EmptyCrate Software. Travel. Stuff.

Quick, which language is the following code written in?

%:include <iostream>
 
int main(int, char *argv<:??))
<%
  int i = 0;
  int j = 0;
 
  for (; i < 10 and j < 11; ++i, ++j) ??< 
    std::cout << not (i bitand j) << std::endl;
  %>
??>

If you guessed standard C++, you'd be correct. C and C++ support a set of alternative tokens and character sequences to account for languages and keyboards that do not have easy access to characters that those of us with US keyboards consider to be normal.

Trigraph Sequences

The trigraph replacements are performed before any other part of the parsing process.

??= becomes #
??( becomes [
??< becomes {
??/ becomes \
??) becomes ]
??> becomes }
??’ becomes ˆ
??! becomes |
??- becomes ∼

Bigraphs and Alternative Tokens

Alternative: <% primary: {
Alternative: %> primary: }
Alternative: <: primary: [
Alternative: :> primary: ]
Alternative: %: primary: #
Alternative: %:%: primary: ##
Alternative: bitor primary: |
Alternative: or_eq primary: |=
Alternative: and primary: &&
Alternative: and_eq primary: &=
Alternative: or primary: ||
Alternative: xor_eq primary: ˆ=
Alternative: xor primary: ˆ
Alternative: not primary: !
Alternative: compl primary: ∼
Alternative: not_eq primary: !=
Alternative: bitand primary: &

Translation of Example

Our example above now becomes:

#include <iostream>
 
int main(int, char *argv[]))
{
  int i = 0;
  int j = 0;
 
  for (; i < 10 && j < 11; ++i, ++j) { 
    std::cout << ! (i & j) << std::endl;
  }
}

What does this gain you? Nothing really. Even though trigraphs are a part of the standard most compilers require a switch to enable them! However, this information might help you work on your next IOCCC entry.

Stackoverflow has more complete discussions on both trigraphs and digraphs.