Maybe A Cast Problem?

Ok I have this code, which I want it convert the string answer into all lowercase letters. I'm using strlwr which you have to use a character array, but when I try to use a character array, it can't be a string(obviously) but when I check if it equals "roll" it doesn't work.

Anyway, i'm not too sure about how to use a cast in this situation, and I would rather not have to, but if it is the only way to get around this, I guess it will have to do.

Code:

---

int main()
{
srand(time(NULL));
string answer;

do
{
cout<<"Enter A Command";
  cin>>answer;
  cin.ignore();

strlwr(answer);

 if(answer=="roll")
  {
    roll_dice();
  }

}
while(1);

cin.get();
return 1;
}

---

Whoops. Forgot the error it gives me.

Code:

---

cannot convert `answer' from type `string' to type `char *'

---

Make your own function:

Code:

---

#include <string>
#include <algorithm>

void char_to_lower(char& c)
{
  // if c is an upper letter
  if(c >= 65 && c <= 90)
    c += 32; // shift to lower character
}

void string_to_lower(std::string& myStr)
{
  std::for_each(myStr.begin(), myStr.end(), char_to_lower);
}

---

Haven't compiled, it should be something like this though.

Quote:

---

Originally Posted by relyt_123

Anyway, i'm not too sure about how to use a cast in this situation, and I would rather not have to, but if it is the only way to get around this, I guess it will have to do.

---

It's good that you see casting as a last option, because it generally it is. Mostly because it is the incorrect choice most of the time -- like this.

Quote:

---

Originally Posted by Desolation

Code:

---

void char_to_lower(char& c)
{
  // if c is an upper letter
  if(c >= 65 && c <= 90)
    c += 32; // shift to lower character
}

---

---

Please try not to be system specific if it isn't necessary.

Hmm...never looked at it that way, but it does work, so thanks man.

Huh, I thought char defaulted to ASCII =/ So you're saying char could be something else than ASCII like UTF-8 or Unicode ? =/

just use what C++ already offers you. Why complicate?

Code:

---

#include <cctype>

char x = 'A';
x = tolower(x);

---

Mario<<Will that work with strings? I'll try it, and go ahead and use that instead if it does, but for now, i'll just use desolations way.

Anyway I'm back with another question:

Code:

---

int main()
{
srand(time(NULL));
string answer;

do
{
cout<<"Enter A Command: ";
  cin>>answer;
  cin.ignore();

string_to_lower(answer);

 if(answer=="roll")
  {
  int numberhold;
  do
    {
    roll_dice();

      cout<<"How Many Dice Would You Like To Hold?";
      cin>>numberhold;
      cin.ignore();

      switch(numberhold)
      {
        case 1:cout<<"test";
        break;
      }
      }
//Line 64
    }

}
while(1);

cin.get();
return 1;
}

---

It seems that there is an error in the switch statement, but I can't seem to find one.

Error is:

Code:

---

parse error before `}' Line 64

---

Edit: Nevermind, stupid mistake, I didn't put the while at the end...

Edit2:Also I can't get the tolower function to work, it won't go through the if statement if I use it. Hmm, what do you put in the ()'s? Because whatever you put there it print's as an ascii value.

Reusing some code in this thread, to make a string lower case

Code:

---

void string_to_lower(std::string& myStr)
{
  std::for_each(myStr.begin(), myStr.end(), tolower);
}

---

Watch for the problem which bubbled up in this thread.

http://cboard.cprogramming.com/showt...hlight=tolower

That was meant as a reply to Desolation.

If you want a less complicated way to change a string case you can use the transform algorithm.

Code:

---

#include <cctype>
#include <string>
#include <iterator>
#include <algorithm>

std::string str = "Hello World";
std::transform( str.begin(), str.end(), str.begin(), toupper );  // all upper case
std::transform( str.begin(), str.end(), str.begin(), tolower );  // all lower case

---

EDIT: Tonto's option is even easier to read

Ok, thanks it's working perfectly in the aspect of making it lowercase, but otherwise...who knows?

>std::for_each(myStr.begin(), myStr.end(), tolower);
>std::transform( str.begin(), str.end(), str.begin(), tolower );
Both of these contain a potential ambiguity that arises due to multiple overloads of tolower in the presence of locales. The chances of hitting the ambiguity are pretty good, even though many compilers will do the right thing. A better solution is to provide a wrapper around tolower that explicitly calls the correct overload:

Code:

---

struct lower_case {
  int operator() ( int ch )
  {
    return tolower ( ch );
  }
};

std::for_each(myStr.begin(), myStr.end(), lower_case());
std::transform(str.begin(), str.end(), str.begin(), lower_case());

---

True.

Or

Code:

---

std::transform( str.begin(), str.end(), str.begin(), (int(*)(int)) tolower );

---

>Or
No, now you have a problem due to the C functions having "extern C" linkage and transform expecting C++ linkage. Yea, it's messed up, but that officially makes your casting solution non-portable.

Well, thanks for that prelude. Didn't know about that. Will definitely go with the functor whenever I need this.

Ok, new question, how would you check if an int has a char in it in a switch statement? I have this so far.

Code:

---

      switch(numberhold)
      {
        case 0:break;

        case 1:cout<<"int = 1";
        break;

        case 2:cout<<"int = 2";
        break;

        case 3:cout<<"int = 3";
        break;

        case 4:cout<<"int = 4";
        break;

        case 5:cout<<"int = 5";
        break;

        default:cout<<"int <0 or >5";
        break;


      }

---

An integer cannot have characters. What do you want ? You want to find if a certain digit is in a certain number ?

Code:

---

// 1 means found
// 0 means not found
// -1 means error
int search_digit(int num, int digit)
{
  if(num == 0 && digit == 0) return 1;
  if(digit >= 10 || digit < 0) return -1;
  while(num > 0)
  {
    if((num % 10) == digit) return 1;
    else num /= 10;
  }
  return 0;
}

---

You don't even need a switch for that. It can be accomplished in a simple pair of if statements:

Code:

---

        if ( numberhold >= 'A' && numberhold <= 'Z' )
                std::cout<<"uppercase character is "<<static_cast<char>(numberhold)<<std::endl;
        else if ( numberhold >= 'a' && numberhold <= 'z' )
                std::cout<<"lowercase character is "<<static_cast<char>(numberhold)<<std::endl;

---

Quote:

---

Originally Posted by Desolation

An integer cannot have characters. What do you want ? You want to find if a certain digit is in a certain number ?

Code:

---

int search_digit(int num, int digit)
{
  if(digit >= 10 || digit < 0) return -1;
  while(num > 0)
  {
    if((num % 10) == digit) return 1;
    else num /= 10;
  }
  return 0;
}

---

---

integers can be represented as characters, by interpreting them with their equivelant ascii value; http://www.lookuptables.com/

No, sorry I explained it wrong. I want to see if the user inputs any thing other than an integer it says an error message. Hold on I have an idea.

Edit:Nope I tried this but it gives a fatal error every time and crashes the program.

Code:

---

      if(isdigit(numberhold)!=1)
        {
        cout<<"Error, Please Input A Whole Number Between 1 and 5";
        }

---

I know that. However, integers cannot contain characters. That is my point.

Code:

---

std::string tmp;
std::cin >> tmp;

for(int i = 0; i < tmp.size(); i++)
{
  if(!isnum(tmp[i]))
  {
    // not a number, do something
  }
}

---

Implicit Declaration of function 'int isnum(...)'

Oh well, can't find what header has it.

Code:

---

inline bool isnum(char c)
{
  return (c >= 48 || <= 57);
}

---

Hmm...well now if I input anything it just exits the program.

>Oh well, can't find what header has it.
That's because isnum isn't a standard function. You're thinking of isdigit from <cctype>. An isnum function would be particularly awkward because it implies that any legal character from any numeric type would return true. That includes non-digit characters that are valid in certain numeric representations such as floating-point and hexadecimal.

>return (c >= 48 || <= 57);
That's kind of broken. ;) You forgot to put c in the second half of the expression. The logic is also faulty because if c is greater than 48, regardless of how much greater, the test will return true. Something like this would be better:

Code:

---

inline bool isnum(char c)
{
  return (c >= 48 && c <= 57);
}

---

You also have two portability issues. The first is that isnum is a reserved name (specifically, any name that begins with "is" and is followed by at least one lower case letter). The second issue is that you assume the values of the digits in your character set. You also assume that they're consecutive, which is okay in this case, but you can't make the same assumption for the rest of the character set. The good news is that you can write this function portably:

Code:

---

bool is_num ( char ch )
{
  return '0' <= ch && ch <= '9';
}

---

Then again, isdigit gives you the same solution without having to worry about locales either. ;)

Huh, I think I had a brain fart there and completely changed isdigit to isnum =/ I think I was pretty tired because that function now looks quite stupid to me with the binary OR and the missing 'c' =/ Oh well, thanks for pointing out =]

case 'a': blabla?