Use moriyoshi's fork of kmyacc, which fixes most of the issues of kmyacc with PHP

grammar/README.md

@@ -5,10 +5,10 @@ What do all those files mean?
  * `zend_language_parser.phpy`: PHP grammer written in a pseudo language
  * `analyzer.php`:              Analyzes the `.phpy`-grammer and outputs some info about it
  * `rebuildParser.php`:         Preprocesses the `.phpy`-grammar and builds the parser using `kmyacc`
- * `php.kmyacc`:                A `kmyacc` parser prototype file for PHP
+ * `kmyacc.php.parser`:         A `kmyacc` parser prototype file for PHP
 
 .phpy pseudo language
-=========================
+=====================
 
 The `.phpy` file is a normal grammer in `kmyacc` (`yacc`) style, with some transformations
 applied to it:
@@ -17,3 +17,9 @@ applied to it:
    `new PHPParser_Node_Name(array('subNode1' => ..., 'subNode2' => ...), $line, $docComment)`
  * `Name::abc` is transformed to `PHPParser_Node_Name::abc`
  * Some function-like constructs are resolved (see `rebuildParser.php` for a list)
+
+Building the parser
+===================
+
+In order to rebuild the parser, you need [moriyoshi's fork of kmyacc](https://github.com/moriyoshi/kmyacc-forked).
+After you compiled/installed it, run the `rebuildParser.php` file.

grammar/rebuildParser.php

@@ -1,6 +1,8 @@
 <?php
 
 const GRAMMAR_FILE = './zend_language_parser.phpy';
+const TMP_FILE     = './tmp_parser.phpy';
+const RESULT_FILE  = './tmp_parser.php';
 
 ///////////////////////////////
 /// Utility regex constants ///
@@ -30,54 +32,22 @@ $grammarCode = file_get_contents(GRAMMAR_FILE);
 $grammarCode = preg_replace('~[A-Z][a-zA-Z_]++::~', 'PHPParser_Node_$0', $grammarCode);
 $grammarCode = resolveNodes($grammarCode);
 $grammarCode = resolveMacros($grammarCode);
-$grammarCode = preg_replace('~\$([a-zA-Z_]++)~', '#$1', $grammarCode);
 
-$tmpGrammarFile = tempnam('.', 'tmpGrammarFile');
+file_put_contents(TMP_FILE, $grammarCode);
-file_put_contents($tmpGrammarFile, $grammarCode);
 
 echo 'Building parser.       Output: "',
-     trim(`kmyacc -l -L c -m php.kmyacc -p PHPParser_Parser $tmpGrammarFile 2>&1`),
+     trim(shell_exec('kmyacc -l -m kmyacc.php.parser -p PHPParser_Parser ' . TMP_FILE . ' 2>&1')),
      '"', "\n";
 
-$code = file_get_contents('y.tab.c');
+rename(RESULT_FILE, '../lib/PHPParser/Parser.php');
-$code = str_replace(
-    array(
-        '"$EOF"',
-        '#',
-    ),
-    array(
-        '\'$EOF\'',
-        '$',
-    ),
-    $code
-);
-
-file_put_contents(dirname(__DIR__) . '/lib/PHPParser/Parser.php', $code);
-unlink(__DIR__ . '/y.tab.c');
 
 echo 'Building debug parser. Output: "',
-     trim(`kmyacc -l -t -L c -m php.kmyacc -p PHPParser_ParserDebug $tmpGrammarFile 2>&1`),
+     trim(shell_exec('kmyacc -t -l -m kmyacc.php.parser -p PHPParser_ParserDebug ' . TMP_FILE . ' 2>&1')),
      '"', "\n";
 
-$code = file_get_contents('y.tab.c');
+rename(RESULT_FILE, '../lib/PHPParser/ParserDebug.php');
-$code = str_replace(
-    array(
-        '"$EOF"',
-        '"$start : start"',
-        '#',
-    ),
-    array(
-        '\'$EOF\'',
-        '\'$start : start\'',
-        '$',
-    ),
-    $code
-);
 
-file_put_contents(dirname(__DIR__) . '/lib/PHPParser/ParserDebug.php', $code);
+unlink(TMP_FILE);
-unlink(__DIR__ . '/y.tab.c');
-
-unlink($tmpGrammarFile);
 
 echo 'The following temporary preproprocessed grammar file was used:', "\n", $grammarCode;
 

lib/PHPParser/Parser.php

@@ -147,7 +147,7 @@ class PHPParser_Parser
     // }}}
 
     private static $yyterminals = array(
-        '$EOF',
+        "EOF",
         "error",
         "T_INCLUDE",
         "T_INCLUDE_ONCE",

lib/PHPParser/ParserDebug.php

@@ -147,7 +147,7 @@ class PHPParser_ParserDebug
     // }}}
 
     private static $yyterminals = array(
-        '$EOF',
+        "EOF",
         "error",
         "T_INCLUDE",
         "T_INCLUDE_ONCE",
@@ -296,7 +296,7 @@ class PHPParser_ParserDebug
     );
 
     private static $yyproduction = array(
-        '$start : start',
+        "start : start",
         "start : top_statement_list",
         "top_statement_list : top_statement_list top_statement",
         "top_statement_list : /* empty */",