[antlr-interest] ANTLRWorks 2 (for ANTLR v4)

Andreas Stefik stefika at gmail.com
Fri Sep 2 10:54:42 PDT 2011


Ter,

Just to keep adding:

>Refactoring

No complaints on this system. It does the most common things.

>displaying decision DFA (well, one person mentioned to say that sometimes it gets stuck)

Do you mean the syntax diagram side? I don't recall seeing it get
stuck. I do, however, use the syntax diagrams constantly and find
having the visualization one of the best features of using AW. Again
an auto-hide like "pin" interface might be nice. Another feature that
could be potentially helpful, although not a huge priority perhaps,
would be to full screen it. That would be nice for class
demonstrations.

>only one person mentioned showing the generated code

Since it's pretty straightforward to generate the code, it's not a
huge deal for me either way.

>group/ungroup rules, rule collapsing

No real preference. Many IDEs have this and it's useful for certain
kinds of boiler plate code (e.g., generated forms). IF there's a use
that's helpful, it "might" include collapsing sections, like
options/members/boilerplate stuff, or just semantic actions, or
something else. I don't use this a whole lot in large scale code
though in other IDEs either, though.

A few other thoughts after reading your post:

1. I definitely agree that the syntax diagram, interpreter, console,
and debugger are sort of odd as tabs. Having them all as pinned
windows is more common in an IDE and has the advantage of letting you
see more than one at a time (if you wish).

2. It might be nice if you could have (on mac) a tabbed interface for
multiple grammars instead of multiple floating windows. My most common
use case is switching back and forth between a tree walker and a
parser. Managing multiple tabs on multiple windows is somewhat more
tedious than having one window where you can click on a new file and
the syntax/interpreter, etc windows auto-respond accordingly.

There's a few more thoughts, for what it's worth,

Stefik




On Fri, Sep 2, 2011 at 12:32 PM, Andreas Stefik <stefika at gmail.com> wrote:
> On Fri, Sep 2, 2011 at 11:58 AM, Terence Parr <parrt at cs.usfca.edu> wrote:
>>
>> On Sep 1, 2011, at 1:20 PM, Andreas Stefik wrote:
>>> Debugger:
>>>
>>> I'll admit that I don't really use the debugger. I have before and I
>>> really like it, but most of the projects I do require that you
>>> integrate into the build cycle of projects in the NetBeans development
>>> environment. In practice, I've never found a way I can really run the
>>> debugger, with all of my complicated build information set together
>>> (e.g., tricky dependencies, ant scripts) in such a way that it is
>>> worth the effort.
>>
>> hi Andreas,So it's hard to use the remote debug feature whereby AW you can listen to socket events from the parser? All I do is turn on -debug, recompile and start my project. when that parser starts up, it blocks waiting for a connection from AW.
>
>
> Hmm, the honest answer is that I'm not sure, as I haven't tried it
> that way. When I debug the parser on my bigger projects, I'm usually
> touching the semantics/other phases anyway, so I usually fire up
> NetBeans, set a breakpoint in the generated code, and fire up the
> NetBeans debugger. So I don't know. What you mentioned certainly
> doesn't sound difficult, although I don't know for sure if it would
> help me in my most common use cases.
>
> Honestly, though, like I said, I use the rest of AW far more often
> than I use the debugger, so I'm not the best judge on that side.
>
> Stefik
>


More information about the antlr-interest mailing list