transmogrify-development — Developers' mailing list

[Transmogrify-development] Matt Benson has started following you on reference.me!

[reference.me <em...@re...>]

Hi Transmogrify Development List,

Matt Benson has started following you on reference.me!
Wondering who else is following you? click here.



Tired of hearing from us? Click here to <% unsubscribe %>

Re: [Transmogrify-development] (no subject)

[Matt B. <gud...@ya...>]

Andrew:
  If you have time I did still want to hear your ideas
on how better to incrementally analyze the syntax tree
in Transmogrify.  Been busy for the past month and
find myself with a little free time again now.

Thanks,
Matt


--- Matt Benson <gud...@ya...> wrote:
> Thanks for the replies, Andrew and John!
> 
> First, you say that you don't have the itch anymore
> since using Eclipse?  It's been awhile since I've
> looked into it, but are you saying that Eclipse has
> plugins that do Transmogrify's job at least as well
> as
> Transmogrify?
> 
> Next, you mentioned that Transmogrify is ill-suited
> for dynamic things such as what I'm going after...
> wish you'd told me sooner! :)  but anyway, I would
> definitely like to hear anything you can offer in
> this
> regard.
> 
> -Matt
> 
> --- "Andrew McCormick (smileyy)"
> <sm...@fi...> wrote:
> > On Thu, 4 Apr 2002, Matt Benson wrote:
> > 
> > > Okay, maybe I shouldn't have included a huge
> stack
> > > trace in a message.  Has Transmogrify been
> > abandoned?
> > 
> > Abandoned?  Hmm...maybe, unfortunately. 
> Personally,
> > I'm lacking the
> > appropriate combination of time and interesting to
> > be working on it.  Add
> > to the fact that I've been using IBM's Eclipse as
> my
> > IDE, and I don't even
> > have the itch to scratch any more.
> > 
> > This bit:
> > 
> > net.sourceforge.transmogrify.symtab.UnknownClass
> >         at
> >
>
net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java:1235)
> >         at
> >
>
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Resolver.java:1205)
> > 
> > means that its having problems with some sort of
> > arithmetic expression
> > (any of the arithmetic operators, I believe).  Not
> > having the code in
> > front of me, I can't tell you more, but maybe that
> > helps narrow things
> > down some.  That seems odd, because the result
> type
> > of that isn't hard to
> > figure out.  It would probably help if
> transmogrify
> > did a better job of
> > indicating where in the code the error is
> happening
> > (I thought it did =/)
> > -- shouldn't be too hard, all the AST nodes have
> > file and line
> > information in them.
> > 
> > Also, IIRC, transmogrify is pretty ill-suited for
> > dynamic situations like
> > you're going after.  At the very least, it needs
> to
> > to a better job of
> > incrementally reanalyzing the AST.  I've had a few
> > ideas on this, but
> > never implemented them at all.  I can share if
> > anyone's interested.
> > 
> > andrew
> > -- 
> > sm...@fi...
> > 
> > 
> > 
> > _______________________________________________
> > Transmogrify-development mailing list
> > Tra...@li...
> >
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development
> 
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Tax Center - online filing with TurboTax
> http://taxes.yahoo.com/
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! Shopping - Mother's Day is May 12th!
http://shopping.yahoo.com

Re: [Transmogrify-development] Transmogrify and Jext

[Matt B. <gud...@ya...>]

John:
     I appreciate the offer to help but it will have
to wait until I get what I am working on stable enough
to duplicate the error!  :)  I have already made
changes such that other errors are happening first...
You were correct, the offending line is from my own
Hook implementation.

By the way, everybody, I have just downloaded Eclipse
and am testing out its code completion.  This feature
is pretty impressive but I would really miss a lot of
Jext's features if I were to switch, plus there's just
something about being partly responsible (in however
minor a capacity) for the welfare of that project.  I
started the DL of Eclipse and I kind of got worried as
to how it would be if I liked it a lot and decided to
use it in favor of Jext.  It sort of seemed like how I
would feel if I fell in love with someone other than
my wife... Then I opened up Eclipse and it seems more
like there was nothing to worry about after all... :)

-Matt

--- John P N Pybus <joh...@zo...>
wrote:
> On Monday 01 Apr 2002 6:12 pm, Matt Benson wrote:
> > I'm back!  This list has had very little activity
> over
> > the last several months it seems, so I hope
> someone's
> > still out there listening.  I am trying to finish
> this
> > port to Jext because my reputation is plummeting! 
> :)
> > I have been busy/lazy but I am going to try to
> devote
> > a few hours each week to finishing this.  Anyway,
> I am
> > really close, meaning that all my background code
> is
> > finished and compiles.  I am now trying to debug
> it
> > all.  The first thing I am getting is this stack
> > trace:
> >
> > java.lang.ClassCastException:
> > net.sourceforge.transmogrify.symtab.UnknownClass
> > 	at
> >
>
net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java
> >:1235) at
> >
>
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Re
> >solver.java:1205) at
> >
>
net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.jav
> >a:558) at
> 
> Hi Matt,
> 
> I can actually remember debugging almost this exact
> error a couple of months 
> back writting my own Hook.  It's caused by
> binaryResultType() expecting a 
> numeric type (i.e. subclass of ExternalClass) and
> getting a type of 
> UnknownClass, due to the arguments not having been
> rpoperly resolved.  Can't 
> remember exactly what I found out though, or
> remember whether I could fix it, 
> or find any notes of what I did.  Perhaps I'll have
> left some notes on my 
> home computer, although it's quite possible I didn't
> record anything - how 
> stupid!
> 
> > Error resolving near [+,<PLUS>][748,59] in
> >
>
C:\transmogrify\src\net\sourceforge\transmogrify\hook\jext\JextHook.java
> > class
> >
> net.sourceforge.transmogrify.symtab.UnknownClass[+]
> > class
> >
>
net.sourceforge.transmogrify.symtab.ExternalClass[int]
> >
> >
> > which appears to be caused by the line
> >
> > return getTextArea().getLineStartOffset(lineNumber
> -
> > 1) + offset - 1;
> >
> > in my Hook implementation.  I would think this is
> > indicative of a bug in the parsing.
> 
> What were you trying to parse, your own Hook's
> sources?  Are you using the CVS 
> transmogrify?  Can you send me a recipe to cause
> this.  Preferably a 
> standalone app, and the .java files for it to parse
> which cause the bug.  If 
> so I'll have a look at it again.
> 
> John
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/

Re: [Transmogrify-development] Transmogrify and Jext

[John P N P. <joh...@zo...>]

On Monday 01 Apr 2002 6:12 pm, Matt Benson wrote:
> I'm back!  This list has had very little activity over
> the last several months it seems, so I hope someone's
> still out there listening.  I am trying to finish this
> port to Jext because my reputation is plummeting!  :)
> I have been busy/lazy but I am going to try to devote
> a few hours each week to finishing this.  Anyway, I am
> really close, meaning that all my background code is
> finished and compiles.  I am now trying to debug it
> all.  The first thing I am getting is this stack
> trace:
>
> java.lang.ClassCastException:
> net.sourceforge.transmogrify.symtab.UnknownClass
> 	at
> net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java
>:1235) at
> net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Re
>solver.java:1205) at
> net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.jav
>a:558) at

Hi Matt,

I can actually remember debugging almost this exact error a couple of months 
back writting my own Hook.  It's caused by binaryResultType() expecting a 
numeric type (i.e. subclass of ExternalClass) and getting a type of 
UnknownClass, due to the arguments not having been rpoperly resolved.  Can't 
remember exactly what I found out though, or remember whether I could fix it, 
or find any notes of what I did.  Perhaps I'll have left some notes on my 
home computer, although it's quite possible I didn't record anything - how 
stupid!

> Error resolving near [+,<PLUS>][748,59] in
> C:\transmogrify\src\net\sourceforge\transmogrify\hook\jext\JextHook.java
> class
> net.sourceforge.transmogrify.symtab.UnknownClass[+]
> class
> net.sourceforge.transmogrify.symtab.ExternalClass[int]
>
>
> which appears to be caused by the line
>
> return getTextArea().getLineStartOffset(lineNumber -
> 1) + offset - 1;
>
> in my Hook implementation.  I would think this is
> indicative of a bug in the parsing.

What were you trying to parse, your own Hook's sources?  Are you using the CVS 
transmogrify?  Can you send me a recipe to cause this.  Preferably a 
standalone app, and the .java files for it to parse which cause the bug.  If 
so I'll have a look at it again.

John

Re: [Transmogrify-development] (no subject)

[Matt B. <gud...@ya...>]

Thanks for the replies, Andrew and John!

First, you say that you don't have the itch anymore
since using Eclipse?  It's been awhile since I've
looked into it, but are you saying that Eclipse has
plugins that do Transmogrify's job at least as well as
Transmogrify?

Next, you mentioned that Transmogrify is ill-suited
for dynamic things such as what I'm going after...
wish you'd told me sooner! :)  but anyway, I would
definitely like to hear anything you can offer in this
regard.

-Matt

--- "Andrew McCormick (smileyy)"
<sm...@fi...> wrote:
> On Thu, 4 Apr 2002, Matt Benson wrote:
> 
> > Okay, maybe I shouldn't have included a huge stack
> > trace in a message.  Has Transmogrify been
> abandoned?
> 
> Abandoned?  Hmm...maybe, unfortunately.  Personally,
> I'm lacking the
> appropriate combination of time and interesting to
> be working on it.  Add
> to the fact that I've been using IBM's Eclipse as my
> IDE, and I don't even
> have the itch to scratch any more.
> 
> This bit:
> 
> net.sourceforge.transmogrify.symtab.UnknownClass
>         at
>
net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java:1235)
>         at
>
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Resolver.java:1205)
> 
> means that its having problems with some sort of
> arithmetic expression
> (any of the arithmetic operators, I believe).  Not
> having the code in
> front of me, I can't tell you more, but maybe that
> helps narrow things
> down some.  That seems odd, because the result type
> of that isn't hard to
> figure out.  It would probably help if transmogrify
> did a better job of
> indicating where in the code the error is happening
> (I thought it did =/)
> -- shouldn't be too hard, all the AST nodes have
> file and line
> information in them.
> 
> Also, IIRC, transmogrify is pretty ill-suited for
> dynamic situations like
> you're going after.  At the very least, it needs to
> to a better job of
> incrementally reanalyzing the AST.  I've had a few
> ideas on this, but
> never implemented them at all.  I can share if
> anyone's interested.
> 
> andrew
> -- 
> sm...@fi...
> 
> 
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/

Re: [Transmogrify-development] (no subject)

[John P N P. <joh...@zo...>]

On Friday 05 Apr 2002 5:29 am, Andrew McCormick (smileyy) wrote:

> Abandoned?  Hmm...maybe, unfortunately.  Personally, I'm lacking the
> appropriate combination of time and interesting to be working on it.  Add
> to the fact that I've been using IBM's Eclipse as my IDE, and I don't even
> have the itch to scratch any more.

I still have some interest, although work has meant I have decreasing amounts 
of time to spend on it.

> Also, IIRC, transmogrify is pretty ill-suited for dynamic situations like
> you're going after.  At the very least, it needs to to a better job of
> incrementally reanalyzing the AST.  I've had a few ideas on this, but
> never implemented them at all.  I can share if anyone's interested.

I would.  It's something I'd like to improve in transmogrify.  If you share 
your thoughts it might refire my enthusiasm.

Yours,

John

Re: [Transmogrify-development] (no subject)

[Andrew M. (smileyy) <sm...@fi...>]

On Thu, 4 Apr 2002, Matt Benson wrote:

> Okay, maybe I shouldn't have included a huge stack
> trace in a message.  Has Transmogrify been abandoned?

Abandoned?  Hmm...maybe, unfortunately.  Personally, I'm lacking the
appropriate combination of time and interesting to be working on it.  Add
to the fact that I've been using IBM's Eclipse as my IDE, and I don't even
have the itch to scratch any more.

This bit:

net.sourceforge.transmogrify.symtab.UnknownClass
        at
net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java:1235)
        at
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Resolver.java:1205)

means that its having problems with some sort of arithmetic expression
(any of the arithmetic operators, I believe).  Not having the code in
front of me, I can't tell you more, but maybe that helps narrow things
down some.  That seems odd, because the result type of that isn't hard to
figure out.  It would probably help if transmogrify did a better job of
indicating where in the code the error is happening (I thought it did =/)
-- shouldn't be too hard, all the AST nodes have file and line
information in them.

Also, IIRC, transmogrify is pretty ill-suited for dynamic situations like
you're going after.  At the very least, it needs to to a better job of
incrementally reanalyzing the AST.  I've had a few ideas on this, but
never implemented them at all.  I can share if anyone's interested.

andrew
-- 
sm...@fi...

[Transmogrify-development] (no subject)

[Matt B. <gud...@ya...>]

Okay, maybe I shouldn't have included a huge stack
trace in a message.  Has Transmogrify been abandoned?

Anyway, that complicated expression still looks like a
bug to me.  I am willing to hack the core of
Transmogrify but it's looking very tricky.  One thing
I had hoped to do from the start is to build trees
from interdependent source files not yet compiled.  So
far though it seems like it's not possible, since I
want to build classes that import other classes I am
editing.  Has anyone ever thought about this at all? 
Any response at all will be appreciated just so I know
I'm not out here alone...

-Matt

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/

Re: [Transmogrify-development] Transmogrify and Jext

[Matt B. <gud...@ya...>]

I'm back!  This list has had very little activity over
the last several months it seems, so I hope someone's
still out there listening.  I am trying to finish this
port to Jext because my reputation is plummeting!  :) 
I have been busy/lazy but I am going to try to devote
a few hours each week to finishing this.  Anyway, I am
really close, meaning that all my background code is
finished and compiles.  I am now trying to debug it
all.  The first thing I am getting is this stack
trace:

java.lang.ClassCastException:
net.sourceforge.transmogrify.symtab.UnknownClass
	at
net.sourceforge.transmogrify.symtab.Resolver.binaryResultType(Resolver.java:1235)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Resolver.java:1205)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.java:558)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveArithmeticExpression(Resolver.java:1200)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.java:558)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.java:408)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolveExpression(Resolver.java:408)
	at
net.sourceforge.transmogrify.symtab.Resolver.handleSList(Resolver.java:59)
	at
net.sourceforge.transmogrify.symtab.Resolver.handleMethod(Resolver.java:145)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverse(DefinitionTraverser.java:92)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverse(DefinitionTraverser.java:59)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverseChildren(DefinitionTraverser.java:121)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverse(DefinitionTraverser.java:84)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traversePackage(DefinitionTraverser.java:113)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverse(DefinitionTraverser.java:74)
	at
net.sourceforge.transmogrify.symtab.DefinitionTraverser.traverse(DefinitionTraverser.java:41)
	at
net.sourceforge.transmogrify.symtab.Resolver.resolve(Resolver.java:52)
	at
net.sourceforge.transmogrify.symtab.TableMaker.resolveReferences(TableMaker.java:200)
	at
net.sourceforge.transmogrify.symtab.TableMaker.getTable(TableMaker.java:81)
	at
net.sourceforge.transmogrify.hook.Hook.parseFiles(Hook.java:231)
	at
net.sourceforge.transmogrify.hook.Hook.parseFiles(Hook.java:244)
	at
net.sourceforge.transmogrify.hook.jext.JextHook.parse(JextHook.java:154)
	at
net.sourceforge.transmogrify.hook.jext.JextHook.<init>(JextHook.java:143)
	at
net.sourceforge.transmogrify.hook.jext.JextHookManager.updateProject(JextHookManager.java:153)
	at
net.sourceforge.transmogrify.hook.jext.JextHookManager.projectEventFired(JextHookManager.java:133)
	at
org.jext.project.AbstractProjectManager.fireProjectEvent(Unknown
Source)
	at
org.jext.project.DefaultProjectManager.jextEventFired(Unknown
Source)
	at org.jext.JextFrame.fireJextEvent(Unknown Source)
	at org.jext.JextTabbedPane.stateChanged(Unknown
Source)
	at javax.swing.JTabbedPane.fireStateChanged(Unknown
Source)
	at
javax.swing.JTabbedPane$ModelListener.stateChanged(Unknown
Source)
	at
javax.swing.DefaultSingleSelectionModel.fireStateChanged(Unknown
Source)
	at
javax.swing.DefaultSingleSelectionModel.setSelectedIndex(Unknown
Source)
	at javax.swing.JTabbedPane.setSelectedIndex(Unknown
Source)
	at javax.swing.JTabbedPane.insertTab(Unknown Source)
	at javax.swing.JTabbedPane.addTab(Unknown Source)
	at org.jext.JextTabbedPane.addTab(Unknown Source)
	at javax.swing.JTabbedPane.add(Unknown Source)
	at org.jext.misc.Workspaces.valueChanged(Unknown
Source)
	at
javax.swing.JList.fireSelectionValueChanged(Unknown
Source)
	at
javax.swing.JList$ListSelectionHandler.valueChanged(Unknown
Source)
	at
javax.swing.DefaultListSelectionModel.fireValueChanged(Unknown
Source)
	at
javax.swing.DefaultListSelectionModel.fireValueChanged(Unknown
Source)
	at
javax.swing.DefaultListSelectionModel.setValueIsAdjusting(Unknown
Source)
	at javax.swing.JList.setValueIsAdjusting(Unknown
Source)
	at
javax.swing.plaf.basic.BasicListUI$MouseInputHandler.mouseReleased(Unknown
Source)
	at java.awt.AWTEventMulticaster.mouseReleased(Unknown
Source)
	at java.awt.Component.processMouseEvent(Unknown
Source)
	at java.awt.Component.processEvent(Unknown Source)
	at java.awt.Container.processEvent(Unknown Source)
	at java.awt.Component.dispatchEventImpl(Unknown
Source)
	at java.awt.Container.dispatchEventImpl(Unknown
Source)
	at java.awt.Component.dispatchEvent(Unknown Source)
	at
java.awt.LightweightDispatcher.retargetMouseEvent(Unknown
Source)
	at
java.awt.LightweightDispatcher.processMouseEvent(Unknown
Source)
	at
java.awt.LightweightDispatcher.dispatchEvent(Unknown
Source)
	at java.awt.Container.dispatchEventImpl(Unknown
Source)
	at java.awt.Window.dispatchEventImpl(Unknown Source)
	at java.awt.Component.dispatchEvent(Unknown Source)
	at java.awt.EventQueue.dispatchEvent(Unknown Source)
	at java.awt.EventDispatchThread.pumpOneEvent(Unknown
Source)
	at java.awt.EventDispatchThread.pumpEvents(Unknown
Source)
	at java.awt.EventDispatchThread.run(Unknown Source)
Error resolving near [+,<PLUS>][748,59] in
C:\transmogrify\src\net\sourceforge\transmogrify\hook\jext\JextHook.java
class
net.sourceforge.transmogrify.symtab.UnknownClass[+]
class
net.sourceforge.transmogrify.symtab.ExternalClass[int]


which appears to be caused by the line

return getTextArea().getLineStartOffset(lineNumber -
1) + offset - 1;

in my Hook implementation.  I would think this is
indicative of a bug in the parsing.  Does anyone have
any ideas?

Thanks,
Matt

__________________________________________________
Do You Yahoo!?
Yahoo! Greetings - send holiday greetings for Easter, Passover
http://greetings.yahoo.com/

Re: [Transmogrify-development] refactorings.xml

[Matt B. <gud...@ya...>]

Thanks; sorry if that last question sounded
ignorant--it was, perhaps.  I have worked with XML
very little but now that I look at the JBuilder stuff
more closely it does look remarkably similar to what I
am doing with Jext.  The two share many architecture
similarities.  Hopefully I will be debugging
refactorings soon instead of redesigning my menu
loading algorithms...


--- John P N Pybus <joh...@zo...>
wrote:
> On Friday 23 Nov 2001 5:12 pm, Matt Benson wrote:
> > Do these files have or need a DTD?
> >
> 
> Don't ask me - I didn't do it!
> 
> In fact not only did I not write the jbuilder hook,
> but I don't have jbuilder 
> installed, so have never even used it ;-).
> 
> However, I did look through the code when writing my
> own hook, and AIUI, it 
> doesn't bother with a DTT but just checks for tags
> by name assuming 
> everything's well formed.  I'm reusing bits from the
> jbuilder hook in my own 
> simple swing-based front end.  When that's a little
> tidier I'll check it into 
> the sourceforge repository.
> 
> I'd imagine you'd get by, simply copying the last
> fifty or so lines of 
> JBuilderHook.java, and adapting buildMenuGroup() and
> addTransmogrifier() to 
> create jext style Actions.
> 
> Good luck,
> 
> John
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

Re: [Transmogrify-development] refactorings.xml

[Andrew M. (smileyy) <sm...@fi...>]

On Fri, 23 Nov 2001, Matt Benson wrote:

> Do these files have or need a DTD?

No, and no.  Though if you want to write one... =)

IIRC, the code for the XML parsing in in JBuilderHook.  It seems
reasonable to implement this in JAXP if it isn't already (for parser
neutrality), and push the XML parsing up into Hook.  From there, Lists of
available refactoring and twiddlings could be exposed.  I can't imagine a
transmogrify implementation not wanting to use the XML files...

andrew
-- 
sm...@fi...

Re: [Transmogrify-development] refactorings.xml

[John P N P. <joh...@zo...>]

On Friday 23 Nov 2001 5:12 pm, Matt Benson wrote:
> Do these files have or need a DTD?
>

Don't ask me - I didn't do it!

In fact not only did I not write the jbuilder hook, but I don't have jbuilder 
installed, so have never even used it ;-).

However, I did look through the code when writing my own hook, and AIUI, it 
doesn't bother with a DTT but just checks for tags by name assuming 
everything's well formed.  I'm reusing bits from the jbuilder hook in my own 
simple swing-based front end.  When that's a little tidier I'll check it into 
the sourceforge repository.

I'd imagine you'd get by, simply copying the last fifty or so lines of 
JBuilderHook.java, and adapting buildMenuGroup() and addTransmogrifier() to 
create jext style Actions.

Good luck,

John

Re: [Transmogrify-development] refactorings.xml

[Matt B. <gud...@ya...>]

Do these files have or need a DTD?

--- John P N Pybus <joh...@zo...>
wrote:
> On Friday 23 Nov 2001 2:24 pm, you (Matt Benson)
> wrote:
> > Hello, all... I am spending my Thanksgiving
> holiday
> > working on the port of Transmogrify to Jext.  I
> have
> > little experience with Ant, but I spent yesterday
> > modifying build.xml so that it would build and
> include
> > my ...hook.jext package.  Now I am working on
> making
> > my code as clean as possible.  Jext uses a sort of
> > Action object to carry out its tasks, and rather
> than
> > write one for each refactoring and have a
> maintenance
> > job on my hands, I want to dynamically create
> Actions
> > that correspond to refactorings (probably from
> xml).
> > I have noticed the files refactorings.xml and
> > twiddles.xml that were in the root of the cvs
> tree.
> > It looks as if it would be little trouble for me
> to
> > use these, but I wanted to know for what purpose
> they
> > were actually intended...
> 
> Take a look at .../hook/jbuilder/JBuilderHook.java
> from about line 388.  The 
> jbuilder hook builds its menu items from those XML
> files.  Sounds ideal for 
> your purpose.
> 
> Best wishes,
> 
> John
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

Re: [Transmogrify-development] refactorings.xml

[John P N P. <joh...@zo...>]

On Friday 23 Nov 2001 2:24 pm, you (Matt Benson) wrote:
> Hello, all... I am spending my Thanksgiving holiday
> working on the port of Transmogrify to Jext.  I have
> little experience with Ant, but I spent yesterday
> modifying build.xml so that it would build and include
> my ...hook.jext package.  Now I am working on making
> my code as clean as possible.  Jext uses a sort of
> Action object to carry out its tasks, and rather than
> write one for each refactoring and have a maintenance
> job on my hands, I want to dynamically create Actions
> that correspond to refactorings (probably from xml).
> I have noticed the files refactorings.xml and
> twiddles.xml that were in the root of the cvs tree.
> It looks as if it would be little trouble for me to
> use these, but I wanted to know for what purpose they
> were actually intended...

Take a look at .../hook/jbuilder/JBuilderHook.java from about line 388.  The 
jbuilder hook builds its menu items from those XML files.  Sounds ideal for 
your purpose.

Best wishes,

John

[Transmogrify-development] (no subject)

[Matt B. <gud...@ya...>]

Hello, all... I am spending my Thanksgiving holiday
working on the port of Transmogrify to Jext.  I have
little experience with Ant, but I spent yesterday
modifying build.xml so that it would build and include
my ...hook.jext package.  Now I am working on making
my code as clean as possible.  Jext uses a sort of
Action object to carry out its tasks, and rather than
write one for each refactoring and have a maintenance
job on my hands, I want to dynamically create Actions
that correspond to refactorings (probably from xml). 
I have noticed the files refactorings.xml and
twiddles.xml that were in the root of the cvs tree. 
It looks as if it would be little trouble for me to
use these, but I wanted to know for what purpose they
were actually intended...

-Matt

__________________________________________________
Do You Yahoo!?
Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

RE: [Transmogrify-development] java grammer 1.3 (printing comment s)

[Nascif Abousalh-N. <na...@no...>]

What about a new Transmogrify milestone to comemorate the new grammar? :-)

> -----Original Message-----
> From: Matt Benson [mailto:gud...@ya...]
> Sent: Tuesday, November 20, 2001 1:31 PM
> To: tra...@li...
> Subject: Re: [Transmogrify-development] java grammer 1.3 (printing
> comments)
> 
> 
> I haven't worked on the Jext port of Transmogrify in
> quite some time; I first had to put together a working
> prototype of Jext's Project API which is being built
> for the purpose of better integrating Transmogrify (we
> hope to use its analytics for all Java parsing tasks).
>  I am about ready to resume work on the port but had
> not heard that comments had been fixed, so assumed
> otherwise.  Again those were a must so I am glad to
> hear that they are no longer a problem.  I will advise
> the list of any difficulties I encounter whenever I
> get back to working on this (hence my comment about
> having a life being a problem for open-source
> developers :)  ).
> 
> --- John P N Pybus <joh...@zo...>
> wrote:
> > On Tuesday 20 Nov 2001 4:39 pm, you (Matt Benson)
> > wrote:
> > > How about those comments, then?? ;) 
> > 
> > Hi there Matt,
> > 
> > Yes, I have had a look at the comments issue, all
> > the machinery is already 
> > there in Transmogrify, comments are stored with the
> > AST, and (apart from a 
> > few minor issues such as the final comment in a file
> > being missed out) the 
> > printer package should be able to write them back
> > out.
> > 
> > Where exactly do you have problems? Are you using
> > JBuilder?, I'm not using 
> > transmogrify with JBuilder myself.  I've attached a
> > test program which parses 
> > a file, and prints it straight back out (it'll need
> > transmogrify and antlr in 
> > its classpath).  Does that work for you?  It works
> > okay for me, comments and 
> > all.
> > 
> > >       Doesn't having a
> > > life suck?
> > 
> > Aah, I use transmogrify as a displacement activity
> > when I'm bored of work.
> > More subtle than a game of Quake -- unless you get
> > very close, one java dev 
> > session is indistinguishable from another ;-).
> > 
> > John
> > > import net.sourceforge.transmogrify.symtab.*;
> > import
> > net.sourceforge.transmogrify.symtab.printer.*;
> > import net.sourceforge.transmogrify.symtab.parser.*;
> > 
> > import antlr.collections.*;
> > 
> > import java.io.*;
> > 
> > public class PrintTest {
> >     /**
> >      * run with two arguments <filein.java>
> > <fileout.java>
> >      *
> >      * If all is well they should be the same modulo
> > whitespace.
> >      */
> >     public static void main(String[] s)
> >         throws Exception
> >     {
> >         FileParser in = new FileParser();
> >         in.doFile(new File(s[0]));
> >         Printer printer =
> > PrinterFactory.makePrinter(in.getTree());
> >         printer.print(new PrettyPrinter(new
> > FileWriter(s[1])));
> >     }
> > }
> 
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
> http://geocities.yahoo.com/ps/info1
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
> https://lists.sourceforge.net/lists/listinfo/transmogrify-development
> 

Re: [Transmogrify-development] java grammer 1.3 (printing comments)

[Matt B. <gud...@ya...>]

I haven't worked on the Jext port of Transmogrify in
quite some time; I first had to put together a working
prototype of Jext's Project API which is being built
for the purpose of better integrating Transmogrify (we
hope to use its analytics for all Java parsing tasks).
 I am about ready to resume work on the port but had
not heard that comments had been fixed, so assumed
otherwise.  Again those were a must so I am glad to
hear that they are no longer a problem.  I will advise
the list of any difficulties I encounter whenever I
get back to working on this (hence my comment about
having a life being a problem for open-source
developers :)  ).

--- John P N Pybus <joh...@zo...>
wrote:
> On Tuesday 20 Nov 2001 4:39 pm, you (Matt Benson)
> wrote:
> > How about those comments, then?? ;) 
> 
> Hi there Matt,
> 
> Yes, I have had a look at the comments issue, all
> the machinery is already 
> there in Transmogrify, comments are stored with the
> AST, and (apart from a 
> few minor issues such as the final comment in a file
> being missed out) the 
> printer package should be able to write them back
> out.
> 
> Where exactly do you have problems? Are you using
> JBuilder?, I'm not using 
> transmogrify with JBuilder myself.  I've attached a
> test program which parses 
> a file, and prints it straight back out (it'll need
> transmogrify and antlr in 
> its classpath).  Does that work for you?  It works
> okay for me, comments and 
> all.
> 
> >       Doesn't having a
> > life suck?
> 
> Aah, I use transmogrify as a displacement activity
> when I'm bored of work.
> More subtle than a game of Quake -- unless you get
> very close, one java dev 
> session is indistinguishable from another ;-).
> 
> John
> > import net.sourceforge.transmogrify.symtab.*;
> import
> net.sourceforge.transmogrify.symtab.printer.*;
> import net.sourceforge.transmogrify.symtab.parser.*;
> 
> import antlr.collections.*;
> 
> import java.io.*;
> 
> public class PrintTest {
>     /**
>      * run with two arguments <filein.java>
> <fileout.java>
>      *
>      * If all is well they should be the same modulo
> whitespace.
>      */
>     public static void main(String[] s)
>         throws Exception
>     {
>         FileParser in = new FileParser();
>         in.doFile(new File(s[0]));
>         Printer printer =
> PrinterFactory.makePrinter(in.getTree());
>         printer.print(new PrettyPrinter(new
> FileWriter(s[1])));
>     }
> }


__________________________________________________
Do You Yahoo!?
Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

Re: [Transmogrify-development] java grammer 1.3 (printing comments)

[John P N P. <joh...@zo...>]

On Tuesday 20 Nov 2001 4:39 pm, you (Matt Benson) wrote:
> How about those comments, then?? ;) 

Hi there Matt,

Yes, I have had a look at the comments issue, all the machinery is already 
there in Transmogrify, comments are stored with the AST, and (apart from a 
few minor issues such as the final comment in a file being missed out) the 
printer package should be able to write them back out.

Where exactly do you have problems? Are you using JBuilder?, I'm not using 
transmogrify with JBuilder myself.  I've attached a test program which parses 
a file, and prints it straight back out (it'll need transmogrify and antlr in 
its classpath).  Does that work for you?  It works okay for me, comments and 
all.

>       Doesn't having a
> life suck?

Aah, I use transmogrify as a displacement activity when I'm bored of work.
More subtle than a game of Quake -- unless you get very close, one java dev 
session is indistinguishable from another ;-).

John

Re: [Transmogrify-development] java grammer 1.3

[Matt B. <gud...@ya...>]

How about those comments, then?? ;)  Doesn't having a
life suck?


--- John P N Pybus <joh...@zo...>
wrote:
> I've updated the symtab packages in transmogrify to
> use the java1.3 version 
> of java.g from the antlr website.  I've checked the
> changes into the 
> sourceforge CVS.
> 
> There are only a few changes in the grammer: the
> addition of strictfp; 
> improvements in the float/long handling; and the
> explicit recognition of 
> (super)constructor calls at the beginning of a
> constructor definition.
> 
> I've left the indenting in antlr style, rather than
> that used by the rest of 
> transmogrify.  It'll make it easier to integrate
> future changes to the 
> grammer, and it is ANTLR code not java.
> 
> It passes all the tests, but it would be good if
> people could check that it 
> still runs over their sources.
> 
> Cheers,
> 
> John
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

[Transmogrify-development] java grammer 1.3

[John P N P. <joh...@zo...>]

I've updated the symtab packages in transmogrify to use the java1.3 version 
of java.g from the antlr website.  I've checked the changes into the 
sourceforge CVS.

There are only a few changes in the grammer: the addition of strictfp; 
improvements in the float/long handling; and the explicit recognition of 
(super)constructor calls at the beginning of a constructor definition.

I've left the indenting in antlr style, rather than that used by the rest of 
transmogrify.  It'll make it easier to integrate future changes to the 
grammer, and it is ANTLR code not java.

It passes all the tests, but it would be good if people could check that it 
still runs over their sources.

Cheers,

John

Re: [Transmogrify-development] JParse

[John P N P. <joh...@zo...>]

On Sunday 28 Oct 2001 3:23 pm, Matt Benson wrote:
> James:
>   I am writing on behalf of an open-source refactoring
> project called Transmogrify.  This is developed by
> some people with Thoughtworks, Inc. and is currently
> available in plugin form for the JBuilder and NetBeans
> IDEs with others in progress.  It, like your JParse,
> uses Antlr but has not yet developed its own grammar,
> which means, as I'm sure you know, that comments and
> whitespace are thrown away when the source is
> regenerated from the tree.

In fact this problem with transmogrify is not with the java parser, but with 
the printer classes which turn the abstract syntax tree back into code.  
Antlr makes it very easy to keep the whitespace/comments in the tree as 
hidden tokens -- see <http://www.jguru.com/faq/view.jsp?EID=203533> and note 
that symtab.parser.SymTabToken and symtab.parser.SymTabAST do extend 
antlr.CommonHiddenStreamToken and antlr.CommonASTWithHiddenTokens 
respectively.

When I have the java 1.3 grammer working (see below) I'll look to fix the 
printer classes to make use of the hidden tokens to reinstate the comments as 
they print out.

> My question to you is,
> what modifications were made from Antlr's java.g to
> your own?  I have tried doing a simple diff but it
> seems quite extensive, is this so?  The reason I am
> asking is that since the last release of JParse, a 1.3
> Java grammar has been released on Antlr's site so I
> would like to see about integrating your changes to
> the 1.2 grammar into the 1.3 grammar.

Apart from reformatting of white space, the java grammer used in transmogrify 
is very little changed from that in the antlr 2.7.1 distribution.  I have 
applied those few changes to the 1.3 grammer, and just have to adapt the rest 
of the parser and symbol table package to the changes in java syntax between 
the 1.2 and 1.3 parsers (it currently fails about 10 unit tests).

It probably makes sense to track the antlr version of java.g as closely as 
possible to ease the amount of work needed to keep up with new java releases. 
 JDK 1.4 will bring language changes with assert, and beyond that it looks 
like there'll be fun when generics make it into 1.5 ...

I'll check this stuff into the CVS as I find time to get it working.

Yours,

John

[Transmogrify-development] Re: JParse

[<ja...@ee...>]

Matt Benson <gud...@ya...>, on Sun, 28 Oct 2001 at 07:23:36
-0800 (PST) you wrote:
>   I am writing on behalf of an open-source refactoring
> project called Transmogrify.  This is developed by
> some people with Thoughtworks, Inc. and is currently
> available in plugin form for the JBuilder and NetBeans
> IDEs with others in progress.  It, like your JParse,
> uses Antlr but has not yet developed its own grammar,
> which means, as I'm sure you know, that comments and
> whitespace are thrown away when the source is
> regenerated from the tree.  My question to you is,
> what modifications were made from Antlr's java.g to
> your own?  I have tried doing a simple diff but it
> seems quite extensive, is this so?

Yes.  I used ANTLR's java.g as a starting point, but I added quite a lot
of machinery to do the things I wanted.  If you strip out all of the
actions, what remains is pretty close to (but not quite identical with)
ANTLR's java.g.  I made it use heterogeneous AST nodes, so that I could
store semantic information in a convenient way, for one thing.  Type
resolution is another addition.

I should also point out one thing I did NOT add: error handling.  JParse
will choke and die on bad input.  I've had such good intentions, but
have never gotten around to it.

>                                     The reason I am
> asking is that since the last release of JParse, a 1.3
> Java grammar has been released on Antlr's site so I
> would like to see about integrating your changes to
> the 1.2 grammar into the 1.3 grammar.

Oh, I see!  Well, I was synched up with version 1.17 of java.g, so all
that has to be worried about are the new things in version 1.18.  The
file header claims that those are:

  o  primary didn't include boolean.class kind of thing

I already fixed that in JParse.

  o  constructor calls parsed explicitly now: see
     explicitConstructorInvocation

I did this in JParse, too, but in a different way.

  o  add strictfp modifier

This is already in JParse.

  o  missing objBlock after new expression in tree grammar

This is already in JParse.

  o  merged local class definition alternatives, moved after declaration

This one would have to be looked into.

  o  fixed problem with ClassName.super.field

If this is the problem I think it is, then I already fixed it as well.
However, I would want to look more closely to be sure.

  o  reordered some alternatives to make things more efficient

This one would have to be looked into as well.

  o  long and double constants were not differentiated from int/float

Ditto.

  o  whitespace rule was inefficient: matched only one char

Ditto.

  o  add an examples directory with some nasty 1.3 cases

Ditto.

  o  made Main.java use buffered IO and a Reader for Unicode support

I did something like this, but would want to compare approaches.

  o  supports UNICODE?

This would have to be looked into.

The way this should be handled is to do a diff between versions 1.17 and
1.18 of the ANTLR Java grammar.  Inspection of the changes there should
reveal the ways in which JParse should be changed to conform.

>                                        Also, I would
> like to ask your opinion about what ramifications
> there might be to using a derivative of your grammar
> with processing that may add nodes to and remove nodes
> from the tree?  I have copied the Transmogrify
> development list on this message; for their benefit I
> will also include the address of JParse's homepage: 
> http://www.ittc.ku.edu/JParse

I did precisely that with a derivative of JParse that I used for the
source-to-source compiler for the Kan project:
<URL:http://www.ittc.ku.edu/kan/>.  It can be done, with care.

JParse has been released under the LGPL, so you are free to modify it as
you wish, so long as it is distributed in conformance with the LGPL.
However, I would appreciate you keeping me informed of what you do, at
the least, if not involving me in it.

Thanks,
-- 
Jerry James, Assistant Professor
EECS Department
University of Kansas

[Transmogrify-development] JParse

[Matt B. <gud...@ya...>]

James:
  I am writing on behalf of an open-source refactoring
project called Transmogrify.  This is developed by
some people with Thoughtworks, Inc. and is currently
available in plugin form for the JBuilder and NetBeans
IDEs with others in progress.  It, like your JParse,
uses Antlr but has not yet developed its own grammar,
which means, as I'm sure you know, that comments and
whitespace are thrown away when the source is
regenerated from the tree.  My question to you is,
what modifications were made from Antlr's java.g to
your own?  I have tried doing a simple diff but it
seems quite extensive, is this so?  The reason I am
asking is that since the last release of JParse, a 1.3
Java grammar has been released on Antlr's site so I
would like to see about integrating your changes to
the 1.2 grammar into the 1.3 grammar.  Also, I would
like to ask your opinion about what ramifications
there might be to using a derivative of your grammar
with processing that may add nodes to and remove nodes
from the tree?  I have copied the Transmogrify
development list on this message; for their benefit I
will also include the address of JParse's homepage: 
http://www.ittc.ku.edu/JParse

Thanks,
Matt Benson

__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com

Re: [Transmogrify-development] strictfp

[Matt B. <gud...@ya...>]

Antlr's 1.3 Java grammar appears to handle strictfp.

--- John P N Pybus <joh...@zo...>
wrote:
> 
> I've been running ReferenceCounter on some large
> source trees to see how much 
> memory it uses.  The javax.* tree in the JDK1.3.1
> src.jar (about 9Mb of 
> source) ran, but used a little over 200Mb of RAM (I
> had to increase the JVM 
> heap size to 200Mb).
> 
> This is what it said:
> 
> References found:      226011
> References resolved:   225587
> References unresolved: 424
>  
> real    9m49.294s
> user    4m4.140s
> sys     0m14.070s
> 
> I then tried it on the java.* tree but it crashed
> with a complaint about an 
> unrecognized token strictfp:
> 
>
net.sourceforge.transmogrify.symtab.parser.ParsingException:
>     unexpected token: strictfp
>       at
>
net.sourceforge.transmogrify.symtab.parser.FileParser.parseFile(FileParser.java:126)
> <...>
> 
> It looks like the java parser is a bit out of date,
> and will need updateing 
> to include the addition of strictfp to the spec.
> 
> John
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
>
https://lists.sourceforge.net/lists/listinfo/transmogrify-development


__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com

RE: [Transmogrify-development] Starting Unit Tests

[Nascif Abousalh-N. <na...@no...>]

I understand. Hopefully I'll have some more time soon to remove the
dependency on gnuserv, which should make it easier for you to get it to
work.
As for the bsh/JDE dependency, I think I can't get it to work without them -
unless.... unless we make a "refactoring server" that implements Hook and
provides a socket-based interface. It could talk directly to Emacs without
using Java(or the Beanshell) at all. Then again, it is yet another JVM
running in your machine, and another editor-to-library API to define... 

Regards,
	Nascif


> -----Original Message-----
> From: John P N Pybus [mailto:joh...@zo...]
> Sent: Friday, October 12, 2001 12:30 PM
> To: tra...@li...
> Subject: [Transmogrify-development] Starting Unit Tests
> 
> 
> On Wednesday 12 Sep 2001 6:23 pm, Andrew McCormick (smileyy) wrote:
> > On Wed, 12 Sep 2001, John P N Pybus wrote:
> > > [snip]                                                    
>              Do
> > > you know of a way to make the junit test's GUI close 
> automatically if all
> > > the tests complete successfully?
> >
> > The best thing I can think of is to use the JUnit's text UI 
> instead of the
> > Swing UI.  In the particular test suite that you're running, change:
> >
> >   junit.swingui.TestRunner.run( ... );
> >
> > to
> >
> >   junit.textui.TestRunner.run( ... );
> >
> > I can't immediately think of a way to make this a project-wide
> > configurable option, though.
> 
> I've added two new classes to net/sourceforge/transmogrify/test/.   
> TestStarter uses a system property to decide which TestRunner 
> to run.  I've 
> edited every test in the tree which had a main() to use 
> TestStarter.startRun(someTestClass.class), rather than 
> junit.swingui.TestRunner.run(someTestClass.class).  I've 
> tested them all[1] 
> by running every test possible standalone, as well as as part 
> of their 
> respective TestSuite.
> 
> I've also added a class ExitingTestRunner, which subclasses 
> junit.swingui.TestRunner to provide a version which exits at 
> the end of a 
> successful run.  Now I can get my way - a pretty green/red 
> bar, and ant 
> completing its build without blocking if the bar stays green.
> 
> Unless you have any objections, I'll check in these changes.
> 
> > > I don't know about an IDE (I tend to use Xemacs for all 
> my developments),
> > > but a standalone app to browse source and apply 
> refactorings, would be a
> > > great idea.
> >
> > That would the general idea -- something that integrates well with
> > "command-line" development tools.
> 
> Since I've so far failed in my various attempts to get 
> Transmogrify to work 
> wth Xemacs.  I don't have a copy of JBuilder, and haven't had 
> time to install 
> plain Emacs with all the required libs (sorry Nascif), I've 
> started putting 
> together a simple swing based file viewer with Hook to apply 
> refactorings.  
> So far I've only got a couple of the refactorings hard-wired 
> into it's menu.  
> Once I've added a bit more code to build the menus from the 
> .xml files as the 
> jbuilder hook does, I'll add this to the project so you can 
> have a look.
> 
> Yours,
> 
> John Pybus
> 
> [1] In the process I came across NameAmbiguityTest, which I 
> notice doesn't 
> pass, tut tut. ;-)
> 
> _______________________________________________
> Transmogrify-development mailing list
> Tra...@li...
> https://lists.sourceforge.net/lists/listinfo/transmogrify-development
>