phpython Code

--- a
+++ b/trunk/python/python-tokenizer.php
@@ -0,0 +1,153 @@
+<!--
+
+http://svn.python.org/view/python/trunk/Lib/tokenize.py?view=markup
+
+
+Whitespace = r'[ \f\t]*'
+Comment = r'#[^\r\n]*'
+Ignore = Whitespace + any(r'\\\r?\n' + Whitespace) + maybe(Comment)
+Name = r'[a-zA-Z_]\w*'
+
+Hexnumber = r'0[xX][\da-fA-F]+[lL]?'
+Octnumber = r'(0[oO][0-7]+)|(0[0-7]*)[lL]?'
+Binnumber = r'0[bB][01]+[lL]?'
+Decnumber = r'[1-9]\d*[lL]?'
+Intnumber = group(Hexnumber, Binnumber, Octnumber, Decnumber)
+Exponent = r'[eE][-+]?\d+'
+Pointfloat = group(r'\d+\.\d*', r'\.\d+') + maybe(Exponent)
+Expfloat = r'\d+' + Exponent
+Floatnumber = group(Pointfloat, Expfloat)
+Imagnumber = group(r'\d+[jJ]', Floatnumber + r'[jJ]')
+Number = group(Imagnumber, Floatnumber, Intnumber)
+
+# Tail end of ' string.
+Single = r"[^'\\]*(?:\\.[^'\\]*)*'"
+# Tail end of " string.
+Double = r'[^"\\]*(?:\\.[^"\\]*)*"'
+# Tail end of ''' string.
+Single3 = r"[^'\\]*(?:(?:\\.|'(?!''))[^'\\]*)*'''"
+# Tail end of """ string.
+Double3 = r'[^"\\]*(?:(?:\\.|"(?!""))[^"\\]*)*"""'
+Triple = group("[uU]?[rR]?'''", '[uU]?[rR]?"""')
+# Single-line ' or " string.
+String = group(r"[uU]?[rR]?'[^\n'\\]*(?:\\.[^\n'\\]*)*'",
+               r'[uU]?[rR]?"[^\n"\\]*(?:\\.[^\n"\\]*)*"')
+
+# Because of leftmost-then-longest match semantics, be sure to put the
+# longest operators first (e.g., if = came before ==, == would get
+# recognized as two instances of =).
+Operator = group(r"\*\*=?", r">>=?", r"<<=?", r"<>", r"!=",
+                 r"//=?",
+                 r"[+\-*/%&|^=<>]=?",
+                 r"~")
+
+Bracket = '[][(){}]'
+Special = group(r'\r?\n', r'[:;.,`@]')
+Funny = group(Operator, Bracket, Special)
+
+PlainToken = group(Number, Funny, String, Name)
+Token = Ignore + PlainToken
+
+# First (or only) line of ' or " string.
+ContStr = group(r"[uU]?[rR]?'[^\n'\\]*(?:\\.[^\n'\\]*)*" +
+                group("'", r'\\\r?\n'),
+                r'[uU]?[rR]?"[^\n"\\]*(?:\\.[^\n"\\]*)*' +
+                group('"', r'\\\r?\n'))
+PseudoExtras = group(r'\\\r?\n', Comment, Triple)
+PseudoToken = Whitespace + group(PseudoExtras, Number, Funny, ContStr, Name)
+
+tokenprog, pseudoprog, single3prog, double3prog = map(
+    re.compile, (Token, PseudoToken, Single3, Double3))
+endprogs = {"'": re.compile(Single), '"': re.compile(Double),
+            "'''": single3prog, '"""': double3prog,
+            "r'''": single3prog, 'r"""': double3prog,
+            "u'''": single3prog, 'u"""': double3prog,
+            "ur'''": single3prog, 'ur"""': double3prog,
+            "R'''": single3prog, 'R"""': double3prog,
+            "U'''": single3prog, 'U"""': double3prog,
+            "uR'''": single3prog, 'uR"""': double3prog,
+            "Ur'''": single3prog, 'Ur"""': double3prog,
+            "UR'''": single3prog, 'UR"""': double3prog,
+            "b'''": single3prog, 'b"""': double3prog,
+            "br'''": single3prog, 'br"""': double3prog,
+            "B'''": single3prog, 'B"""': double3prog,
+            "bR'''": single3prog, 'bR"""': double3prog,
+            "Br'''": single3prog, 'Br"""': double3prog,
+            "BR'''": single3prog, 'BR"""': double3prog,
+            'r': None, 'R': None, 'u': None, 'U': None,
+            'b': None, 'B': None}
+
+triple_quoted = {}
+for t in ("'''", '"""',
+          "r'''", 'r"""', "R'''", 'R"""',
+          "u'''", 'u"""', "U'''", 'U"""',
+          "ur'''", 'ur"""', "Ur'''", 'Ur"""',
+          "uR'''", 'uR"""', "UR'''", 'UR"""',
+          "b'''", 'b"""', "B'''", 'B"""',
+          "br'''", 'br"""', "Br'''", 'Br"""',
+          "bR'''", 'bR"""', "BR'''", 'BR"""'):
+    triple_quoted[t] = t
+single_quoted = {}
+for t in ("'", '"',
+          "r'", 'r"', "R'", 'R"',
+          "u'", 'u"', "U'", 'U"',
+          "ur'", 'ur"', "Ur'", 'Ur"',
+          "uR'", 'uR"', "UR'", 'UR"',
+          "b'", 'b"', "B'", 'B"',
+          "br'", 'br"', "Br'", 'Br"',
+          "bR'", 'bR"', "BR'", 'BR"' ):
+    single_quoted[t] = t
+
+tabsize = 8
+-->
+
+<?php
+	function group($re) {
+		if (is_array($re)) {
+			return "(".implode("|",$re).")";
+		} else {
+			return "(".$re.")";
+		}
+	}
+	function any($re) {	
+		return group($re)."*";
+	}
+	function maybe($re) {
+		return group($re)."?";	
+	}
+
+	$preg_whitespace 	= "[ \f\t]*";
+	$preg_comment 		= "#[^\r\n]*";
+	$preg_ignore 		= $preg_whitespace + any("\\\r?\n" + $preg_whitespace) + maybe($preg_comment);
+	$preg_name 			= "[a-zA-Z_]\w*";
+	
+
+	function python_tokenize($pycode) {	
+		$line = $pycode;
+
+		$out = array();
+		while (strlen($line)) {
+			if (preg_match('/^[ \t]+(\.[0-9]*)?/', $line, $regs)) {
+				# Ignored
+				$out[] = $regs[0];
+				$line = substr($line, strlen($regs[0]));
+			
+			} else if (preg_match('/^[0-9]+(\.[0-9]*)?/', $line, $regs)) {
+				# It's a number
+				$out[] = $regs[0];
+				$line = substr($line, strlen($regs[0]));
+
+			} else if (preg_match('/^[A-Za-z]+/', $line, $regs)) {
+				# It's a variable name
+				$out[] = $regs[0];
+				$line = substr($line, strlen($regs[0]));
+
+			} else {
+				# It's some other character
+				$out[] = $line[0];
+				$line = substr($line, 1);
+			}
+		}
+		return $out;
+	}
+?>