shithub: trueawk

Info  •  Files  •  Log  •  Branches

fixed (I think) the behavior of Nan, thanks to code from Arnold Robbins; also updated awk.1 (ditto), though I think it needs more work

--- a/awk.1

+++ b/awk.1

@@ -76,6 +76,12 @@

 .I fs

 option defines the input field separator to be the regular expression

 .IR fs .

+The

+.B \-\^\-csv

+option causes

+.I awk

+to process records using (more or less) standard comma-separated values

+(CSV) format.

 .PP

 An input line is normally made up of fields separated by white space,

 or by the regular expression

@@ -576,7 +582,8 @@

 The scope rules for variables in functions are a botch;

 the syntax is worse.

 .PP

-Only eight-bit characters sets are handled correctly.

+Input is expected to be UTF-8 encoded. Other multibyte

+character sets are not handled.

 .SH UNUSUAL FLOATING-POINT VALUES

 .I Awk

 was designed before IEEE 754 arithmetic defined Not-A-Number (NaN)

--- a/run.c

+++ b/run.c

@@ -884,10 +884,15 @@

 	int i;

 	Cell *x, *y;

 	Awkfloat j;

+	bool x_is_nan, y_is_nan;

 	x = execute(a[0]);

 	y = execute(a[1]);

+	x_is_nan = isnan(x->fval);

+	y_is_nan = isnan(y->fval);

 	if (x->tval&NUM && y->tval&NUM) {

+		if ((x_is_nan || y_is_nan) && n != NE)

+			return(False);

 		j = x->fval - y->fval;

 		i = j<0? -1: (j>0? 1: 0);

 	} else {

@@ -900,7 +905,8 @@

 			else return(False);

 	case LE:	if (i<=0) return(True);

 			else return(False);

-	case NE:	if (i!=0) return(True);

+	case NE:	if (x_is_nan && y_is_nan) return(True);

+			else if (i!=0) return(True);

 			else return(False);

 	case EQ:	if (i == 0) return(True);

 			else return(False);

-- 

⑨