Fixed greedy multiline comment bug, ready for 0.0.2

ANNOUNCE

@@ -0,0 +1,12 @@
+I am pleased to announce version 0.0.2 of language-javascript, a parser for ECMAScript.
+
+Changes in language-javascript 0.0.2 vs. 0.0.1:
+
+  * Multiline comments were processed in greedy form, now end of
+    comment recognised properly. Thanks to Tony Morris for reporting this.
+
+Distributions can be obtained from hackage
+
+  http://hackage.haskell.org/package/language-javascript
+
+language-javascript is distributed under a BSD-style license.

README

@@ -39,4 +39,16 @@ $ cabal install
 
 This will put a utf8 enabled alex in ~/.cabal/bin
 
+To switch back to the standard version, 
+
+$ cabal install --reinstall alex
+
+Changes
+
+0.0.2 Multiline comments were processed in greedy form, now end of
+comment recognised properly. Thanks to Tony Morris for reporting this.
+
+0.0.1 Initial release
+
+EOF
 

language-javascript.cabal

@@ -1,5 +1,5 @@
 Name:                language-javascript
-Version:             0.0.1
+Version:             0.0.2
 Synopsis:            Parser for JavaScript
 Description:         Parses Javascript into an Abstract Syntax Tree (AST).  Initially intended as frontend to hjsmin.
 Homepage:            https://github.com/alanz/language-javascript
@@ -13,6 +13,7 @@ Build-type:          Simple
 homepage:            http://github.com/alanz/language-javascript
 bug-reports:         http://github.com/alanz//language-javascript/issues
 Extra-source-files:  README
+                     ANNOUNCE
                      .ghci
                      buildall.sh
 

runtests.hs

@@ -224,6 +224,10 @@ testSuite = testGroup "Parser"
 
    , testCase "122_jsexec" (testProg "v = getValue(execute(n[0], x)) in getValue(execute(n[1], x));" "Right (JSSourceElementsTop [JSExpression [JSElement \"assignmentExpression\" [JSIdentifier \"v\",JSOperator \"=\",JSExpressionBinary \" in \" [JSIdentifier \"getValue\",JSArguments [[JSIdentifier \"execute\",JSArguments [[JSMemberSquare [JSIdentifier \"n\"] (JSExpression [JSDecimal \"0\"])],[JSIdentifier \"x\"]]]]] [JSIdentifier \"getValue\",JSArguments [[JSIdentifier \"execute\",JSArguments [[JSMemberSquare [JSIdentifier \"n\"] (JSExpression [JSDecimal \"1\"])],[JSIdentifier \"x\"]]]]]]],JSLiteral \";\"])")
 
+
+   , testCase "bug1" (testProg "/* */\nfunction f() {\n/*  */\n}\n" "Right (JSSourceElementsTop [JSFunction (JSIdentifier \"f\") [] (JSFunctionBody [])])")
+   , testCase "bug1" (testProg "/* **/\nfunction f() {\n/*  */\n}\n" "Right (JSSourceElementsTop [JSFunction (JSIdentifier \"f\") [] (JSFunctionBody [])])")
+
     ]
 
 srcHelloWorld = "Hello"

src/Language/JavaScript/Parser/Grammar.y

@@ -149,8 +149,8 @@ import qualified Language.JavaScript.Parser.AST as AST
 -- Comment End   = '*/'
 -- Comment Line  = '//'
 
--- ! ------------------------------------------------- Rules
 
+-- ! ------------------------------------------------- Rules
 
 
 

src/Language/JavaScript/Parser/Lexer.x

@@ -66,6 +66,9 @@ $RegExpFirstChar = [$printable] # [ $cr $lf \* \\ \/]
 -- ~ ( LineTerminator | BSLASH | DIV )
 $RegExpChars = [$printable] # [ $cr $lf \\ \/]
 
+$MultiLineNotAsteriskChar               = [$any_char] # [\*]
+$MultiLineNotForwardSlashOrAsteriskChar = [$any_char] # [\* \/]
+
 -- WhiteSpace ::
 --      <TAB>
 --      <VT>
@@ -89,8 +92,27 @@ tokens :-
 -- Skip one line comment
 <reg,divide> "//"($not_eol_char)*   ;
 
+
+-- ---------------------------------------------------------------------
+-- Comment definition from the ECMAScript spec, ver 3
+
+-- MultiLineComment ::
+--        /* MultiLineCommentChars(opt) */
+-- MultiLineCommentChars ::
+--        MultiLineNotAsteriskChar MultiLineCommentChars(opt)
+--        * PostAsteriskCommentChars(opt)
+-- PostAsteriskCommentChars ::
+--        MultiLineNotForwardSlashOrAsteriskChar MultiLineCommentChars(opt)
+--        * PostAsteriskCommentChars(opt)
+-- MultiLineNotAsteriskChar ::
+--        SourceCharacter but not asterisk *
+-- MultiLineNotForwardSlashOrAsteriskChar ::
+--        SourceCharacter but not forward-slash / or asterisk *
+
 -- Skip multi-line comments. Note: may not nest
-<reg,divide> "/*"($any_char)*"*/"  ;	
+-- <reg,divide> "/*"($any_char)*"*/"  ;	
+<reg,divide> "/*" (($MultiLineNotAsteriskChar)*| ("*")+ ($MultiLineNotForwardSlashOrAsteriskChar) )* ("*")+ "/"  ;	
+
 
 
 -- Identifier    = {ID Head}{ID Tail}*

src/Language/JavaScript/Parser/SrcLocation.hs

@@ -17,21 +17,21 @@ module Language.JavaScript.Parser.SrcLocation (
   SrcLocation (..),
   SrcSpan (..),
   Span (..),
-  spanning,
+  -- spanning,
   mkSrcSpan,
-  combineSrcSpans,
+  -- combineSrcSpans,
   initialSrcLocation,
   spanStartPoint,
   -- * Modification
   incColumn, 
   decColumn,
   incLine,
   incTab,
-  endCol,
+  -- endCol,
   -- * Projection of components of a span
-  endRow,
-  startCol,
-  startRow
+  -- endRow,
+  -- startCol,
+  -- startRow
 ) where
 
 import Data.Data