Add the -b option (break at byte position, not column number) and the -s

option (try to break at word bounaries) for SUSv3 conformance.

Partially based on the NetBSD version, with the following changes:
 - style(9)
 - break on <blank>s, not spaces, per POSIX (and GNU)
 - when looking for last space on line, search backwards instead of
   forwards; less comparisons needed this way.
 - use LINE_MAX macro instead of a magic number and a comment saying it is
   LINE_MAX.

PR:		36245
Reviewed by:	mike
Obtained from:	NetBSD (partially)

2 files changed, 97 insertions, 29 deletions

diff --git a/usr.bin/fold/fold.1 b/usr.bin/fold/fold.1
index 63a52f2..071f39a 100644
--- a/usr.bin/fold/fold.1
+++ b/usr.bin/fold/fold.1
@@ -32,7 +32,7 @@
 .\"	@(#)fold.1	8.1 (Berkeley) 6/6/93
 .\" $FreeBSD$
 .\"
-.Dd June 6, 1993
+.Dd April 17, 2002
 .Dt FOLD 1
 .Os
 .Sh NAME
@@ -40,6 +40,7 @@
 .Nd "fold long lines for finite width output device"
 .Sh SYNOPSIS
 .Nm
+.Op Fl bs
 .Op Fl w Ar width
 .Op Ar
 .Sh DESCRIPTION
@@ -50,6 +51,14 @@ breaking the lines to have a maximum of 80 characters.
 .Pp
 The following option is available:
 .Bl -tag -width indent
+.It Fl b
+Count
+.Ar width
+in bytes rather than column positions.
+.It Fl s
+Fold line after the last blank character within the first
+.Ar width
+column positions (or bytes).
 .It Fl w Ar width
 Specify a line width to use instead of the default 80 characters.
 .Ar Width
@@ -61,5 +70,10 @@ before using
 .El
 .Sh SEE ALSO
 .Xr expand 1
+.Sh STANDARDS
+The
+.Nm
+utility conforms to
+.St -p1003.1-2001 .
 .Sh BUGS
 If underlining is present it may be messed up by folding.
diff --git a/usr.bin/fold/fold.c b/usr.bin/fold/fold.c
index 33208ae..68a3ebe 100644
--- a/usr.bin/fold/fold.c
+++ b/usr.bin/fold/fold.c
@@ -49,16 +49,23 @@ static char sccsid[] = "@(#)fold.c	8.1 (Berkeley) 6/6/93";
 #include <sys/cdefs.h>
 __FBSDID("$FreeBSD$");
 
+#include <ctype.h>
 #include <err.h>
+#include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
+#include <string.h>
 #include <unistd.h>
 
 #define	DEFLINEWIDTH	80
 
 void fold(int);
+static int newpos(int, int);
 static void usage(void);
 
+int bflag;			/* Count bytes, not columns */
+int sflag;			/* Split on word boundaries */
+
 int
 main(argc, argv)
 	int argc;
@@ -69,8 +76,14 @@ main(argc, argv)
 	char *p;
 
 	width = -1;
-	while ((ch = getopt(argc, argv, "0123456789w:")) != -1)
+	while ((ch = getopt(argc, argv, "0123456789bsw:")) != -1)
 		switch (ch) {
+		case 'b':
+			bflag = 1;
+			break;
+		case 's':
+			sflag = 1;
+			break;
 		case 'w':
 			if ((width = atoi(optarg)) <= 0) {
 				errx(1, "illegal width value");
@@ -107,57 +120,98 @@ main(argc, argv)
 static void
 usage()
 {
-	(void)fprintf(stderr, "usage: fold [-w width] [file ...]\n");
+	(void)fprintf(stderr, "usage: fold [-bs] [-w width] [file ...]\n");
 	exit(1);
 }
 
+/*
+ * Fold the contents of standard input to fit within WIDTH columns (or bytes)
+ * and write to standard output.
+ *
+ * If sflag is set, split the line at the last space character on the line.
+ * This flag necessitates storing the line in a buffer until the current
+ * column > width, or a newline or EOF is read.
+ *
+ * The buffer can grow larger than WIDTH due to backspaces and carriage
+ * returns embedded in the input stream.
+ */
 void
 fold(width)
 	register int width;
 {
-	register int ch, col, new;
+	static char *buf;
+	static int buf_max;
+	int ch, col, i, indx, space;
 
-	for (col = 0;;) {
-		switch (ch = getchar()) {
-		case EOF:
-			return;
-		case '\b':
-			new = col ? col - 1 : 0;
-			break;
-		case '\n':
-		case '\r':
-			new = 0;
-			break;
-		case '\t':
-			new = (col + 8) & ~7;
-			break;
-		default:
-			new = col + 1;
-			break;
+	col = indx = 0;
+	while ((ch = getchar()) != EOF) {
+		if (ch == '\n') {
+			if (indx != 0)
+				fwrite(buf, 1, indx, stdout);
+			putchar('\n');
+			col = indx = 0;
+			continue;
 		}
-
-		if (new > width) {
+		if ((col = newpos(col, ch)) > width) {
+			if (sflag) {
+				i = indx;
+				while (--i >= 0 && !isblank(buf[i]))
+					;
+				space = i;
+			}
+			if (sflag && space != -1) {
+				space++;
+				fwrite(buf, 1, space, stdout);
+				memmove(buf, buf + space, indx - space);
+				indx -= space;
+				col = 0;
+				for (i = 0; i < indx; i++)
+					col = newpos(col, buf[i]);
+			} else {
+				fwrite(buf, 1, indx, stdout);
+				col = indx = 0;
+			}
 			putchar('\n');
-			col = 0;
+			col = newpos(col, ch);
+		}
+		if (indx + 1 > buf_max) {
+			buf_max += LINE_MAX;
+			if ((buf = realloc(buf, buf_max)) == NULL)
+				err(1, "realloc()");
 		}
-		putchar(ch);
+		buf[indx++] = ch;
+	}
 
+	if (indx != 0)
+		fwrite(buf, 1, indx, stdout);
+}
+
+/*
+ * Update the current column position for a character.
+ */
+static int
+newpos(col, ch)
+	int col, ch;
+{
+
+	if (bflag)
+		++col;
+	else
 		switch (ch) {
 		case '\b':
 			if (col > 0)
 				--col;
 			break;
-		case '\n':
 		case '\r':
 			col = 0;
 			break;
 		case '\t':
-			col += 8;
-			col &= ~7;
+			col = (col + 8) & ~7;
 			break;
 		default:
 			++col;
 			break;
 		}
-	}
+
+	return (col);
 }