chars: implement mbtowc() ourselves, for more efficiency

This saves a function call, and the passing and checking of the
MAXCHARLEN parameter, and the checking whether wc is maybe NULL
(which for nano is never the case), and who knows what other
overheads mbtowc() has, and our workaround for glibc.

Code was written after looking at gnulib/lib/mbrtowc-impl-utf8.h.

src/chars.c

@@ -180,12 +180,49 @@ char control_mbrep(const char *c, bool isdata)
  * the number of bytes in the sequence, or -1 for an invalid sequence. */
 int mbtowide(wchar_t *wc, const char *c)
 {
-	int count = mbtowc(wc, c, MAXCHARLEN);
+#ifdef ENABLE_UTF8
+	if ((signed char)*c < 0 && use_utf8) {
+		unsigned char v1 = (unsigned char)c[0];
+		unsigned char v2 = (unsigned char)c[1] ^ 0x80;
 
-	if (count < 0 || *wc > 0x10FFFF)
+		if (v2 > 0x3F || v1 < 0xC2)
-		return -1;
+			return -1;
-	else
+
-		return count;
+		if (v1 < 0xE0) {
+			*wc = (((unsigned int)(v1 & 0x1F) << 6) | (unsigned int)v2);
+			return 2;
+		}
+
+		unsigned char v3 = (unsigned char)c[2] ^ 0x80;
+
+		if (v3 > 0x3F)
+			return -1;
+
+		if (v1 < 0xF0) {
+			if ((v1 > 0xE0 || v2 >= 0x20) && (v1 != 0xED || v2 < 0x20)) {
+				*wc = (((unsigned int)(v1 & 0x0F) << 12) |
+							((unsigned int)v2 << 6) | (unsigned int)v3);
+				return 3;
+			} else
+				return -1;
+		}
+
+		unsigned char v4 = (unsigned char)c[3] ^ 0x80;
+
+		if (v4 > 0x3F || v1 > 0xF4)
+			return -1;
+
+		if ((v1 > 0xF0 || v2 >= 0x10) && (v1 != 0xF4 || v2 < 0x10)) {
+			*wc = (((unsigned int)(v1 & 0x07) << 18) | ((unsigned int)v2 << 12) |
+							((unsigned int)v3 << 6) | (unsigned int)v4);
+			return 4;
+		} else
+			return -1;
+	}
+#endif
+
+	*wc = (unsigned int)*c;
+	return 1;
 }
 
 /* Return the width in columns of the given (multibyte) character. */