diff options
| -rw-r--r-- | transform.php | 194 |
1 files changed, 194 insertions, 0 deletions
diff --git a/transform.php b/transform.php new file mode 100644 index 0000000..788b5fd --- /dev/null +++ b/transform.php @@ -0,0 +1,194 @@ +<?php + +class TextNode +{ + public $text; +} + +class TagNode +{ + public $name; + public $params; + public $sub; +} + +class VariableNode +{ + public $name; + public $arrayfields; +} + +/* $text must start after the first opening bracket */ +function find_closing_bracket($text, $opening, $closing) +{ + $counter = 1; + $len = strlen($text); + for($i = 0; $i < $len; ++$i) + { + switch($text[$i]) + { + case $opening: + ++$counter; + break; + case $closing: + --$counter; + break; + } + if($counter == 0) + break; + } + + if($counter > 0) + throw new Exception("Missing closing \"$closing\". Stop."); + + return $i; +} + +function unescape_text($text) +{ + $text = preg_replace("/(?:(?<!\\\\)\\\\\\\$)/s", "$", $text); + return str_replace("\\\\", "\\", $text); +} + +function tokenize_text($text) +{ + $tokens = array(); + /* Find next non-escaped $-char */ + if(preg_match("/(?:(?<!\\\\)\\$)/s", $text, $match, PREG_OFFSET_CAPTURE) == 0) + { + $node = new TextNode(); + $node->text = unescape_text($text); + return (strlen($node->text) == 0) ? array() : array($node); + } + + if($match[0][1] > 0) + { + $node = new TextNode(); + $node->text = unescape_text(substr($text, 0, $match[0][1])); + $tokens[] = $node; + } + + if($text[$match[0][1] + 1] == "{") + { + $varend = find_closing_bracket(substr($text, $match[0][1] + 2), "{", "}") + $match[0][1] + 2; + return array_merge( + $tokens, + tokenize_text("\$" . substr($text, $match[0][1] + 2, ($varend - 1) - ($match[0][1] + 1))), + tokenize_text(substr($text, $varend + 1)) + ); + } + + $text = substr($text, $match[0][1] + 1); + if(preg_match("/^[a-zA-Z0-9_]+/s", $text, $match, PREG_OFFSET_CAPTURE) == 0) + { + $nexttokens = tokenize_text($text); + if($nexttokens[0] instanceof Text) + $nexttokens[0]->text = "\$" . $nexttokens[0]->text; + else + { + $node = new TextNode(); + $node->text = "\$"; + $tokens[] = $node; + } + return array_merge($tokens, $nexttokens); + } + + $node = new VariableNode(); + $node->name = $match[0][0]; + $node->arrayfields = array(); + + $text = substr($text, $match[0][1] + strlen($match[0][0])); + while(@$text[0] == "[") + { + $text = substr($text, 1); + $fieldend = find_closing_bracket($text, "[", "]"); + $node->arrayfields[] = tokenize_text(substr($text, 0, $fieldend)); + $text = substr($text, $fieldend + 1); + } + + $tokens[] = $node; + + return strlen($text) > 0 ? array_merge($tokens, tokenize_text($text)) : $tokens; +} + +function mk_ast($code) +{ + $ast = array(); + + if(preg_match("/\\<\\s*ste:([a-zA-Z0-9_]*)/s", $code, $matches, PREG_OFFSET_CAPTURE) == 0) + return tokenize_text($code); + + $ast = tokenize_text(substr($code, 0, $matches[0][1])); + + $tag = new TagNode(); + $tag->name = $matches[1][0]; + + $code = substr($code, $matches[0][1] + strlen($matches[0][0])); + + $tag->params = array(); + + while(preg_match("/^\\s+([a-zA-Z0-9_]+)=((?:\"(?:.*?)(?<!\\\\)\")|(?:'(?:.*?)(?<!\\\\)'))/s", $code, $matches, PREG_OFFSET_CAPTURE) > 0) + { + $paramval = substr($code, $matches[2][1] + 1, strlen($matches[2][0]) - 2); + $paramval = str_replace("\\\"", "\"", $paramval); + $paramval = str_replace("\\'", "'", $paramval); + $tag->params[$matches[1][0]] = tokenize_text($paramval); + $code = substr($code, strlen($matches[0][0])); + } + + if(preg_match("/^\\s*([\\/]?)\\s*\\>/s", $code, $matches) == 0) + throw new Exception("Missing closing '>' in \"" . $tag->name . "\"-Tag. Stop. (:::CODE START:::$code:::CODE END:::)"); + + $code = substr($code, strlen($matches[0])); + + $tag->sub = array(); + + if($matches[1][0] != "/") + { + $tags_open = 1; + $off = 0; + $last_tag_start = 0; + while(preg_match("/\\<((?:\\s*)|(?:\\s*\\/\\s*))ste:([a-zA-Z0-9_]*)(?:\\s+(?:[a-zA-Z0-9_]+)=(?:(?:\"(?:.*?)(?<!\\\\)\")|(?:'(?:.*?)(?<!\\\\)')))*\\s*(?<!\\/)\\>/s", $code, $matches, PREG_OFFSET_CAPTURE, $off) > 0) /* RegEx from hell! Matches all non-selfclosing <ste:> Tags. Opening and closing ones. */ + { + $closingtag = trim($matches[1][0]); + if($closingtag[0] == "/") + $tags_open--; + else + $tags_open++; + $last_tag_start = $matches[0][1]; + $off = $last_tag_start + strlen($matches[0][0]); + if($tags_open == 0) + break; + } + + if(($tags_open != 0) or ($tag->name != $matches[2][0])) + throw new Exception("Missing closing \"ste:" + $tag->name + "\"-Tag. Stop."); + + $tag->sub = mk_ast(substr($code, 0, $last_tag_start)); + $code = substr($code, $off); + } + + $ast[] = $tag; + return array_merge($ast, strlen($code) > 0 ? mk_ast($code) : array()); +} + +function parse($code) +{ + /* Precompiling... */ + $code = preg_replace("/\\<\\s*ste:comment\\s*\\>.*?\\<\\s*\\/\\s*ste:comment\\s*\\>/s", "", $code); /* Remove comments */ + $code = preg_replace( /* Transform short form of if-clause (?{cond|then|else}) to long form */ + "/(?:(?<!\\\\)\\?)(?:(?<!\\\\)\\{)(.*?)(?:(?<!\\\\)\\|)(.*?)(?:(?<!\\\\)\\|)(.*?)(?:(?<!\\\\)\\})/s", + "<ste:if>\$1<ste:then>\$2</ste:then><ste:else>\$3</ste:else></ste:if>", + $code + ); + /* Unescape \? \{ \} \| */ + $code = preg_replace("/(?:(?<!\\\\)\\\\\\?)/s", "?", $code); + $code = preg_replace("/(?:(?<!\\\\)\\\\\\{)/s", "{", $code); + $code = preg_replace("/(?:(?<!\\\\)\\\\\\})/s", "}", $code); + $code = preg_replace("/(?:(?<!\\\\)\\\\\\|)/s", "|", $code); + + /* Create abstract syntax tree */ + return mk_ast($code); +} + +?> |