Printing "hello , wolrd" with out using semicolon

[Peter Nilsson]

#define P(x) printf(#x
#define L(x) #x
#define S L(| |)[!Z]
int f(int l,int i){if((i=l,!Z)&&l==Z&&P(%c%c%d\n),S,S,Z)&&f
(++i,i)){}else if(--l==Z&&P(\n))&&f(++i,i)){}else if(--l==Z

The \ in \n is not at the end of a line, not part of a character or
string literal, not part of a universal character, not part of an
include header, and does not constitute a punctuator or other
preprocessor token.

Macro processing, including # stringising macros, is done in translation
phase 4; escape sequences are translated in phase 5.

But conversion to preprocessing tokens occurs in phase 3, by which
time an isolated backslash character like the one above is _not_ a
preprocessor token.

 

[Martin Dickopp]

#define P(x) printf(#x
#define L(x) #x
#define S L(| |)[!Z]
int f(int l,int i){if((i=l,!Z)&&l==Z&&P(%c%c%d\n),S,S,Z)&&f
(++i,i)){}else if(--l==Z&&P(\n))&&f(++i,i)){}else if(--l==Z

The \ in \n is not at the end of a line, not part of a character or
string literal, not part of a universal character, not part of an
include header, and does not constitute a punctuator or other
preprocessor token.

Macro processing, including # stringising macros, is done in translation
phase 4; escape sequences are translated in phase 5.

But conversion to preprocessing tokens occurs in phase 3, by which
time an isolated backslash character like the one above is _not_ a
preprocessor token.

Why not? 6.4#3:

| [...] A preprocessing token is the minimal lexical element of the
| language in translation phases 3 through 6. The categories of
| preprocessing tokens are: header names, identifiers, preprocessing
| numbers, character constants, string literals, punctuators, and single
| non-white-space characters that do not lexically match the other
| preprocessing token categories. [...]

A single backslash character seems to fit into the last category.

Martin

 

[Peter Nilsson]

(e-mail address removed) (Peter Nilsson) wrote:

#define P(x) printf(#x
#define L(x) #x
#define S L(| |)[!Z]
int f(int l,int i){if((i=l,!Z)&&l==Z&&P(%c%c%d\n),S,S,Z)&&f
(++i,i)){}else if(--l==Z&&P(\n))&&f(++i,i)){}else if(--l==Z

The \ in \n is not at the end of a line, not part of a character or
string literal, not part of a universal character, not part of an
include header, and does not constitute a punctuator or other
preprocessor token.

Macro processing, including # stringising macros, is done in translation
phase 4; escape sequences are translated in phase 5.

But conversion to preprocessing tokens occurs in phase 3, by which
time an isolated backslash character like the one above is _not_ a
preprocessor token.

Why not? 6.4#3:

| [...] A preprocessing token is the minimal lexical element of the
| language in translation phases 3 through 6. The categories of
| preprocessing tokens are: header names, identifiers, preprocessing
| numbers, character constants, string literals, punctuators, and single
| non-white-space characters that do not lexically match the other
| preprocessing token categories. [...]

A single backslash character seems to fit into the last category.

Yup, my (very) bad.