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

[Andreas Stefik]

Ter and all,

>>> Refactoring
>>
>> No complaints on this system. It does the most common things.
>
> But do youuse it very often?


Not very much in AW. I think my most common use case is in languages
where I have quite a few files with a number of dependencies. Probably
most often, I rename methods in java. However, the grammar for my
language realistically doesn't change all that terribly much nowadays,
so refactoring isn't used much. So yaa, not a huge priority for me
specifically.

>> 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
>
>  can you tell me what they pinned window is exactly?

Many IDEs have them. I could probably describe it badly in English,
but if you pull up NetBeans and look at the "Projects" tab, you'll
see. In short, the windows auto-hide, can be moved around,
restructured, made big/small, and are generally very flexible.


>> My most common
>> use case is switching back and forth between a tree walker and a
>> parser.
>
>  true, but isn't a nicer to see two Windows next to each other to compare?

This is probably in the realm of personal preference. In my case, I
use multiple monitors and usually have a web browser, a number of
terminals, NetBeans (or other) or other stuff open when I'm working on
the grammar anyway. So, for me, it's just more clutter having multiple
windows to keep track of in AW. But, others might have only AW open
when they work or really just have a different standard way of
working. If we took a poll, I'm not sure what the result would be.


On the last point from another email, with porting to the NetBeans
platform, I can certainly understand your hesitation, considering
there are multiple IDE's out there (e.g., at least eclipse), where
things could be ported, and the learning curve on any of the platforms
can be pretty steep. As I'm sure you know, with something like
NetBeans, you would get 90% of the way there just by starting with:
http://kenai.com/projects/nbantlr, which I think is posted on your
site, but yaa, I definitely sympathize not wanting to spend time
learning complicated APIs.

There might be some other options. I haven't looked into what
technique you use to draw the syntax diagrams, run the interpreter, or
whatever, but if those were made available as public
APIs/implementation in AW as jar files you could load up, others could
really easily integrate them into NetBeans, as you can just embed
swing components/hook into the debugger API, etc. That could,
potentially, give a boost to the kenai project.

What we do on our projects is to create a set of API jar files and
split them apart as much as possible. Then, __any__ interested person
can simply download the jars and hook them into any IDE that supports
that, which probably includes at least NetBeans and Eclipse. So, for
example, our tools have a text-to-speech engine. You can get the
separately, but we also (trivially) have it hooked in through
NetBeans, which makes the API available. There's a wizard for hooking
it in even, so it's pretty easy.

Anyway, just a thought for version 3,4,5, or whenever. Like others
have said, the editor issue is still #1 most important on my end.

Stefik