The enhancement request for the publication attribute is the following
http://issues.apache.org/jira/browse/IVY-395 Thanks, Paolo On 1/17/07, Xavier Hanin <[EMAIL PROTECTED]> wrote:
On 1/17/07, Ross Clewley <[EMAIL PROTECTED]> wrote: > > Hi > > Is there a way to get hold of the published artifacts for a module that > you are about to build? I think the answer is probably no but it could > be made possible with an enhancement to the /info/ task. > > Let me explain why I'm interested in being able to do this. We have a > modular build system, using Ivy of course, which is not entirely > dissimilar to the 'multiple projects environment' tutorial. Like that > example, we have a common build file that builds each project. The base > variant of this common build file just builds the contents of the > project into a jar. However, unlike in the tuturial, we can't always > guarantee that the jar name is the same as the ant project name. Given > that I'd like to be able to define the name of the jar in just one > place, it'd be useful if, in my common modular build file, I could get > hold of the name (or names) and types etc. of the artifacts declared to > be produced in the ivy file. > > At the moment we do it the other way around and set the jar name in a > property in a property file that's local to each project and then use > that property definition in the Ivy file. However, I'd prefer to keep > the data describing what each project produces all together in the Ivy > file. > > It seems like an extension to the /info/ task could set properties for > the declared artifacts in the Ant project. Since there are variable > numbers of artifacts it would presumably want to do this using patterns > for the property names and values just like the /artifactproperty/ task > does for resolved artifacts. It makes sense. There is also another user asking for the info task to be able to read the publication attribute, and I think reading the branch attribute could be fine too. Not too difficult to implement, maybe a jira issue for all of this would be fine. Xavier thanks > > Ross >
