===================================================================
RCS file: /cvsrepo/anoncvs/cvs/src/usr.bin/column/column.c,v
retrieving revision 1.24
retrieving revision 1.25
diff -c -r1.24 -r1.25
*** src/usr.bin/column/column.c	2016/08/31 20:43:57	1.24
--- src/usr.bin/column/column.c	2016/09/04 20:33:36	1.25
***************
*** 1,4 ****
! /*	$OpenBSD: column.c,v 1.24 2016/08/31 20:43:57 martijn Exp $	*/
  /*	$NetBSD: column.c,v 1.4 1995/09/02 05:53:03 jtc Exp $	*/
  
  /*
--- 1,4 ----
! /*	$OpenBSD: column.c,v 1.25 2016/09/04 20:33:36 martijn Exp $	*/
  /*	$NetBSD: column.c,v 1.4 1995/09/02 05:53:03 jtc Exp $	*/
  
  /*
***************
*** 36,45 ****
--- 36,48 ----
  #include <ctype.h>
  #include <err.h>
  #include <limits.h>
+ #include <locale.h>
  #include <stdio.h>
  #include <stdlib.h>
  #include <string.h>
  #include <unistd.h>
+ #include <wchar.h>
+ #include <wctype.h>
  
  void  c_columnate(void);
  void *ereallocarray(void *, size_t, size_t);
***************
*** 60,66 ****
  int eval;			/* exit value */
  int *maxwidths;			/* longest record per column */
  struct field **table;		/* one array of pointers per line */
! char *separator = "\t ";	/* field separator for table option */
  
  int
  main(int argc, char *argv[])
--- 63,69 ----
  int eval;			/* exit value */
  int *maxwidths;			/* longest record per column */
  struct field **table;		/* one array of pointers per line */
! wchar_t *separator = L"\t ";	/* field separator for table option */
  
  int
  main(int argc, char *argv[])
***************
*** 71,76 ****
--- 74,81 ----
  	char *p;
  	const char *errstr;
  
+ 	setlocale(LC_CTYPE, "");
+ 
  	termwidth = 0;
  	if ((p = getenv("COLUMNS")) != NULL)
  		termwidth = strtonum(p, 1, INT_MAX, NULL);
***************
*** 92,98 ****
  				errx(1, "%s: %s", errstr, optarg);
  			break;
  		case 's':
! 			separator = optarg;
  			break;
  		case 't':
  			tflag = 1;
--- 97,108 ----
  				errx(1, "%s: %s", errstr, optarg);
  			break;
  		case 's':
! 			if ((separator = reallocarray(NULL, strlen(optarg) + 1,
! 			    sizeof(*separator))) == NULL)
! 				err(1, NULL);
! 			if (mbstowcs(separator, optarg, strlen(optarg) + 1) ==
! 			    (size_t) -1)
! 				err(1, "sep");
  			break;
  		case 't':
  			tflag = 1;
***************
*** 106,112 ****
  	}
  
  	if (!tflag)
! 		separator = "";
  	argv += optind;
  
  	if (*argv == NULL) {
--- 116,122 ----
  	}
  
  	if (!tflag)
! 		separator = L"";
  	argv += optind;
  
  	if (*argv == NULL) {
***************
*** 222,231 ****
  	static int maxentry = 0;
  	static int maxcols = 0;
  	static struct field *cols = NULL;
! 	int col, width;
  	size_t blen;
  	ssize_t llen;
  	char *p, *s, *buf = NULL;
  
  	while ((llen = getline(&buf, &blen, fp)) > -1) {
  		if (buf[llen - 1] == '\n')
--- 232,243 ----
  	static int maxentry = 0;
  	static int maxcols = 0;
  	static struct field *cols = NULL;
! 	int col, width, twidth;
  	size_t blen;
  	ssize_t llen;
  	char *p, *s, *buf = NULL;
+ 	wchar_t wc;
+ 	int wlen;
  
  	while ((llen = getline(&buf, &blen, fp)) > -1) {
  		if (buf[llen - 1] == '\n')
***************
*** 236,251 ****
  
  			/* Skip lines containing nothing but whitespace. */
  
! 			for (s = p; s != '\0'; s++)
! 				if (!isspace((unsigned char)*s))
  					break;
  			if (*s == '\0')
  				break;
  
  			/* Skip leading, multiple, and trailing separators. */
  
! 			while (*p != '\0' && strchr(separator, *p) != NULL)
! 				p++;
  			if (*p == '\0')
  				break;
  
--- 248,265 ----
  
  			/* Skip lines containing nothing but whitespace. */
  
! 			for (s = p; (wlen = mbtowc(&wc, s, MB_CUR_MAX)) > 0;
! 			     s += wlen)
! 				if (!iswspace(wc))
  					break;
  			if (*s == '\0')
  				break;
  
  			/* Skip leading, multiple, and trailing separators. */
  
! 			while ((wlen = mbtowc(&wc, p, MB_CUR_MAX)) > 0 &&
! 			    wcschr(separator, wc) != NULL)
! 				p += wlen;
  			if (*p == '\0')
  				break;
  
***************
*** 256,266 ****
  
  			s = p;
  			width = 0;
! 			while (*p != '\0' && strchr(separator, *p) == NULL) {
! 				if (*p++ == '\t')
! 					INCR_NEXTTAB(width);
! 				else
  					width++;
  			}
  
  			if (col + 1 >= maxcols) {
--- 270,290 ----
  
  			s = p;
  			width = 0;
! 			while (*p != '\0') {
! 				if ((wlen = mbtowc(&wc, p, MB_CUR_MAX)) == -1) {
  					width++;
+ 					p++;
+ 					continue;
+ 				}
+ 				if (wcschr(separator, wc) != NULL)
+ 					break;
+ 				if (*p == '\t')
+ 					INCR_NEXTTAB(width);
+ 				else  {
+ 					width += (twidth = wcwidth(wc)) == -1 ?
+ 					    1 : twidth;
+ 				}
+ 				p += wlen;
  			}
  
  			if (col + 1 >= maxcols) {
***************
*** 284,291 ****
  			cols[col].width = width;
  			if (maxwidths[col] < width)
  				maxwidths[col] = width;
! 			if (*p != '\0')
! 				*p++ = '\0';
  			if ((cols[col].content = strdup(s)) == NULL)
  				err(1, NULL);
  		}
--- 308,317 ----
  			cols[col].width = width;
  			if (maxwidths[col] < width)
  				maxwidths[col] = width;
! 			if (*p != '\0') {
! 				*p = '\0';
! 				p += wlen;
! 			}
  			if ((cols[col].content = strdup(s)) == NULL)
  				err(1, NULL);
  		}