[antlr-interest] JavaScript Target

Joey Hurst jhurstus at gmail.com
Sat May 17 23:56:10 PDT 2008


George,
Yes, yes, and yes.  I've been keeping an eye on the ES4 target as I
developed the JS target, so there is already quite a bit of parity between
the two.  Your comments about maintenance and migration are well taken.

I'm travelling this week and won't have much free time, but when I get back
I'll send you and Terence a more detailed comparison of the current state of
the JS and ES4 targets in addition to some thoughts on how to achieve the
ideas you set out below.

Thanks,
Joey

On 5/16/08, George Scott <george.scott at gmail.com> wrote:
>
> Hi Joey,
>
> Great work!  I maintain the ActionScript 3/ECMAScript 4 target for ANTLR,
> we should sync up on the work that you've done since the targets should be
> very similar.  It would be good to align the runtime APIs and code
> generation decisions as much as possible, so there is an evolutionary path
> from ECMAScript 3 to 4 that is consistent.  If done right, much of the code
> generation templates and even some runtime code could be shared between the
> two versions, which would simplify maintenance going forward.
>
>
> George
>
>
> On Fri, May 16, 2008 at 1:39 PM, Terence Parr <parrt at cs.usfca.edu> wrote:
>
>>
>> On May 15, 2008, at 9:19 PM, Joey Hurst wrote:
>>
>> Hello all,
>>> For the last few months I've been working on a JavaScript ANTLR target.
>>>
>>
>> hooray!
>>
>> Things are coming along quite well; as of today the target passes most of
>>> the tests (ported to JsUnit) shipped with the 3.0.1 source distribution.
>>>  I've focused on the tests in the runtime/Python/tests/ directory, although
>>> there are others too.  I've maintained API compatibility with the Java
>>> target where possible, but there are a few places where JavaScript couldn't
>>> quite handle the task at hand (e.g. there is no reliable way to do runtime
>>> stack reflection in EcmaScript/JavaScript version 3).
>>>
>>> I write this message because I wanted to get a couple legal and
>>> scheduling issues cleared up before I continue any further:
>>>
>>> 1) 99% of the JavaScript target code is original (although heavily
>>> inspired, of course, by the other targets).  The remaining 1% consists of a
>>> modified version of the Yahoo class provided by the Yahoo UI Library.  This
>>> code is licensed under the BSD license like ANTLR, so I assume this won't
>>> cause any problems.  If this is an issue I could rewrite the code, but it
>>> would be a hassle.  Here's a link to the YUI license:
>>> http://developer.yahoo.com/yui/license.html
>>> Here's a link to the Yahoo class docs for anyone who is curious:
>>> http://developer.yahoo.com/yui/yahoo/
>>>
>>
>> seems ok. so you would simply be including a source code file with a
>> license a the top? Should be fine.
>>
>> 2) It appears that the 3.1 release is imminent.  My work has been based
>>> off the 3.01 release to date.  Do y'all think it would be valuable to create
>>> a finished 3.0.1 compatible version of the JavaScript target, or should I
>>> port what I have now to 3.1 and release that?
>>>
>>
>> We could include 3.0.1 JavaScript target with 3.1 as long as the templates
>> worked and the runtime was consistent.
>>
>> I'll send out the documentation and code in about another month.  It
>>> should be complete and pretty robust at that time.
>>>
>>
>> Oh okay, it seems like you will miss the 3.1 deadline; uh, like this
>> weekend ;)
>>
>> Terence: I'm a fellow SF denizen.  I'll have to buy you a beer sometime in
>>> appreciation for this great tool!
>>>
>>
>> Heh,bring on the beer, dude. ok, wine, but same idea... what part of the
>> city do you live in? If you hear gunshots, that's probably my neighborhood
>> too. [might one take this discussion to an offlist thread]
>>
>> Ter
>>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20080517/acd858ce/attachment-0001.html 


More information about the antlr-interest mailing list