RE: Phing reweite

Hi Alex,

>> The new parser will parse the phile and instantiate the 
>> task objects during parsing time. Then the interal arrays 
>> filled with the objects are executed as
>> parsed considering dependencies and all the other stuff.

> What do you think about implementing XMLUtils Configuration 
> (or maybe subclassing) which would allow you to gain 
> control, instead? The reason I say that is I think we 
> probably want those capabilities elsewhere in the
> system, also.

That sounds interesting. Would be killer. I thought of the parsing stuff 
basically going like this

<root>
  <tagname attrib="value" attrib2="value2" />
</root>


Parser calls 

$data = new tagname()
$data->SetAttrib("value")
$data->SetAttrib2("value2")

The problem is to handle subtags smoothly. I've not idea how to do this. In 
phing we have a kinda fixed structure and that's not so a big problem:

<root>

 <mapper />
 <property />
 <fileset />
 <...>

 <target>
   <tasks>
 </target>

</root>

>> phing. Well, it works but it's more or less a quick hack 
>> to see if a php make can replace a gnu make.

> heh, andi's idea of a quick hack. not.

:D

Andi