/** * Parse a reference */ public function evaluate(Lexer $lexer) { switch (true) { case $lexer->isNextToken(Lexer::T_SHORT_P): throw new ParserException('Property \\p (Unicode Property) not supported use \\x to specify unicode character or range'); break; case $lexer->isNextToken(Lexer::T_SHORT_UNICODE_X): $lexer->moveNext(); if ($lexer->lookahead['value'] !== '{') { throw new ParserException('Expecting character { after \\X none found'); } $tokens = array(); while ($lexer->moveNext() && $lexer->lookahead['value'] !== '}') { # check if we nested eg.{ddd{d} if ($lexer->lookahead['value'] === '{') { throw new ParserException('Nesting hex value ranges is not allowed'); } if ($lexer->lookahead['value'] !== " " && ctype_xdigit($lexer->lookahead['value']) === false) { throw new ParserException(sprintf('Character %s is not a hexdeciaml digit', $lexer->lookahead['value'])); } $tokens[] = $lexer->lookahead['value']; } # check that current lookahead is a closing character as it's possible to iterate to end of string (i.e. lookahead === null) if ($lexer->lookahead['value'] !== '}') { throw new ParserException('Closing quantifier token `}` not found'); } if (count($tokens) === 0) { throw new ParserException('No hex number found inside the range'); } $number = trim(implode('', $tokens)); return Utf8::chr(hexdec($number)); break; case $lexer->isNextToken(Lexer::T_SHORT_X): // only allow another 2 hex characters $glimpse = $lexer->glimpse(); if ($glimpse['value'] === '{') { throw new ParserException('Braces not supported here'); } $tokens = array(); $count = 2; while ($count > 0 && $lexer->moveNext()) { $tokens[] = $lexer->lookahead['value']; --$count; } $value = trim(implode('', $tokens)); return Utf8::chr(hexdec($value)); break; default: throw new ParserException('No Unicode expression to evaluate'); } }
/** * Parse the current token for new Quantifiers * * @access public * @return ReverseRegex\Generator\Scope a new head * @param ReverseRegex\Generator\Scope $head * @param ReverseRegex\Generator\Scope $set * @param ReverseRegex\Lexer $lexer */ public function parse(Scope $head, Scope $set, Lexer $lexer) { if ($lexer->lookahead['type'] !== Lexer::T_SET_OPEN) { throw new ParserException('Opening character set token not found'); } $peek = $lexer->glimpse(); if ($peek['type'] === Lexer::T_SET_NEGATED) { throw new ParserException('Negated Character Set ranges not supported at this time'); } $normal_lexer = new Lexer($this->normalize($head, $set, $lexer)); while ($normal_lexer->moveNext() && !$normal_lexer->isNextToken(Lexer::T_SET_CLOSE)) { $glimpse = $normal_lexer->glimpse(); if ($glimpse['type'] === Lexer::T_SET_RANGE) { continue; //value be included in range when `-` character is passed } switch (true) { case $normal_lexer->isNextToken(Lexer::T_SET_RANGE): $range_start = $normal_lexer->token['value']; $normal_lexer->moveNext(); if ($normal_lexer->isNextToken(Lexer::T_ESCAPE_CHAR)) { $normal_lexer->moveNext(); } $range_end = $normal_lexer->lookahead['value']; $this->fillRange($head, $range_start, $range_end); break; case $normal_lexer->isNextToken(Lexer::T_LITERAL_NUMERIC) || $normal_lexer->isNextToken(Lexer::T_LITERAL_CHAR): $index = (int) Utf8::ord($normal_lexer->lookahead['value']); $head->setLiteral($index, $normal_lexer->lookahead['value']); break; default: # ignore } } $head->getLiterals()->sort(); return $head; }
/** * Will parse the regex into generator * * @access public * @return */ public function parse($sub = false) { try { while ($this->lexer->moveNext()) { $result = null; $scope = null; $parser = null; switch (true) { case $this->lexer->isNextToken(Lexer::T_GROUP_OPEN): # is the group character the first token? is the regex wrapped in brackets. //if($this->lexer->token === null) { // continue; //} # note this is a new group create new parser instance. $parser = new Parser($this->lexer, new Scope(), new Scope()); $this->left = $parser->parse(true)->getResult(); $this->head->attach($this->left); break; case $this->lexer->isNextToken(Lexer::T_GROUP_CLOSE): # group is finished don't want to contine this loop break = 2 break 2; break; case $this->lexer->isNextTokenAny(array(Lexer::T_LITERAL_CHAR, Lexer::T_LITERAL_NUMERIC)): # test for literal characters (abcd) $this->left = new LiteralScope(); $this->left->addLiteral($this->lexer->lookahead['value']); $this->head->attach($this->left); break; case $this->lexer->isNextToken(Lexer::T_SET_OPEN): # character classes [a-z] $this->left = new LiteralScope(); self::createSubParser('character')->parse($this->left, $this->head, $this->lexer); $this->head->attach($this->left); break; case $this->lexer->isNextTokenAny(array(Lexer::T_DOT, Lexer::T_SHORT_D, Lexer::T_SHORT_NOT_D, Lexer::T_SHORT_W, Lexer::T_SHORT_NOT_W, Lexer::T_SHORT_S, Lexer::T_SHORT_NOT_S)): # match short (. \d \D \w \W \s \S) $this->left = new LiteralScope(); self::createSubParser('short')->parse($this->left, $this->head, $this->lexer); $this->head->attach($this->left); break; case $this->lexer->isNextTokenAny(array(Lexer::T_SHORT_P, Lexer::T_SHORT_UNICODE_X, Lexer::T_SHORT_X)): # match short (\p{L} \x \X ) $this->left = new LiteralScope(); self::createSubParser('unicode')->parse($this->left, $this->head, $this->lexer); $this->head->attach($this->left); break; case $this->lexer->isNextTokenAny(array(Lexer::T_QUANTIFIER_OPEN, Lexer::T_QUANTIFIER_PLUS, Lexer::T_QUANTIFIER_QUESTION, Lexer::T_QUANTIFIER_STAR, Lexer::T_QUANTIFIER_OPEN)): # match quantifiers self::createSubParser('quantifer')->parse($this->left, $this->head, $this->lexer); break; case $this->lexer->isNextToken(Lexer::T_CHOICE_BAR): # match alternations $this->left = $this->head; $this->head = new Scope(); $this->result->useAlternatingStrategy(); $this->result->attach($this->head); break; default: # ignore character } } } catch (ParserException $e) { $pos = $this->lexer->lookahead['position']; $compressed = $this->compress(); throw new ParserException(sprintf('Error found STARTING at position %s after `%s` with msg %s ', $pos, $compressed, $e->getMessage())); } return $this; }
public function convertWhiteSpaceToRange(Scope $head, Scope $result, Lexer $lexer) { if ($lexer->isNextToken(Lexer::T_SHORT_S)) { # spaces, tabs, and line breaks #0009 #0010 #0012 #0013 #0032 $head->addLiteral(chr(9)); $head->addLiteral(chr(10)); $head->addLiteral(chr(12)); $head->addLiteral(chr(13)); $head->addLiteral(chr(32)); } else { # not spaces, tabs, and line breaks #0000-0008 #0011 #0014 - #0031 for ($i = 0; $i <= 8; $i++) { $head->addLiteral(chr($i)); } $head->addLiteral(chr(11)); for ($i = 14; $i <= 31; $i++) { $head->addLiteral(chr($i)); } } }
/** * Parse the current token for closers : {###} { ## } {##,##} * * @access public * @return ReverseRegex\Generator\Scope a new head * @param ReverseRegex\Generator\Scope $head * @param ReverseRegex\Generator\Scope $result * @param ReverseRegex\Lexer $lexer */ public function quantifyClosure(Scope $head, Scope $result, Lexer $lexer) { $tokens = array(); $min = $head->getMinOccurances(); $max = $head->getMaxOccurances(); # move to the first token inside the quantifer. # parse for the minimum , move lookahead until read end of the closure or the `,` while ($lexer->moveNext() === true && !$lexer->isNextToken(Lexer::T_QUANTIFIER_CLOSE) && $lexer->lookahead['value'] !== ',') { if ($lexer->isNextToken(Lexer::T_QUANTIFIER_OPEN)) { throw new ParserException('Nesting Quantifiers is not allowed'); } $tokens[] = $lexer->lookahead; } $min = $this->convertInteger($tokens); # do we have a maximum after the comma? if ($lexer->lookahead['value'] === ',') { # make sure we have values to gather ie not {778,} $tokens = array(); # move to the first token after the `,` character # grab the remaining numbers while ($lexer->moveNext() && !$lexer->isNextToken(Lexer::T_QUANTIFIER_CLOSE)) { if ($lexer->isNextToken(Lexer::T_QUANTIFIER_OPEN)) { throw new ParserException('Nesting Quantifiers is not allowed'); } $tokens[] = $lexer->lookahead; } $max = $this->convertInteger($tokens); } else { $max = $min; } $head->setMaxOccurances($max); $head->setMinOccurances($min); # skip the lexer to the closing token $lexer->skipUntil(Lexer::T_QUANTIFIER_CLOSE); # check if the last matched token was the closing bracket # not going to stop errors like {#####,###{[a-z]} {#####{[a-z]} if (!$lexer->isNextToken(Lexer::T_QUANTIFIER_CLOSE)) { throw new ParserException('Closing quantifier token `}` not found'); } return $head; }