, op1, op2);
        push (&opndstk, value);
        }
      return (pop(&opndstk));
}

float oper(symb, operand1, operand2)
int symb;
float operand1, operand2;
        {
          switch(symb){
                              case '+': return (operand1 + operand2);
                              case '-': return (operand1 - operand2);
                              case '*': return (operand1 * operand2);
                              case '/': return (operand1 / operand2);
                              default: printf("%s", "illegal operation");
                              exit(1);
                            }
        }

push(ps, x)
struct stack *ps;
float x;
    {
      if(ps->top == MAXCOLS-1)
            {
            printf("%s", "stack overflow");
            exit(1);
            }
else
      ps->items[++(ps->top)] =x;
      return;
      }

float pop(ps)
struct stack *ps;
    {
    int top;
if (top==-1)
          {
printf("%s", "stack underflow");
exit(1);
          }
else
      return(ps->items[ps->top--]);
    }

---

Can you try to answer as many of these questions as possible:

What compiler are you using?
Does it support enabling warnings?
If so, please do enable warnings.
Does it support ANSI standard?
If so enable it, and fix up your code to use ANSI.


Your string-to-float code uses double, whilst the rest of the code is using float - is there a reason for that?

The pop operation has an uninitialized variable to check for top == -1, which means that if you get a stack underflow, it's not going to be detected (and if you are really unlucky, the uninitialized variable turns into -1 by chance, and you get a stack-underflow when there isn't one).

What happens if you see a "space" in the expr in eval?

Does this look about right?

Code:

---

Enter postfix expression with space as delimiter: 23 45 + 18 * 19 /
postfix expression is: 23 45 + 18 * 19 /
value is 64.421051

---

--
Mats

Mat I'm using dev-c++, I've just enabled checked the compiler settings as folows:
*Support all ANSI standard C programs---"YES"
*Inhibit all warning messages---"NO"
*Display one error per line ---"YES"
*Attempt to support some aspects of traditonal C pre----"YES"

I used if(c==' ') position++; in the program though I missed it out on the page here.

Could these be the problem? Is the loop ok as it is? Please put me through, I've really toiled over this.

Now fix all the warnings that the compiler gives!

It is VERY GOOD PRACTICE to use compiler warnings to tell you when you do something wrong in the code. I made your code compile on MS VC 7 without warnings, so you should be able to too. And NO I'm not going to just give you the code.

You probably want to look over where you do your check for space, and how you do that. I think that will fix your problem. But I want (as thanks for the help) see your code when you're finished, and I want it to NOT give any warnings when I compile it.

--
Mats

I've still tried it but it's not working, only for two operands and one operator

Add more code to debug the problem then. Print what is happening at each stage - you will sooner or later find where it's going wrong.

Note: You should aim to change the code as little as possible when adding these prints. Just going round making random changes to the code isn't really that helpfull unless you have a fair idea what you're trying to fix.

Is your code compiling without warnings now?

--
Mats

It compiles without warnings, or any error messages. You said something about space
I used if(c==' ') position++; before the isdigit(), is it correct? Again I ask is the loop I used ok. Just help me out so I can move forward, I have some other code to write, this just got me stucked for days now. Please so I can move forward.

I haven't added any loop to your code, but you probably don't want to just update the position, as "c" is still a space after that. It's probably better to leave the "position++" in one place in the for-loop (at least that's what I did). And I didn't do much more than that (and remove some unused variables, fixed the stack underflow detection problem I told you about earlier).

--
Mats

I've amended th pop as follows I guess that should check for the underflow, but still it's not working

Code:

---

float pop(ps)
struct stack *ps;
    {
   
if (ps->top==-1)
          {
printf("%s", "stack underflow");
exit(1);
          }
else
      return(ps->items[ps->top--]);
    }

---

Are you saying the position++ is not supposed to appear under the if(c==' ') ?

Quote:

---

Originally Posted by matsp

I haven't added any loop to your code, but you probably don't want to just update the position, as "c" is still a space after that. It's probably better to leave the "position++" in one place in the for-loop (at least that's what I did). And I didn't do much more than that (and remove some unused variables, fixed the stack underflow detection problem I told you about earlier).

--
Mats

---

But the position++ is in the for-loop, I don't get you hear, are u saying I should remove the one after the if(c==' ') ?

Mats, I got it thanks for your time but try to be more lenient.

I'm trying to teach you how do solve the problems - because me solving them is a few minutes because I've solved MANY similar problems.

I'm sure you've spent a lot of time on this, but in all honesty, if you had some experience in debugging your own code, you wouldn't have spent as much time - and that's what I'm trying to teach you.

If I just say add a "if (c != ' ')" before the else, you won't learn how to debug, you will just have code that "happens to work because someone else solved that particular problem".

So does "23 45 + +" work "correctly" now?

--
Mats

You are right, it's working properly now, but I still have a problem and that's the next stage in my code. You know I'm using numbers as direct infix input. I want to use variable instead of numbers as infix. Say I have red*blue+green as infix, where these are red, blue, green are variables representing array columns for n*3 array. Get the row values and use it to substitute the variables in the infix before proceding to postfix. That's the next thing I'm trying to do. I hope you guide me through.

Doesn't sound too difficult.

Start with getting a piece of code that can recognise a variable, then add substitution for the actual value of the variable - the rest should be pretty easy.

I would keep a very simple data structure for the variables, just a string and the value.

--
Mats

ok, I'm on it now

this is what I came up with so far but my but being that the array is double it's a bit difficult getting the elements into string expression and passing it to my infix.

Code:

---

#include<stdio.h>
#include<stdlib.h>
#define MAXCOLS 80

int main()
{
  const int num_operand =3;
    const int num_operatr = 2;
  const int strglen=num_operand+num_operatr;
    char operand;
    int i, j;
    char string[MAXCOLS];
    char getexp[strglen+1];
    double array[3][3]={23.4,13,12,3,6.5,21,5,4,6};
               

    printf("Enter an expression:\n");
    fgets(getexp, sizeof getexp, stdin);

    for(j=0; getexp[j] != '\0'; j++)
    {
      operand=getexp[j];
     
      if(operand >= 'A' && operand <='Z')
          string[j]=array[operand-(int)('A')];
      else
          string[j]=(char)operand;
    }
    string[j] ='\0';

    printf("Expression\n");
    printf("&#37;s", string);
   
    getchar();
    return 0;
 }

---