Clean up all major UTF-8 issues in cvt.c.

This also allows to delete the buggy, now unused function put_wchar().
OK millert@

3 files changed, 11 insertions, 46 deletions

diff --git a/usr.bin/less/charset.c b/usr.bin/less/charset.c
index 363d1a80cba..ec1365414ed 100644
--- a/usr.bin/less/charset.c
+++ b/usr.bin/less/charset.c
@@ -329,48 +329,6 @@ get_wchar(const char *p)
 }
 
 /*
- * Store a character into a UTF-8 string.
- */
-void
-put_wchar(char **pp, LWCHAR ch)
-{
-	if (!utf_mode || ch < 0x80) {
-		/* 0xxxxxxx */
-		*(*pp)++ = (char)ch;
-	} else if (ch < 0x800) {
-		/* 110xxxxx 10xxxxxx */
-		*(*pp)++ = (char)(0xC0 | ((ch >> 6) & 0x1F));
-		*(*pp)++ = (char)(0x80 | (ch & 0x3F));
-	} else if (ch < 0x10000) {
-		/* 1110xxxx 10xxxxxx 10xxxxxx */
-		*(*pp)++ = (char)(0xE0 | ((ch >> 12) & 0x0F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 6) & 0x3F));
-		*(*pp)++ = (char)(0x80 | (ch & 0x3F));
-	} else if (ch < 0x200000) {
-		/* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
-		*(*pp)++ = (char)(0xF0 | ((ch >> 18) & 0x07));
-		*(*pp)++ = (char)(0x80 | ((ch >> 12) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 6) & 0x3F));
-		*(*pp)++ = (char)(0x80 | (ch & 0x3F));
-	} else if (ch < 0x4000000) {
-		/* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
-		*(*pp)++ = (char)(0xF0 | ((ch >> 24) & 0x03));
-		*(*pp)++ = (char)(0x80 | ((ch >> 18) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 12) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 6) & 0x3F));
-		*(*pp)++ = (char)(0x80 | (ch & 0x3F));
-	} else {
-		/* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
-		*(*pp)++ = (char)(0xF0 | ((ch >> 30) & 0x01));
-		*(*pp)++ = (char)(0x80 | ((ch >> 24) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 18) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 12) & 0x3F));
-		*(*pp)++ = (char)(0x80 | ((ch >> 6) & 0x3F));
-		*(*pp)++ = (char)(0x80 | (ch & 0x3F));
-	}
-}
-
-/*
  * Step forward or backward one character in a string.
  */
 LWCHAR
diff --git a/usr.bin/less/cvt.c b/usr.bin/less/cvt.c
index ed7c72d69a1..70a58851e36 100644
--- a/usr.bin/less/cvt.c
+++ b/usr.bin/less/cvt.c
@@ -60,7 +60,8 @@ cvt_text(char *odst, char *osrc, int *chpos, int *lenp, int ops)
 	char *edst = odst;
 	char *src;
 	char *src_end;
-	LWCHAR ch;
+	wchar_t ch;
+	int len;
 
 	if (lenp != NULL)
 		src_end = osrc + *lenp;
@@ -70,8 +71,10 @@ cvt_text(char *odst, char *osrc, int *chpos, int *lenp, int ops)
 	for (src = osrc, dst = odst; src < src_end; ) {
 		int src_pos = src - osrc;
 		int dst_pos = dst - odst;
-		ch = step_char(&src, +1, src_end);
+		if ((len = mbtowc(&ch, src, src_end - src)) < 1)
+			ch = L'\0';
 		if ((ops & CVT_BS) && ch == '\b' && dst > odst) {
+			src++;
 			/* Delete backspace and preceding char. */
 			do {
 				dst--;
@@ -83,11 +86,16 @@ cvt_text(char *odst, char *osrc, int *chpos, int *lenp, int ops)
 			while (src < src_end)
 				if (!is_ansi_middle(*src++))
 					break;
+		} else if (len < 1) {
+			*dst++ = *src++;
+			if (chpos != NULL)
+				chpos[dst_pos] = src_pos;
 		} else {
+			src += len;
 			/* Just copy the char to the destination buffer. */
 			if ((ops & CVT_TO_LC) && iswupper(ch))
 				ch = towlower(ch);
-			put_wchar(&dst, ch);
+			dst += wctomb(dst, ch);
 			/* Record the original position of the char. */
 			if (chpos != NULL)
 				chpos[dst_pos] = src_pos;
diff --git a/usr.bin/less/funcs.h b/usr.bin/less/funcs.h
index 5c24ba89e96..87439f06e2e 100644
--- a/usr.bin/less/funcs.h
+++ b/usr.bin/less/funcs.h
@@ -62,7 +62,6 @@ char *prutfchar(LWCHAR);
 int utf_len(char);
 int is_utf8_well_formed(const char *);
 LWCHAR get_wchar(const char *);
-void put_wchar(char **, LWCHAR);
 LWCHAR step_char(char **, int, char *);
 int is_composing_char(LWCHAR);
 int is_ubin_char(LWCHAR);