how do compiler differentiate strings and intigers?

[tonysonney]

i was programming an assembler as part of my assignment. In that to check errors in the program, i wanted to check whether the argument is an intiger or a character? Can any one help me how to do that?
Thanks in advance

[horace1]

i was programming an assembler as part of my assignment. In that to check errors in the program, i wanted to check whether the argument is an intiger or a character? Can any one help me how to do that?
Thanks in advance

this is not simple as once a character has been read into memory it is represented by an integer character code such as ASCII, see
http://www.asciitable.com/

you can check if an integer represents a character code such as a digit, alphabetic, etc using functions such as isdigit(), isalpha(), see
http://www.cppreference.com/stdstring/index.html

however, although decimal 65 is the ASCII code for A it could have been entered as an integer

[DeMan]

a byte is a byte is a byte is a byte.....

(Incedently a character is 1 byte (2^8 or 0 - 255 ) and an integer is 2 bytes (2 ^16 or 0 - 65536 ) but that does't help because you can't necessarily tell the difference between 1 integer or 2 characters).

Compilers are very trusting beasts. They store all the variables you declare in a look-up table. When they come accross a variable, they look it up in the table and check that it's type is the same as what you said before (this is about where you start getting warnings (sometimes errors, but usually warnings I think)that there are problems with types).

Assuming someone has defined a var to be something beforehand, I would reccommend you do something similar, store the variables, their types (and possibly value) in a table and make sure their type is never changed.



Slightly (a lot perhaps) off topic (as always): This is why casting can be a bad idea unless you know for a fact you want to. The cast get's rid of the warning/error because the compiler says "This doesn't look right...wait a minute, there's an explicit cast so this is derliberate...I'll assume this chap knows what he's doing". You should always be very careful in casts, because even though the compiler can see what you are doing is strange, by adding a cast you are assuring it that you definitely intend to use that value as a different type.

[Ganon11]

Incedently a character is 1 byte (2^8 or 0 - 255 ) and an integer is 2 bytes (2 ^16 or 0 - 65536)

Strictly speaking, an integer's byte value varies from system to system. Your 2 byte integer is specifically known as a short int, whereas there can be a 4 byte integer (0 to 4,294,967,296 (unsigned)) known as a long int. Using just int will use your system default - for mine, it is the long int.

But a char is always 1 byte.