Loading ...
Sorry, an error occurred while loading the content.

RE: FW: [antlr-interest] Java Cross Referencer "head start" at an tlr. org

Expand Messages
  • Jim O'Connor
    Hi Matt, I m not sure what strategy would be best for class loading. I would be interested in your AggregateClassLoader. The end result of this work will be a
    Message 1 of 2 , Aug 25, 2003
    • 0 Attachment
      Hi Matt,
      I'm not sure what strategy would be best for class loading. I would be
      interested in your AggregateClassLoader.

      The end result of this work will be a commercial application. I would be
      happy to share any work on getting the refs.g grammar completed to resolve
      references. That should lead to one or more methods in CodeProject.java
      like refMethod(). I see the needs of interested programmers diverging at
      that point, anyway.

      Jim

      -----Original Message-----
      From: Matt Benson [mailto:gudnabrsam@...]
      Sent: Monday, August 25, 2003 12:06 PM
      To: antlr-interest@yahoogroups.com
      Subject: Re: FW: [antlr-interest] Java Cross Referencer "head start" at
      antlr. org


      Jim: I would recommend that on your phase II, you
      open up the class loading to pull from any specified
      classloader or set of classloaders (I have an
      implementation of an AggregateClassLoader for this
      purpose if you want it). This way a client of your
      code should be able to use a different JRE than the
      runtime, or they can use the default if they like.
      What license do you plan to use with your work?

      -Matt


      --- Jim O'Connor <Jim.OConnor@...> wrote:
      >
      >
      >
      > Hi Terence,
      > The code is coherent - the grammars "Tool" and the
      > code compiles. The
      > only minor difficulty is the defs.g and refs.g could
      > only "see" the
      > JavaTreeParser when the -glib java.tree.g option was
      > used.
      >
      > I did an extended run-through of the code. I made
      > an Eclipse project to
      > hold/compile/debug the code. I had to Tool refs.g
      > and defs.g outside
      > Eclipse. Expandedrefs.g and Expandeddefs.g helped.
      > MethodCallExpression and objectOrClassExpression
      > are the most
      > interesting/involved of the rules. Are these rules
      > complete/accurate? The
      > resolution of overloaded methods aside, are there
      > deficiencies in the logic?
      > Is there a reason, other than lack of time, for not
      > extending the actions to
      > other areas of the grammar ( expression, etc...)?
      > If anyone else would like
      > to see my extended notes, let me know.
      >
      > My plan:
      > I. Resolve field/variable references as to
      > create a project.refField()
      > method similar to project.refMethod(). This method
      > does nothing other than
      > print out the referencingMethod, referencedMethod
      > and the AST of the method
      > call. Should the project, AST or something else
      > (symbol table?) hold this
      > unused information? I will try to hold off
      > "putting" the information
      > somewhere. The grammars are true to "each phase as
      > simple as possible".
      > Ia. Get file (already there), line and column
      > information to each node.
      > II. Use reflection to load information about
      > imports from jars on
      > classpath.
      > III. Extract/dump information at the end of the
      > REFS phase for each
      > Clase
      > IV. Extract/dump information for whole project
      > where the at the end of
      > Main.java
      > V. Figure out the memory and speed problems
      > VI. Catch and logically proceed on exceptions.
      >
      > Jim
      >
      >
      > Extended notes deleted
      >
      >
      >
      > Your use of Yahoo! Groups is subject to
      > http://docs.yahoo.com/info/terms/
      >
      >


      __________________________________
      Do you Yahoo!?
      Yahoo! SiteBuilder - Free, easy-to-use web site design software
      http://sitebuilder.yahoo.com



      Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    • parrt2000
      ... Makes sense. Let s keep the core as free and open given that I gave this (originally commercial) head start out. It would be great to have a nice java
      Message 2 of 2 , Aug 27, 2003
      • 0 Attachment
        --- In antlr-interest@yahoogroups.com, "Jim O'Connor" <Jim.OConnor@m...> wrote:
        > Hi Matt,
        > I'm not sure what strategy would be best for class loading. I would be
        > interested in your AggregateClassLoader.
        >
        > The end result of this work will be a commercial application. I would be
        > happy to share any work on getting the refs.g grammar completed to resolve
        > references. That should lead to one or more methods in CodeProject.java
        > like refMethod(). I see the needs of interested programmers diverging at
        > that point, anyway.

        Makes sense. Let's keep the core as free and open given that I gave this (originally
        commercial) head start out. It would be great to have a nice java cross-referencer as
        BSD or public domain so I'll help all I can. Sorry about the vague state of the software
        as I left it on the side 2 years ago. The grammars should be just about up to date wi
        the antlr.org grammar.

        Stupid viruses caused yahoogroups to stop sending me antlr-interest mail so I'm
        behind. School starts tomorrow too, but I should have time to help.

        Perhaps Jim you and I should speak by phone to get started. Just email me directly.

        Best regards,
        Terence
      Your message has been successfully submitted and would be delivered to recipients shortly.