ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Donald <dona...@apache.org>
Subject Ant2 and properties
Date Sat, 02 Dec 2000 03:32:41 GMT
Hi,

Just a few questions about properties and ant2. From what I understand
properties will encompass any named data that is available to different
tasks in system. So for instance I envision the folowing being as properties

1. text properties
2. file properties (ie indicate a file object)
3. fileset/patternset/filterset/whateverset objects
4. evaluations of 3 (ie replacing of embedded variables and applying if at
a aprticular point in time)
5. Possibly other "temp" data - ie task 1 generates X, while task to grabs
X from property and uses it.
etc

So somethings (like patternsets) can not be directly used but have to be
evaluated in a context (aka Scope) for a task to find it useful. The
fileset will be evaluated to create an array of files which can then in
turn be bound to a property.

I see the following as the scopes
1. Workspace
2. Project
3. Target
4. Task

Scope 4 is completely seperate from all others so can effectively be
ignored. Scope 1 is the toplevel scope. For instance you could place
projects; tomcat, ant, turbine, velocity and X in one workspace because
they are all needed together to build and deploy X. 

Now given that we want to write a standard set of property names so that we
can set "dist.base" in workspace and have all the different projects build
distributions to that particular base directory. We could also set up a set
of standard directory structures so that Tomcat that relied on Ant could
use the ant jars from "${dist.base}/ant/lib/ant.jar" etc.

I don't think anyone has decided how property setting will work (ie are
overwrites allowed, do we always write to current scope or parent scope or
...) but other than that do I grok the current ideas ?


Cheers,

Pete

*-----------------------------------------------------*
| "Faced with the choice between changing one's mind, |
| and proving that there is no need to do so - almost |
| everyone gets busy on the proof."                   |
|              - John Kenneth Galbraith               |
*-----------------------------------------------------*


Mime
View raw message