core: encoding: add utf-8 functions from b_string to b_wchar interface

core/encoding.c

@@ -1191,3 +1191,191 @@ bool b_wchar_is_punct(b_wchar c)
 {
 	return iswpunct((wchar_t)c);
 }
+
+bool b_wchar_utf8_is_valid_scalar(b_wchar c)
+{
+	return (((c) >= 0x0000 && (c) <= 0xD7FF)
+	        || ((c) >= 0xE000 && (c) <= 0x10FFFF));
+}
+
+unsigned int b_wchar_utf8_header_decode(char c)
+{
+	unsigned int len = 0;
+
+	if (!(c & 0x80)) {
+		len = 1;
+	} else if (c & 0xC0 && !(c & 0x20)) {
+		len = 2;
+	} else if (c & 0xE0 && !(c & 0x10)) {
+		len = 3;
+	} else if (c & 0xF0 && !(c & 0x08)) {
+		len = 4;
+	} else {
+		len = 0;
+	}
+
+	return len;
+}
+
+unsigned int b_wchar_utf8_codepoint_size(b_wchar c)
+{
+	if (!b_wchar_utf8_is_valid_scalar(c)) {
+		return 0;
+	}
+
+	if (c <= 0x7F) {
+		return 1;
+	}
+
+	if (c <= 0x7FF) {
+		return 2;
+	}
+
+	if (c <= 0xFFFF) {
+		return 3;
+	}
+
+	if (c <= 0x10FFFF) {
+		return 4;
+	}
+
+	return 0;
+}
+
+static int32_t decode_utf8_trailer_byte(char c)
+{
+	if (!(c & 0x80) || (c & 0x40)) {
+		return -1;
+	}
+
+	return c & 0x3F;
+}
+
+b_wchar b_wchar_utf8_codepoint_decode(const char *s)
+{
+	b_wchar result = 0;
+	int len = 0;
+
+	if (!(s[0] & 0x80)) {
+		len = 1;
+		result = s[0] & 0x7F;
+	} else if (s[0] & 0xC0 && !(s[0] & 0x20)) {
+		len = 2;
+		result = s[0] & 0x1F;
+		result <<= 6;
+	} else if (s[0] & 0xE0 && !(s[0] & 0x10)) {
+		len = 3;
+		result = s[0] & 0x0F;
+		result <<= 12;
+	} else if (s[0] & 0xF0 && !(s[0] & 0x08)) {
+		len = 4;
+		result = s[0] & 0x07;
+		result <<= 18;
+	} else {
+		return B_WCHAR_INVALID;
+	}
+
+	for (int i = 1; i < len; i++) {
+		int32_t c = decode_utf8_trailer_byte(s[i]);
+		if (c == -1) {
+			return B_WCHAR_INVALID;
+		}
+
+		c <<= 6 * (len - i - 1);
+		result |= c;
+	}
+
+	if (!b_wchar_utf8_is_valid_scalar(result)) {
+		return B_WCHAR_INVALID;
+	}
+
+	return result;
+}
+
+unsigned int b_wchar_utf8_codepoint_encode(b_wchar c, char s[4])
+{
+	unsigned int len = b_wchar_utf8_codepoint_size(c);
+
+	switch (len) {
+	case 1:
+		s[0] = c & 0x7F;
+		break;
+	case 2:
+		s[0] = ((c >> 6) & 0x1F) | 0xC0;
+		s[1] = (c & 0x3F) | 0x80;
+		break;
+	case 3:
+		s[0] = ((c >> 12) & 0x0F) | 0xE0;
+		s[1] = ((c >> 6) & 0x3F) | 0x80;
+		s[2] = (c & 0x3F) | 0x80;
+		break;
+	case 4:
+		s[0] = ((c >> 18) & 0x07) | 0xF0;
+		s[1] = ((c >> 12) & 0x3F) | 0x80;
+		s[2] = ((c >> 6) & 0x3F) | 0x80;
+		s[3] = (c & 0x3F) | 0x80;
+		break;
+	default:
+		return 0;
+	}
+
+	return len;
+}
+
+unsigned int b_wchar_utf8_codepoint_stride(const char *s)
+{
+	char c = *s;
+
+	if (!(c & 0x80)) {
+		return 1;
+	}
+
+	if ((c & 0xC0) && !(c & 0x20)) {
+		return 2;
+	}
+
+	if ((c & 0xE0) && !(c & 0x10)) {
+		return 3;
+	}
+
+	if ((c & 0xF0) && !(c & 0x08)) {
+		return 4;
+	}
+
+	return 0;
+}
+
+size_t b_wchar_utf8_codepoint_count(const char *s, size_t nr_bytes)
+{
+	size_t nr_codepoints = 0;
+	const char *end = s + nr_bytes;
+
+	while (*s && s < end) {
+		size_t stride = b_wchar_utf8_codepoint_stride(s);
+		if (stride == 0) {
+			/* invalid codepoint */
+			return 0;
+		}
+
+		nr_codepoints++;
+		s += stride;
+	}
+
+	return nr_codepoints;
+}
+
+size_t b_wchar_utf8_string_encoded_size(const b_wchar *s, size_t nr_codepoints)
+{
+	size_t len = 0;
+	for (size_t i = 0; i < nr_codepoints; i++) {
+		size_t l = b_wchar_utf8_codepoint_size(s[i]);
+		if (l == 0) {
+			/* invalid codepoint */
+			return 0;
+		}
+
+		len += l;
+	}
+
+	return len;
+}