[G A B R I E L]

I have this code:

 if (aChar == '0' || '1' || '2' || '3' || '4' || '5'|| '6' || '7' || '8' || '9')
		return DIGIT  // (I've defined DIGIT as something else I need to use later

Basically my question is

(1) Is there an easier way to do that i.e. if I were to do it with say, the upper case alphabet would I have to write out every single letter (or is there is an easier way)

(2) If I just wanted to return DIGIT for digits, and say, return SOMETHING I(which i've defined as somethign) for anything thats not a digit (space, underscore, uppercase, lowercase, etc), how would I do that

[mwnciau]

(1) You could use regular expressions

(2) Expanding on (1), you could use regular expressions to check if the char is a digit in which case return DIGIT, otherwise return SOMETHING

[Psyk]

You could do the check based on numeric ASCII values. So just check that aChar is between 48 and 57 (inclusive) as they are the ASCII values for the characters 0 and 9.

[G A B R I E L]

Alright, thanks I get that. But I've defined aChar as char, so wouldn't I need to do: int(aChar) to get the ASCII value?

[G A B R I E L]

Originally Posted by spikeymike

See code edit. You put a cast around the number.

My original code was wrong.

What does that do? (as opposed to what I suggested)?

[spikeymike]

Originally Posted by G A B R I E L

What does that do? (as opposed to what I suggested)?

It's the equivalent to yours but for some reason it doesn't work properly. Forget my code... this works however.

@Psyk: Why did you mention ASCII? More complicated than what it should have been...

#include <iostream>
using namespace std;

int main ()
{
	int aChar;
	cin >> aChar;
	if (aChar >= 0 && aChar <= 9)
		cout << "ok";

	system("PAUSE");
}

Basically, if aChar is greater than or equal to 0 and less than/equal to 9 then print "ok".

[G A B R I E L]

I did think about that but later on in the program I need to consider the whole uppercase alphabet so I guess resorting to ASCII would be what needs to be done. But as I said I wasnt sure if int(aChar) is the right way

[spikeymike]

Originally Posted by G A B R I E L

I did think about that but later on in the program I need to consider the whole uppercase alphabet so I guess resorting to ASCII would be what needs to be done. But as I said I wasnt sure if int(aChar) is the right way

Ok, going back to the ASCII stuff. And you're right aChar cannot be int if you're using Psyk's suggestion.

Your aChar has to be a char, and the number that represents a particular letter/number etc has to be put into a cast..

char aChar;
cin << aChar;

if (aChar == char(101))   // checks for 'e'
   // do this

[G A B R I E L]

Originally Posted by spikeymike

Ok, going back to the ASCII stuff. And you're right aChar cannot be int if you're using Psyk's suggestion.

Your aChar has to be a char, and the number that represents a particular letter/number etc has to be put into a cast..

char aChar;
cin << aChar;

if (aChar == char(101))   // checks for 'e'
   // do this

Alright thanks, that works.

[DFranklin]

Originally Posted by spikeymike

Basically, if aChar is greater than or equal to 0 and less than/equal to 9 then print "ok".

Are you sure your code works? You want

if (aChar >= '0' && aChar <= '9') // note the '0' and '9'
...

surely?

And for the alphabet:

if (aChar >= 'A' && aChar <='Z') etc...

Edit: you might also want to google isalpha and isdigit.

[spikeymike]

Originally Posted by DFranklin

Are you sure you're code works? You want

if (aChar >= '0' && aChar <= '9') // note the '0' and '9'
...

surely?

And for the alphabet:

if (aChar >= 'A' && aChar <='Z') etc...

Edit: you might also want to google isalpha and isdigit.

The bit in orange... code works, tried it before posting.

The bit in bold... I doubt very much that would work. I've always used ASCII if I'm validating a letter.

[DFranklin]

Originally Posted by spikeymike

The bit in orange... code works, tried it before posting.

Actually it doesn't, not really.

I hadn't noticed that aChar was an int (nice choice of variable name - this is why I hate naming schemes that include the variable type).

So you're not actually doing what was asked - the input is being interpreted as decimal - you're not actually doing anything based on the value of the characters entered. (To be clear, the OP has a variable aChar of type char. If you were to write "int aInt = aChar;" and then "if (aInt >=0 && aInt <=9) ..." you'd get the wrong result).

If you enter 10, it won't say OK, which I very much doubt is the desired outcome.

Also, if the user doesn't enter an digit, then aChar doesn't get written to, so you're left with what ever was in there before.

So it *can* still say OK even if you don't enter a digit. (Try initialising aChar to 5 before the call to cin to see this happen).

The bit in bold... I doubt very much that would work. I've always used ASCII if I'm validating a letter.

It works fine (aChar has to be a char, (or an int created by casting a char to an int)) of course. It's arguably more portable, although it still goes horribly wrong for something like EBCDIC where the alphabet isn't a single contiguous range.

[Psyk]

I'm sure you can do a comparison between a char and an int without any casting. But I guess it's sensible to do the cast even if it's not strictly necessary.

Now I think about it, you don't really need to use the numeric ASCII values, just checking that it's between '0' and '9' should do (note the quotes, they mean the character '0' and not the number 0). The same should work for 'a'...'z' and 'A'...'Z', assuming the platform you're working on uses ASCII (as DFranklin says, this might not work for some character encoding schemes).

[DFranklin]

If the "cast" comment was directed at me, let me clarify:

If you have a variable aChar of type char, then the comparison "aChar >=0 && aChar <=9" does NOT correctly identify if aChar is a decimal digit (assuming the ASCII character set).
The correct comparison is "aChar >= '0' && aChar <= '9'".

The reason spikeymike's code works is because the code "cin >> aInt; " (where aInt is an integer) actually parses the input to create an integer. So the intent of me saying "int created by casting a char to an int" was to contrast casting the input with using cin to parse it. You don't need an explicit cast, but "int aInt = aChar;" still involves a cast operation, even if it's implicit.

[Not that the details hugely matter - it was the contrast with cin I was trying to emphasize].

[Psyk]

Oh ok that makes sense. But am I right in thinking that something like this is valid?

char c = '0';
if(c == 48)
{
    return true;
}

[gregmantis]

Originally Posted by Psyk

Oh ok that makes sense. But am I right in thinking that something like this is valid?

char c = '0';
if(c == 48)
{
    return true;
}

Yes. A char is basically an integer type where sizeof(char) is guaranteed to be 1 (though whether it is signed or unsigned by default is implementation specific).

Here's an interesting little program:

extern "C" int printf( const char*, ... );

int main()
{
  char c = 0;

  do
  {
    printf( "%d '%c'\n", c, c );
  } while ( ++c );
}

[gregmantis]

Originally Posted by gregmantis

extern "C" int printf( const char*, ... );

int main()
{
  char c = 0;

  do
  {
    printf( "%d '%c'\n", c, c );
  } while ( ++c );
}

What's also quite interesting (subjective) is if you play with different types for c.

extern "C" int printf( const char*, ... );

int main()
{
  unsigned short c = 0;

  do
  {
    printf( "%d %d '%c'\n", c, (char)c, c );
  } while ( ++c );
}

[DFranklin]

@gregmantis: "Most" interesting if c is a double, of course...