Why are @ and $ characters not used for anything in C and C++?

Question

In both languages the basic source character set includes every printable ASCII character except @, $ and `. I can understand not using grave accent because it's not always interpreted as a separate character and it also looks very similar to apostrophe. But is there a specific reason why @ and $ don't have any usage or did the language designers just run out of ideas? :)

Answer 1

I can't imagine what capacity they would fill. Perhaps using @ to signify pointers...

But $ and @ are very busy looking symbols, perhaps almost distracting, and if you throw them into the mix with an already diverse syntax, just because they're there, you might end up having a language that reads like a perl regex. Which is to say it doesn't read at all. :P

Answer 2

@ was a bad idea because it was the kill character. If you were typing in a program and you accidentally hit @ then you erased the entire line of input up to that point.

# was more or less a bad idea because it was the erase character. If you were typing in a program and you accidentally hit # then you erased the most recent character.

When the preprocessor was added to the C language, # was accepted in the first column of a line, but not anywhere else. So maybe ed was modified to allow # to be input as the first character of a line, since there was nothing before it to be erased.

So why didn't the preprocessor use $ instead of #? Here we go, I answered half of your question but added to the other half of your question.

Newspaper articles didn't used to include the @ character. After the internet became common, some reporters or editors put the 4-character string "(at)" in newspaper articles because they couldn't or wouldn't use some escape sequence to put an actual @ in the article. Unix's definition of the kill character @ was copied from newspaper equipment.

http://en.wikipedia.org/wiki/Seventh_Edition_Unix_terminal_interface

Answer 3

The first question you should ask is, "Why are only certain characters allowed in C/C++ function and variable names"?

Even I am not quite old enough to answer that... But I would bet many special characters (especially $) were not legal in external symbols in the original Unix. That is, the assembler and linker would choke on them.

So the only use for non-alphanumeric characters was in operators, like + or ->. The original designers presumably had all the operators they needed, so there was no reason to use $ or @ or whatever. (How do you do the back-tick markdown, anyway?)

With the advent of C++ and name mangling, most restrictions on identifier names could presumably be lifted. But even the C++ committee is not going to break with tradition for no reason at all.

Anyway, this is just my guess. I do know that to answer your question definitively, you will need to virtually transport yourself to 1973...

Answer 4

I don't see any specific reason for them to be used.

I mean, maybe $ and @ could have been used to denote scalars and arrays as in perl, but I see little benefits in adding a character for each variable names.

Also, in C, arrays are really just syntactic sugar for pointers, so they can be used in a scalar context of sort.

Maybe they could have been permitted in variable names as any other valid characters.

Or maybe the reason is just they didn't think about it because there wasn't really any reason to put them.

Go ask K&R :)

Answer 5

Perhaps, the Standards committee left out these characters because there were lot of characters to choose from and they simply found these to be odd. We may never know the rationale behind why not unless someone from the Standards committee actually answers this.

Atleast $ is supported as an valid identifier in both MSVC & GCC through extensions.

Following code compiles in both:

struct $Y1$ 
{
   void $Test$() {}
};

int main() 
{
   $Y1$ $x$;
   $x$.$Test$();
   return 0;
}