Although a javadoc_artifact method is a possible improvement, the kind variable issue is a bug. Isn't it?
On Sat, Aug 28, 2010 at 1:03 AM, Alex Boisvert <[email protected]>wrote: > Yeah, in that case we could support eclipse.options.m2_repo that would be a > path and output artifacts with kind="lib". > > alex > > On Fri, Aug 27, 2010 at 2:59 PM, Nikos Maris <[email protected]> wrote: > > > Problem is not solved as eclipse does not work out-of-the-box. When > > setting eclipse.options.m2_repo_var, > > kind="lib" should be added to Eclipse .classpath instead of kind="var" as > > stated here (http://eclim.org/vim/java/classpath.html) > > > > Do you think that a specific issue should be raised? > > > > On Sat, Aug 28, 2010 at 12:28 AM, Nikos Maris <[email protected]> > wrote: > > > > > Thanks for responding in 7 minutes! Which would be an elegant way to > > > configure the task to add javadocpath to Eclipse > > > .classpath "ecliple.options.javadoc = true"? > > > > > > On Sat, Aug 28, 2010 at 12:04 AM, Alex Boisvert < > [email protected] > > >wrote: > > > > > >> In your top-level project, just add: > > >> > > >> eclipse.options.m2_repo_var = Buildr.repositories.local > > >> > > >> This will "hard-code" your repository path into the Eclipse > .classpath. > > >> > > >> alex > > >> > > >> > > >> On Fri, Aug 27, 2010 at 1:56 PM, Nikos Maris <[email protected]> > > wrote: > > >> > > >> > How could I refactor the following code? As I don't want my > classpath > > >> file > > >> > to include the M2_REPO variable, I want this variable to be > unbounded > > >> after > > >> > every invocation of the eclipse task. > > >> > > > >> > desc "usage: task :eclipse do noneedfor_m2eclipse _('.classpath'); > > end" > > >> > def noneedfor_m2eclipse(path) > > >> > if File.exists? path then > > >> > unbounded = File.read(path).gsub(/M2_REPO/, > > >> > repositories.local).gsub(/kind="var"/, 'kind="lib"') > > >> > File.open(path, "w") {|file| file.puts unbounded} > > >> > end > > >> > end > > >> > > > >> > > > > > > > > >
