ObjFW  Check-in [9cfefe9c30]

		@throw [OFOutOfMemoryException
		    exceptionWithClass: [self class]
			 requestedSize: sizeof(*ivars)];

	ivars->cString = cString;
	ivars->cStringLength = cStringLength;

	switch (of_string_check_utf8(ivars->cString, ivars->cStringLength,
	    &ivars->length)) {
	case 1:
		ivars->UTF8 = YES;
		break;
	case -1:
		free(ivars);
		@throw [OFInvalidEncodingException

			table = startTable;
			tableSize = middleTableSize;
		} else {
			table = middleTable;
			tableSize = middleTableSize;
		}

		cLen = of_string_utf8_to_unicode(s->cString + i,
		    s->cStringLength - i, &c);

		if (cLen == 0 || c > 0x10FFFF) {
			[self freeMemory: unicodeString];
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];
		}

	}

	j = 0;

	for (i = 0; i < unicodeLen; i++) {
		size_t d;

		if ((d = of_string_unicode_to_utf8(unicodeString[i],
		    newCString + j)) == 0) {
			[self freeMemory: unicodeString];
			[self freeMemory: newCString];
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];
		}
		j += d;

}

- (void)setCharacter: (of_unichar_t)character
	     atIndex: (size_t)index
{
	char buffer[4];
	of_unichar_t c;
	size_t length, oldLength;

	if (s->UTF8)
		index = of_string_index_to_position(s->cString, index,
		    s->cStringLength);

	if (index > s->cStringLength)
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];


	if (!(character & 0x80) && !(s->cString[index] & 0x80)) {
		s->hashed = NO;
		s->cString[index] = character;
		return;
	}

	if ((length = of_string_unicode_to_utf8(character, buffer)) == 0)
		@throw [OFInvalidEncodingException
		    exceptionWithClass: [self class]];

	if ((oldLength = of_string_utf8_to_unicode(s->cString + index,
	    s->cStringLength - index, &c)) == 0)
		@throw [OFInvalidEncodingException
		    exceptionWithClass: [self class]];

	s->hashed = NO;

	if (length == oldLength) {
		memcpy(s->cString + index, buffer, length);
		return;
	}

	if (length > oldLength) {
		s->cString = [self resizeMemory: s->cString
					   size: s->cStringLength - oldLength +
						 length + 1];

		memmove(s->cString + index + length,
		    s->cString + index + oldLength,
		    s->cStringLength - index - oldLength);
		memcpy(s->cString + index, buffer, length);

		s->cStringLength -= oldLength;
		s->cStringLength += length;
		s->cString[s->cStringLength] = '\0';

		if (character & 0x80)
			s->UTF8 = YES;

		return;
	}

	if (length < oldLength) {
		memmove(s->cString + index + length,
		    s->cString + index + oldLength,
		    s->cStringLength - index - oldLength);
		memcpy(s->cString + index, buffer, length);

		s->cStringLength -= oldLength;
		s->cStringLength += length;
		s->cString[s->cStringLength] = '\0';

		@try {
			s->cString = [self resizeMemory: s->cString
						   size: s->cStringLength + 1];
		} @catch (OFOutOfMemoryException *e) {
			/* We don't really care, as we only made it smaller */
		}

		return;
	}

	assert(0);
}

- (void)appendUTF8String: (const char*)UTF8String
{
	size_t UTF8StringLength = strlen(UTF8String);
	size_t length;

	if (UTF8StringLength >= 3 && !memcmp(UTF8String, "\xEF\xBB\xBF", 3)) {
		UTF8String += 3;
		UTF8StringLength -= 3;
	}

	switch (of_string_check_utf8(UTF8String, UTF8StringLength, &length)) {
	case 1:
		s->UTF8 = YES;
		break;
	case -1:
		@throw [OFInvalidEncodingException
		    exceptionWithClass: [self class]];
	}

	size_t length;

	if (UTF8StringLength >= 3 && !memcmp(UTF8String, "\xEF\xBB\xBF", 3)) {
		UTF8String += 3;
		UTF8StringLength -= 3;
	}

	switch (of_string_check_utf8(UTF8String, UTF8StringLength, &length)) {
	case 1:
		s->UTF8 = YES;
		break;
	case -1:
		@throw [OFInvalidEncodingException
		    exceptionWithClass: [self class]];
	}

	size_t newCStringLength;

	if (index > s->length)
		@throw [OFOutOfRangeException
		    exceptionWithClass: [self class]];

	if (s->UTF8)
		index = of_string_index_to_position(s->cString, index,
		    s->cStringLength);

	newCStringLength = s->cStringLength + [string UTF8StringLength];
	s->hashed = NO;
	s->cString = [self resizeMemory: s->cString
				   size: newCStringLength + 1];

	if (end > s->length)
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];

	s->hashed = NO;
	s->length -= end - start;

	if (s->UTF8) {
		start = of_string_index_to_position(s->cString, start,
		    s->cStringLength);
		end = of_string_index_to_position(s->cString, end,
		    s->cStringLength);
	}

	memmove(s->cString + start, s->cString + end, s->cStringLength - end);
	s->cStringLength -= end - start;
	s->cString[s->cStringLength] = 0;

	if (end > s->length)
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];

	newLength = s->length - (end - start) + [replacement length];

	if (s->UTF8) {
		start = of_string_index_to_position(s->cString, start,
		    s->cStringLength);
		end = of_string_index_to_position(s->cString, end,
		    s->cStringLength);
	}

	newCStringLength = s->cStringLength - (end - start) +
	    [replacement UTF8StringLength];
	s->hashed = NO;
	s->cString = [self resizeMemory: s->cString

	const char *replacementString = [replacement UTF8String];
	size_t searchLength = [string UTF8StringLength];
	size_t replacementLength = [replacement UTF8StringLength];
	size_t i, last, newCStringLength, newLength;
	char *newCString;

	if (s->UTF8) {
		range.start = of_string_index_to_position(s->cString,
		    range.start, s->cStringLength);
		range.length = of_string_index_to_position(s->cString,
		    range.start + range.length, s->cStringLength) - range.start;
	}

	if (range.start + range.length > [self UTF8StringLength])
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];

	if ([string UTF8StringLength] > range.length)

				if ((c1 & 0xFC00) == 0xDC00) {
					free(buffer);
					return nil;
				}

				/* Normal character */
				if ((c1 & 0xFC00) != 0xD800) {
					l = of_string_unicode_to_utf8(c1,
					    buffer + i);
					if (l == 0) {
						free(buffer);
						return nil;
					}

					i += l;

					free(buffer);
					return nil;
				}

				c = (((c1 & 0x3FF) << 10) |
				    (c2 & 0x3FF)) + 0x10000;

				l = of_string_unicode_to_utf8(c, buffer + i);
				if (l == 0) {
					free(buffer);
					return nil;
				}

				i += l;
				*pointer += 11;

			if ((c1 & 0xFC00) == 0xDC00) {
				free(buffer);
				return nil;
			}

			/* Normal character */
			if ((c1 & 0xFC00) != 0xD800) {
				l = of_string_unicode_to_utf8(c1, buffer + i);
				if (l == 0) {
					free(buffer);
					return nil;
				}

				i += l;
				*pointer += 5;

			if (c2 == 0xFFFF) {
				free(buffer);
				return nil;
			}

			c = (((c1 & 0x3FF) << 10) | (c2 & 0x3FF)) + 0x10000;

			l = of_string_unicode_to_utf8(c, buffer + i);
			if (l == 0) {
				free(buffer);
				return nil;
			}

			i += l;
			*pointer += 11;

			if (entity[i] >= '0' && entity[i] <= '9')
				c = (c * 10) + (entity[i] - '0');
			else
				return nil;
		}
	}

	if ((i = of_string_unicode_to_utf8(c, buffer)) == 0)
		return nil;
	buffer[i] = 0;

	return [OFString stringWithUTF8String: buffer
				       length: i];
}

#ifdef OF_HAVE_BLOCKS
typedef void (^of_string_line_enumeration_block_t)(OFString *line, BOOL *stop);
#endif

#ifdef __cplusplus
extern "C" {
#endif
extern int of_string_check_utf8(const char*, size_t, size_t*);
extern size_t of_string_unicode_to_utf8(of_unichar_t, char*);
extern size_t of_string_utf8_to_unicode(const char*, size_t, of_unichar_t*);
extern size_t of_string_position_to_index(const char*, size_t);
extern size_t of_string_index_to_position(const char*, size_t, size_t);
extern size_t of_unicode_string_length(const of_unichar_t*);
extern size_t of_utf16_string_length(const uint16_t*);
#ifdef __cplusplus
}
#endif

@class OFArray;

}

void _reference_to_OFConstantString(void)
{
	[OFConstantString class];
}

int
of_string_check_utf8(const char *cString, size_t cStringLength, size_t *length)
{
	size_t i, tmpLength = cStringLength;
	int UTF8 = 0;

	for (i = 0; i < cStringLength; i++) {
		/* No sign of UTF-8 here */
		if OF_LIKELY (!(cString[i] & 0x80))
			continue;

		UTF8 = 1;

		/* We're missing a start byte here */
		if OF_UNLIKELY (!(cString[i] & 0x40))
			return -1;

		/* 2 byte sequences for code points 0 - 127 are forbidden */
		if OF_UNLIKELY ((cString[i] & 0x7E) == 0x40)
			return -1;

		/* We have at minimum a 2 byte character -> check next byte */
		if OF_UNLIKELY (cStringLength <= i + 1 ||
		    (cString[i + 1] & 0xC0) != 0x80)
			return -1;

		/* Check if we have at minimum a 3 byte character */
		if OF_LIKELY (!(cString[i] & 0x20)) {
			i++;
			tmpLength--;
			continue;
		}

		/* We have at minimum a 3 byte char -> check second next byte */
		if OF_UNLIKELY (cStringLength <= i + 2 ||
		    (cString[i + 2] & 0xC0) != 0x80)
			return -1;

		/* Check if we have a 4 byte character */
		if OF_LIKELY (!(cString[i] & 0x10)) {
			i += 2;
			tmpLength -= 2;
			continue;
		}

		/* We have a 4 byte character -> check third next byte */
		if OF_UNLIKELY (cStringLength <= i + 3 ||
		    (cString[i + 3] & 0xC0) != 0x80)
			return -1;

		/*
		 * Just in case, check if there's a 5th character, which is
		 * forbidden by UTF-8
		 */
		if OF_UNLIKELY (cString[i] & 0x08)
			return -1;

		i += 3;
		tmpLength -= 3;
	}

	if (length != NULL)
		*length = tmpLength;

	return UTF8;
}

size_t
of_string_unicode_to_utf8(of_unichar_t character, char *buffer)
{
	size_t i = 0;

	if (character < 0x80) {
		buffer[i] = character;
		return 1;
	}
	if (character < 0x800) {
		buffer[i++] = 0xC0 | (character >> 6);
		buffer[i] = 0x80 | (character & 0x3F);
		return 2;
	}
	if (character < 0x10000) {
		buffer[i++] = 0xE0 | (character >> 12);
		buffer[i++] = 0x80 | (character >> 6 & 0x3F);
		buffer[i] = 0x80 | (character & 0x3F);
		return 3;
	}
	if (character < 0x110000) {
		buffer[i++] = 0xF0 | (character >> 18);
		buffer[i++] = 0x80 | (character >> 12 & 0x3F);
		buffer[i++] = 0x80 | (character >> 6 & 0x3F);
		buffer[i] = 0x80 | (character & 0x3F);
		return 4;
	}

	return 0;
}

size_t
of_string_utf8_to_unicode(const char *buffer_, size_t length, of_unichar_t *ret)
{
	const uint8_t *buffer = (const uint8_t*)buffer_;

	if (!(*buffer & 0x80)) {
		*ret = buffer[0];
		return 1;
	}

	if ((*buffer & 0xE0) == 0xC0) {
		if OF_UNLIKELY (length < 2)
			return 0;

		*ret = ((buffer[0] & 0x1F) << 6) | (buffer[1] & 0x3F);
		return 2;
	}

	if ((*buffer & 0xF0) == 0xE0) {
		if OF_UNLIKELY (length < 3)
			return 0;

		*ret = ((buffer[0] & 0x0F) << 12) | ((buffer[1] & 0x3F) << 6) |
		    (buffer[2] & 0x3F);
		return 3;
	}

	if ((*buffer & 0xF8) == 0xF0) {
		if OF_UNLIKELY (length < 4)
			return 0;

		*ret = ((buffer[0] & 0x07) << 18) | ((buffer[1] & 0x3F) << 12) |
		    ((buffer[2] & 0x3F) << 6) | (buffer[3] & 0x3F);
		return 4;
	}

	return 0;
}

size_t
of_string_position_to_index(const char *string, size_t position)
{
	size_t i, index = position;

	for (i = 0; i < position; i++)
		if OF_UNLIKELY ((string[i] & 0xC0) == 0x80)
			index--;

	return index;
}

size_t
of_string_index_to_position(const char *string, size_t index, size_t length)
{
	size_t i;

	for (i = 0; i <= index; i++)
		if OF_UNLIKELY ((string[i] & 0xC0) == 0x80)
			if (++index > length)
				return OF_INVALID_INDEX;

	return index;
}

size_t
of_unicode_string_length(const of_unichar_t *string)
{
	const of_unichar_t *string_ = string;

	while (*string_ != 0)
		string_++;

	OFObject *object;

	object = [[[OFObject alloc] init] autorelease];
	UTF8String = [object allocMemoryWithSize: (length * 4) + 1];

	for (i = 0; i < length; i++) {
		char buffer[4];
		size_t characterLen = of_string_unicode_to_utf8(
		    unicodeString[i], buffer);

		switch (characterLen) {
		case 1:
			UTF8String[j++] = buffer[0];
			break;
		case 2:
			UTF8StringLength++;

			memcpy(UTF8String + j, buffer, 2);

{
	const of_unichar_t *unicodeString = [self unicodeString];
	size_t length = [self length];
	size_t i, UTF8StringLength = 0;

	for (i = 0; i < length; i++) {
		char buffer[4];
		size_t characterLen = of_string_unicode_to_utf8(
		    unicodeString[i], buffer);

		if (characterLen == 0)
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		UTF8StringLength += characterLen;
	}

	return UTF8StringLength;
}

- (size_t)cStringLengthWithEncoding: (of_string_encoding_t)encoding
{

			return OF_ORDERED_DESCENDING;
		if (tolower((int)first[i]) < tolower((int)second[i]))
			return OF_ORDERED_ASCENDING;
	}

	return OF_ORDERED_SAME;
}
































































































































































@implementation OFString_UTF8
- init
{
	self = [super init];

	@try {

		}

		s = &s_store;

		s->cString = storage;
		s->cStringLength = UTF8StringLength;

		switch (of_string_check_utf8(UTF8String, UTF8StringLength,
		    &s->length)) {
		case 1:
			s->UTF8 = YES;
			break;
		case -1:
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		s = &s_store;

		s->cString = [self allocMemoryWithSize: cStringLength + 1];
		s->cStringLength = cStringLength;

		if (encoding == OF_STRING_ENCODING_UTF_8 ||
		    encoding == OF_STRING_ENCODING_ASCII) {
			switch (of_string_check_utf8(cString, cStringLength,
			    &s->length)) {
			case 1:
				if (encoding == OF_STRING_ENCODING_ASCII)
					@throw [OFInvalidEncodingException
					    exceptionWithClass: [self class]];

				s->UTF8 = YES;

				if (!(cString[i] & 0x80)) {
					s->cString[j++] = cString[i];
					continue;
				}

				s->UTF8 = YES;
				bytes = of_string_unicode_to_utf8(
				    (uint8_t)cString[i], buffer);

				if (bytes == 0)
					@throw [OFInvalidEncodingException
					    exceptionWithClass: [self class]];

				s->cStringLength += bytes - 1;

			character = table[(uint8_t)cString[i]];

			if (character == 0xFFFD)
				@throw [OFInvalidEncodingException
				    exceptionWithClass: [self class]];

			s->UTF8 = YES;
			characterBytes = of_string_unicode_to_utf8(character,
			    buffer);

			if (characterBytes == 0)
				@throw [OFInvalidEncodingException
				    exceptionWithClass: [self class]];

			s->cStringLength += characterBytes - 1;

		}

		s = &s_store;

		s->cString = (char*)UTF8String;
		s->cStringLength = UTF8StringLength;

		switch (of_string_check_utf8(UTF8String, UTF8StringLength,
		    &s->length)) {
		case 1:
			s->UTF8 = YES;
			break;
		case -1:
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		s->cStringLength = length;
		s->cString = [self allocMemoryWithSize: (length * 4) + 1];
		s->length = length;

		for (i = 0; i < length; i++) {
			char buffer[4];
			size_t characterLen = of_string_unicode_to_utf8(
			    (swap ? of_bswap32(string[i]) : string[i]),
			    buffer);

			switch (characterLen) {
			case 1:
				s->cString[j++] = buffer[0];
				break;

				    (nextCharacter & 0x3FF)) + 0x10000;

				i++;
				s->cStringLength--;
				s->length--;
			}

			characterLen = of_string_unicode_to_utf8(
			    character, buffer);

			switch (characterLen) {
			case 1:
				s->cString[j++] = buffer[0];
				break;
			case 2:
				s->UTF8 = YES;

		    arguments)) == -1)
			@throw [OFInvalidFormatException
			    exceptionWithClass: [self class]];

		s->cStringLength = cStringLength;

		@try {
			switch (of_string_check_utf8(tmp, cStringLength,
			    &s->length)) {
			case 1:
				s->UTF8 = YES;
				break;
			case -1:
				@throw [OFInvalidEncodingException
				    exceptionWithClass: [self class]];

	i = j = 0;

	while (i < s->cStringLength && j < otherCStringLength) {
		of_unichar_t c1, c2;
		size_t l1, l2;

		l1 = of_string_utf8_to_unicode(s->cString + i,
		    s->cStringLength - i, &c1);
		l2 = of_string_utf8_to_unicode(otherCString + j,
		    otherCStringLength - j, &c2);

		if (l1 == 0 || l2 == 0 || c1 > 0x10FFFF || c2 > 0x10FFFF)
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		if (c1 >> 8 < OF_UNICODE_CASEFOLDING_TABLE_SIZE) {

	OF_HASH_INIT(hash);

	for (i = 0; i < s->cStringLength; i++) {
		of_unichar_t c;
		size_t length;

		if ((length = of_string_utf8_to_unicode(s->cString + i,
		    s->cStringLength - i, &c)) == 0)
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		OF_HASH_ADD(hash, (c & 0xFF0000) >> 16);
		OF_HASH_ADD(hash, (c & 0x00FF00) >>  8);
		OF_HASH_ADD(hash,  c & 0x0000FF);

	if (index >= s->length)
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];

	if (!s->UTF8)
		return s->cString[index];

	index = of_string_index_to_position(s->cString, index,
	    s->cStringLength);

	if (!of_string_utf8_to_unicode(s->cString + index,
	    s->cStringLength - index, &character))
		@throw [OFInvalidEncodingException
		    exceptionWithClass: [self class]];

	return character;
}

- (void)getCharacters: (of_unichar_t*)buffer

	if (cStringLength > s->cStringLength)
		return of_range(OF_INVALID_INDEX, 0);

	if (options & OF_STRING_SEARCH_BACKWARDS) {
		for (i = s->cStringLength - cStringLength;; i--) {
			if (!memcmp(s->cString + i, cString, cStringLength))
				return of_range(
				    of_string_position_to_index(s->cString, i),
				    [string length]);

			/* Did not match and we're at the last char */
			if (i == 0)
				return of_range(OF_INVALID_INDEX, 0);
		}
	} else {
		for (i = 0; i <= s->cStringLength - cStringLength; i++)
			if (!memcmp(s->cString + i, cString, cStringLength))
				return of_range(
				    of_string_position_to_index(s->cString, i),
				    [string length]);
	}

	return of_range(OF_INVALID_INDEX, 0);
}

- (BOOL)containsString: (OFString*)string

	size_t start = range.start;
	size_t end = range.start + range.length;

	if (end > s->length)
		@throw [OFOutOfRangeException exceptionWithClass: [self class]];

	if (s->UTF8) {
		start = of_string_index_to_position(s->cString, start,
		    s->cStringLength);
		end = of_string_index_to_position(s->cString, end,
		    s->cStringLength);
	}

	return [OFString stringWithUTF8String: s->cString + start
				       length: end - start];
}

	i = 0;
	j = 0;

	while (i < s->cStringLength) {
		of_unichar_t c;
		size_t cLen;

		cLen = of_string_utf8_to_unicode(s->cString + i,
		    s->cStringLength - i, &c);

		if (cLen == 0 || c > 0x10FFFF)
			@throw [OFInvalidEncodingException
			    exceptionWithClass: [self class]];

		ret[j++] = c;