Really? I guess that could be done for the jars, but not for the ivy.xml files. Indeed, the ivy file must be read in order to get the extra fields values. So the pattern of your ivy file can not be based on those values.
Am I wrong? Gilles 2007/6/25, Xavier Hanin <[EMAIL PROTECTED]>:
On 6/22/07, Scott Goldstein <[EMAIL PROTECTED]> wrote: > > I have a use case in which a product contains a number of submodules which > can be used individually. So, in essence, the submodules are identified by: > > organization/product/revision/submodule > > It doesn't appear that Ivy has such a notion in its depedency resolution > mechanism. Am I correct? > > I could replace one of these tokens with the "branch" element, but I > don't think that this was the intended purpose. > > I suppose another approach, here, would be to write a custom > DependencyResolver in which for the module, I would specify something like > "product:submodule" and my resolver would split these up and apply them to a > pattern. > > Are there any other approaches that would make more sense? Yes, I think you should better use extra attributes in this case, they are meant for that and very easy to use. All you have to do is declare a submodule attribute in your info element of your ivy files, and in your dependency elements. Then you can use this submodule as a token in your patterns. HTH, Xavier Thanks. > > Scott > -- Xavier Hanin - Independent Java Consultant Manage your dependencies with Ivy! http://incubator.apache.org/ivy/
