jastrachan | 31 Jan 20:52 2005
Picon

Re: Groovy and TestNG

Agreed - the easiest & most productive thing is probably just to get 
the annotation stuff in the AST (and in the new parser) actually 
spitting out 1.5 bytecode, then TestNG can work straight off Groovy 
classes without any special hack.

On 31 Jan 2005, at 19:44, Russel Winder wrote:
> On Mon, 2005-01-31 at 19:06 +0000, jastrachan@... wrote:
>> On 31 Jan 2005, at 16:35, Robert McIntosh wrote:
>>
>>> There is now a version of TestNG that works with javadoc comments on
>>> JDK1.4, but I don't know if that helps you or not.
>>
>> Cool. I wonder if it uses JAM / Annogen - if so we could maybe wire in
>> JAM/Annogen into the AST to avoid having to write a groovy javadoc.
>>
>> Or we could just wire in the bytecode generation of the annotations in
>> the New Groovy parser to enable us to work on Java 5 with TestNG...
>
> I wonder how many people actually want to use TestNG with J2SE1.4?
>
> For me TestNG is a J2SE5.0 thing that happens to have a hack retrofit
> into J2SE1.4.  Personally I believe we should not spend time worrying
> about a tool designed for J2SE5.0 retrofitted to J2SE1.4, there is
> enough to do to get Groovy to a release state already.  If we want to
> support TestNG then I think it should be by doing it properly, i.e. not
> bolting Annogen into Groovy but putting annotation processing into the
> new parser.  This is forward looking rather than backward looking.
>
> NB ignoring the J2SE1.4 mode of TestNG is nothing to do with worrying
> about Groovy working with J2SE1.4 and J2SE5.0 that is a given, it must
> be a least for the short to medium term.
>
> -- 
> Russel.
> ====================================================
> Dr Russel Winder                +44 20 7585 2200
> 41 Buckmaster Road              +44 7770 465 077
> London SW11 1EN, UK             russel@...
>

James
-------
http://radio.weblogs.com/0112098/


Gmane