/**
* Return parse info for parse type.
*
* @return ?map Parse info (NULL: error)
*/
function _js_parse_js()
{
// Choice{"FUNCTION" "IDENTIFIER "BRACKET_OPEN" comma_parameters "BRACKET_CLOSE" command | command}*
$next = parser_peek();
$program = array();
$program['functions'] = array();
$program['main'] = array();
while (!is_null($next)) {
switch ($next) {
case 'FUNCTION':
$_function = _js_parse_function_dec();
if (is_null($_function)) {
return NULL;
}
foreach ($program['functions'] as $_) {
if ($_['name'] == $_function['name']) {
js_log_warning('PARSER', 'Duplicated function \'' . $_function['name'] . '\'');
}
}
//log_special('defined',$_function['name']);
$program['functions'][] = $_function;
// Sometimes happens when people get confused by =function() and function blah() {};
$next_2 = parser_peek();
if ($next_2 == 'COMMAND_TERMINATE') {
parser_next();
}
break;
default:
$command = _js_parse_command();
if (is_null($command)) {
return NULL;
}
$program['main'] = array_merge($program['main'], $command);
break;
}
$next = parser_peek();
}
return $program;
}
/**
* Do type checking for something specific.
*
* @param list List of allowed types
* @param string Actual type involved
* @param integer Current parse position
* @param ?string Specific error message to give (NULL: use default)
* @return boolean Whether it type-checks
*/
function js_ensure_type($_allowed_types, $actual_type, $pos, $alt_error = NULL)
{
if ($actual_type == '!Object') {
return true;
}
// We can't check it
global $JS_PROTOTYPES;
// Tidy up our allow list to be a nice map
$allowed_types = array('Undefined' => 1, 'Null' => 1);
foreach ($_allowed_types as $type) {
if ($type == '') {
continue;
}
// Weird
if ($type[0] == '!') {
$allowed_types += $JS_PROTOTYPES[substr($type, 1)][2];
$allowed_types[substr($type, 1)] = 1;
} else {
$allowed_types[$type] = 1;
}
}
// The check
if (substr($actual_type, 0, 1) == '!') {
$actual_type = substr($actual_type, 1);
}
if (isset($allowed_types[$actual_type])) {
return true;
}
js_log_warning('CHECKER', is_null($alt_error) ? 'Type mismatch' : $alt_error, $pos);
return false;
}
/**
* Lex some Javascript code.
*
* @param string The code
* @return list List of lexed tokens
*/
function js_lex($text)
{
global $CONTINUATIONS, $TOKENS, $JS_TAG_RANGES, $JS_VALUE_RANGES, $JS_TEXT, $JS_LEX_TOKENS;
// So that we don't have to consider end-of-file states as much.
$JS_TEXT = $text . "\n";
$JS_LEX_TOKENS = array();
// We will be lexing into this list of tokens
$special_token_value = '';
// This will be used during special lexing modes to build up the special token value being lexed
$lex_state = LEXER_FREE;
$escape_flag = false;
// Used for string_literal escaping
// Lex the code. Hard coded state changes occur. Understanding of tokenisation implicit. Trying to match tokens to $TOKENS, otherwise an identifier.
$char = '';
$i = 0;
while (true) {
switch ($lex_state) {
case LEXER_FREE:
// Jump over any white space in our way
do {
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 3;
}
} while (trim($char) == '');
// We need to know where our token is starting
$i--;
$i_current = $i;
// Try and work out what token we're looking at next
$maybe_applicable_tokens = $TOKENS;
$applicable_tokens = array();
$token_so_far = '';
while (count($maybe_applicable_tokens) != 0) {
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 3;
}
$token_so_far .= $char;
$_ = $token_so_far[0];
// To strict stupid optimiser
// Filter out any tokens that no longer match
$cnt = count($JS_LEX_TOKENS);
foreach ($maybe_applicable_tokens as $token_name => $token_value) {
// Hasn't matched (or otherwise, may still match)
if (substr($token_value, 0, strlen($token_so_far)) !== $token_so_far) {
unset($maybe_applicable_tokens[$token_name]);
} else {
// Is it a perfect match?
if (strlen($token_so_far) == strlen($token_value) && (!array_key_exists($token_so_far[0], $CONTINUATIONS) || !array_key_exists($JS_TEXT[$i], $CONTINUATIONS))) {
if ($token_name != 'FUNCTION' || !isset($JS_LEX_TOKENS[$cnt - 1]) || $JS_LEX_TOKENS[$cnt - 1][0] != 'NEW') {
$applicable_tokens[] = $token_name;
}
unset($maybe_applicable_tokens[$token_name]);
}
}
}
}
if (in_array('DIV_EQUAL', $applicable_tokens)) {
$previous = isset($JS_LEX_TOKENS[count($JS_LEX_TOKENS) - 1]) ? $JS_LEX_TOKENS[count($JS_LEX_TOKENS) - 1][0] : 'BRACKET_OPEN';
if ($previous == 'BRACKET_OPEN' || $previous == 'COMMA') {
$applicable_tokens = array('DIVIDE');
// Actually, a regular expression
}
}
// If we have any applicable tokens, find the longest and move $i so it's as we just read it
$i = $i_current;
if (count($applicable_tokens) != 0) {
usort($applicable_tokens, '_strlen_sort');
$token_found = $applicable_tokens[count($applicable_tokens) - 1];
$i += strlen($TOKENS[$token_found]);
// Is it a special state jumping token?
if ($token_found == 'START_ML_COMMENT') {
$lex_state = LEXER_ML_COMMENT;
break;
} elseif ($token_found == 'COMMENT') {
$lex_state = LEXER_COMMENT;
break;
} elseif ($token_found == 'DIVIDE' && !in_array(@$JS_LEX_TOKENS[count($JS_LEX_TOKENS) - 1][0], array('number_literal', 'IDENTIFIER', 'EXTRACT_CLOSE', 'BRACKET_CLOSE'))) {
$lex_state = LEXER_REGEXP;
break;
} elseif ($token_found == 'DOUBLE_QUOTE') {
$lex_state = LEXER_DOUBLE_QUOTE_STRING_LITERAL;
break;
} elseif ($token_found == 'SINGLE_QUOTE') {
$lex_state = LEXER_SINGLE_QUOTE_STRING_LITERAL;
break;
}
$JS_LEX_TOKENS[] = array($token_found, $i);
} else {
// Otherwise, we've found an identifier or numerical literal token, so extract it
$token_found = '';
$numeric = NULL;
do {
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 3;
}
if (is_null($numeric)) {
$numeric = array_key_exists($char, array('0' => 1, '1' => 1, '2' => 1, '3' => 1, '4' => 1, '5' => 1, '6' => 1, '7' => 1, '8' => 1, '9' => 1));
}
if (!array_key_exists($char, $CONTINUATIONS) && ($numeric === false || $char != '.' || !is_numeric($JS_TEXT[$i]))) {
break;
}
$token_found .= $char;
} while (true);
$i--;
if ($numeric) {
if (strpos($token_found, '.') !== false) {
$JS_LEX_TOKENS[] = array('number_literal', floatval($token_found), $i);
} elseif (strpos($token_found, 'x') !== false) {
$JS_LEX_TOKENS[] = array('number_literal', intval(base_convert($token_found, 16, 10)), $i);
} elseif ($token_found[0] == '0') {
$JS_LEX_TOKENS[] = array('number_literal', intval(base_convert($token_found, 8, 10)), $i);
} else {
$JS_LEX_TOKENS[] = array('number_literal', intval($token_found), $i);
}
$JS_VALUE_RANGES[] = array($i - strlen($token_found), $i);
} else {
if ($token_found == '') {
js_log_warning('LEXER', 'Bad token found', $i, true);
return array();
}
$JS_LEX_TOKENS[] = array('IDENTIFIER', $token_found, $i);
$JS_TAG_RANGES[] = array($i - strlen($token_found), $i);
}
}
break;
case LEXER_COMMENT:
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 2;
}
// Exit case
if ($char == chr(10)) {
$lex_state = LEXER_FREE;
$JS_LEX_TOKENS[] = array('comment', $special_token_value, $i);
$special_token_value = '';
$i--;
break;
}
// Normal case
$special_token_value .= $char;
break;
case LEXER_ML_COMMENT:
list($reached_end, $i, $char) = lex__get_next_chars($i, 2);
if ($reached_end) {
break 2;
}
// Exit case
if ($char == '*/') {
$lex_state = LEXER_FREE;
$JS_LEX_TOKENS[] = array('comment', $special_token_value, $i);
$special_token_value = '';
break;
}
$i -= 1;
if (!isset($char[0])) {
break 2;
}
$char = $char[0];
// Normal case
$special_token_value .= $char;
break;
case LEXER_REGEXP:
list($reached_end, $i, $char) = lex__get_next_chars($i, 1);
if ($reached_end) {
break 2;
}
// Exit case
if ($char == '/' && ($i < 2 || $JS_TEXT[$i - 2] != '\\' || $JS_TEXT[$i - 3] == '\\')) {
do {
list($reached_end, $i, $char) = lex__get_next_chars($i, 1);
} while ($char == 'g' || $char == 'i' || $char == 'm');
$i--;
$lex_state = LEXER_FREE;
$JS_LEX_TOKENS[] = array('NEW', $i);
$JS_LEX_TOKENS[] = array('IDENTIFIER', 'RegExp', $i);
$JS_LEX_TOKENS[] = array('BRACKET_OPEN', $i);
$JS_LEX_TOKENS[] = array('string_literal', $special_token_value, $i);
$JS_LEX_TOKENS[] = array('BRACKET_CLOSE', $i);
$JS_VALUE_RANGES[] = array($i - strlen($special_token_value), $i);
$special_token_value = '';
break;
}
// Normal case
$special_token_value .= $char;
break;
case LEXER_DOUBLE_QUOTE_STRING_LITERAL:
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 2;
}
if ($char == "\n" && (strlen($special_token_value) == 0 || $special_token_value[strlen($special_token_value) - 1] == '\\')) {
js_log_warning('LEXER', 'String literals may not contain explicit new lines without special escaping', $i, true);
}
// Exit case
if ($char == '"' && !$escape_flag) {
$lex_state = LEXER_FREE;
$JS_LEX_TOKENS[] = array('string_literal', $special_token_value, $i);
$JS_VALUE_RANGES[] = array($i - strlen($special_token_value) - 1, $i - 1);
$special_token_value = '';
break;
}
// Escape flag based filtering
$actual_char = $char;
if ($escape_flag) {
if ($char == 'n') {
$actual_char = "\n";
} elseif ($char == 'r') {
$actual_char = "\r";
} elseif ($char == 't') {
$actual_char = "\t";
}
} else {
if ($char == '\\') {
$actual_char = '';
}
}
// Normal case
$special_token_value .= $actual_char;
$escape_flag = !$escape_flag && $char == '\\';
break;
case LEXER_SINGLE_QUOTE_STRING_LITERAL:
list($reached_end, $i, $char) = lex__get_next_char($i);
if ($reached_end) {
break 2;
}
if ($char == "\n") {
js_log_warning('LEXER', 'String literals may not contain explicit new lines', $i, true);
}
// Exit case
if ($char == "'" && !$escape_flag) {
$lex_state = LEXER_FREE;
$JS_LEX_TOKENS[] = array('string_literal', $special_token_value, $i);
$JS_VALUE_RANGES[] = array($i - strlen($special_token_value) - 1, $i - 1);
$special_token_value = '';
break;
}
// Escape flag based filtering
$actual_char = $char;
if ($escape_flag) {
if ($char == "'") {
$actual_char = "'";
} elseif ($char == '\\') {
$actual_char = '\\';
} else {
$actual_char = '\\' . $char;
}
} elseif ($char == '\\') {
$actual_char = '';
}
// Normal case
$special_token_value .= $actual_char;
$escape_flag = !$escape_flag && $char == '\\';
break;
}
}
return $JS_LEX_TOKENS;
}