void
linked_list_free (linked_list list)
{
ASSERT_LIST (list);
linked_list_header *header = (linked_list_header *) list;
if (header->next)
{
linked_list_free ((linked_list) header->next);
}
jsp_mm_free (list);
}
void
scopes_tree_free (scopes_tree tree)
{
assert_tree (tree);
if (tree->t.children_num != 0)
{
for (uint8_t i = 0; i < tree->t.children_num; ++i)
{
scopes_tree_free (*(scopes_tree *) linked_list_element (tree->t.children, i));
}
linked_list_free (tree->t.children);
}
linked_list_free (tree->instrs);
linked_list_free (tree->var_decls);
jsp_mm_free (tree);
}
/**
* Create token of specified type from charset, transforming escape sequences.
*
* @return token descriptor
*/
static token
lexer_create_token_for_charset_transform_escape_sequences (token_type tt, /**< token type */
const lit_utf8_byte_t *charset_p, /**< charset buffer */
lit_utf8_size_t size) /**< size of the charset */
{
lit_utf8_byte_t *converted_str_p = (lit_utf8_byte_t *) jsp_mm_alloc (size);
lit_utf8_size_t converted_size = lexer_transform_escape_sequences (charset_p,
size,
converted_str_p);
token ret = lexer_create_token_for_charset (tt,
converted_str_p,
converted_size);
jsp_mm_free (converted_str_p);
return ret;
} /* lexer_create_token_for_charset_transform_escape_sequences */
/**
* Convert specified string to token of specified type, transforming escape sequences
*
* @return token descriptor
*/
static token
convert_string_to_token_transform_escape_seq (token_type tok_type, /**< type of token to produce */
const jerry_api_char_t *source_str_p, /**< string to convert,
* located in source buffer */
size_t source_str_size) /**< size of the string */
{
token ret;
if (source_str_size == 0)
{
return convert_string_to_token (tok_type,
lit_get_magic_string_utf8 (LIT_MAGIC_STRING__EMPTY),
0);
}
else
{
JERRY_ASSERT (source_str_p != NULL);
}
lit_utf8_byte_t *str_buf_p = (lit_utf8_byte_t*) jsp_mm_alloc (source_str_size);
const lit_utf8_byte_t *source_str_iter_p = source_str_p;
lit_utf8_byte_t *str_buf_iter_p = str_buf_p;
bool is_correct_sequence = true;
bool every_char_islower = true;
bool every_char_allowed_in_identifier = true;
while (source_str_iter_p < source_str_p + source_str_size)
{
ecma_char_t converted_char;
if (*source_str_iter_p != '\\')
{
converted_char = (lit_utf8_byte_t) *source_str_iter_p++;
JERRY_ASSERT (str_buf_iter_p <= str_buf_p + source_str_size);
JERRY_ASSERT (source_str_iter_p <= source_str_p + source_str_size);
}
else
{
source_str_iter_p++;
const lit_utf8_byte_t escape_character = (lit_utf8_byte_t) *source_str_iter_p++;
JERRY_ASSERT (source_str_iter_p <= source_str_p + source_str_size);
if (isdigit (escape_character))
{
if (escape_character == '0')
{
JERRY_UNIMPLEMENTED ("<NUL> character is not currently supported.\n");
}
else
{
/* Implementation-defined (ECMA-262 v5, B.1.2): octal escape sequences are not implemented */
is_correct_sequence = false;
break;
}
}
else if (escape_character == 'u'
|| escape_character == 'x')
{
const uint32_t hex_chars_num = (escape_character == 'u' ? 4u : 2u);
if (source_str_iter_p + hex_chars_num > source_str_p + source_str_size)
{
is_correct_sequence = false;
break;
}
bool chars_are_hex = true;
uint16_t char_code = 0;
for (uint32_t i = 0; i < hex_chars_num; i++)
{
const lit_utf8_byte_t byte = (lit_utf8_byte_t) *source_str_iter_p++;
if (!isxdigit (byte))
{
chars_are_hex = false;
break;
}
else
{
/*
* Check that highest 4 bits are zero, so the value would not overflow.
*/
JERRY_ASSERT ((char_code & 0xF000u) == 0);
char_code = (uint16_t) (char_code << 4u);
char_code = (uint16_t) (char_code + ecma_char_hex_to_int (byte));
}
}
JERRY_ASSERT (str_buf_iter_p <= str_buf_p + source_str_size);
JERRY_ASSERT (source_str_iter_p <= source_str_p + source_str_size);
if (!chars_are_hex)
{
is_correct_sequence = false;
break;
}
/*
* In CONFIG_ECMA_CHAR_ASCII mode size of ecma_char_t is 1 byte, so the conversion
* would ignore highest part of 2-byte value, and in CONFIG_ECMA_CHAR_UTF16 mode this
* would be just an assignment of 2-byte value.
*/
converted_char = (ecma_char_t) char_code;
}
else if (ecma_char_is_line_terminator (escape_character))
{
if (source_str_iter_p + 1 <= source_str_p + source_str_size)
{
lit_utf8_byte_t byte = *source_str_iter_p;
if (escape_character == '\x0D'
&& byte == '\x0A')
{
source_str_iter_p++;
}
}
continue;
}
else
{
convert_single_escape_character ((ecma_char_t) escape_character, &converted_char);
}
}
TODO ("Support surrogate paris.")
str_buf_iter_p += lit_code_unit_to_utf8 (converted_char, str_buf_iter_p);
JERRY_ASSERT (str_buf_iter_p <= str_buf_p + source_str_size);
if (!islower (converted_char))
{
every_char_islower = false;
if (!isalpha (converted_char)
&& !isdigit (converted_char)
&& converted_char != '$'
&& converted_char != '_')
{
every_char_allowed_in_identifier = false;
}
}
}
if (is_correct_sequence)
{
lit_utf8_size_t length = (lit_utf8_size_t) (str_buf_iter_p - str_buf_p);
ret = empty_token;
if (tok_type == TOK_NAME)
{
if (every_char_islower)
{
ret = decode_keyword (str_buf_p, length);
}
else if (!every_char_allowed_in_identifier)
{
PARSE_ERROR ("Malformed identifier name", source_str_p - buffer_start);
}
}
if (is_empty (ret))
{
ret = convert_string_to_token (tok_type, str_buf_p, length);
}
}
else
{
PARSE_ERROR ("Malformed escape sequence", source_str_p - buffer_start);
}
jsp_mm_free (str_buf_p);
return ret;
} /* convert_string_to_token_transform_escape_seq */
/**
* Create token of specified type from charset
*
* @return token descriptor
*/
static token
lexer_create_token_for_charset (jsp_token_type_t tt, /**< token type */
const lit_utf8_byte_t *charset_p, /**< charset buffer */
lit_utf8_size_t size) /**< size of the charset */
{
JERRY_ASSERT (charset_p != NULL);
lit_utf8_iterator_t iter = lit_utf8_iterator_create (charset_p, (lit_utf8_size_t) size);
lit_utf8_size_t new_size = 0;
lit_utf8_size_t new_length = 0;
bool should_convert = false;
while (!lit_utf8_iterator_is_eos (&iter))
{
if (iter.buf_pos.is_non_bmp_middle)
{
should_convert = true;
}
lit_utf8_iterator_incr (&iter);
new_size += LIT_CESU8_MAX_BYTES_IN_CODE_UNIT;
}
lit_utf8_byte_t *converted_str_p;
if (unlikely (should_convert))
{
lit_utf8_iterator_seek_bos (&iter);
converted_str_p = (lit_utf8_byte_t *) jsp_mm_alloc (new_size);
while (!lit_utf8_iterator_is_eos (&iter))
{
ecma_char_t ch = lit_utf8_iterator_read_next (&iter);
new_length += lit_code_unit_to_utf8 (ch, converted_str_p + new_length);
}
}
else
{
converted_str_p = (lit_utf8_byte_t *) charset_p;
new_length = size;
JERRY_ASSERT (lit_is_cesu8_string_valid (converted_str_p, new_length));
}
lit_literal_t lit = lit_find_literal_by_utf8_string (converted_str_p, new_length);
if (lit != NULL)
{
if (unlikely (should_convert))
{
jsp_mm_free (converted_str_p);
}
return create_token_from_lit (tt, lit);
}
lit = lit_create_literal_from_utf8_string (converted_str_p, new_length);
rcs_record_type_t type = rcs_record_get_type (lit);
JERRY_ASSERT (RCS_RECORD_TYPE_IS_CHARSET (type)
|| RCS_RECORD_TYPE_IS_MAGIC_STR (type)
|| RCS_RECORD_TYPE_IS_MAGIC_STR_EX (type));
if (unlikely (should_convert))
{
jsp_mm_free (converted_str_p);
}
return create_token_from_lit (tt, lit);
} /* lexer_create_token_for_charset */
