* into TaskSegments. This class represents the execution context of one such task segment.
 * </p>
 * <p>
 * Wise voices have suggested that maybe aggregators shouldn't be bound to the ordinary
 * lifecycle at all, in which case we wouldn't be needing this class at all ( and
 * ProjectBuildList.getByTaskSegments). Or maybe they should be introduced in the calculation
 * of the execution plan instead, which seems much nicer.
 * </p>
 * <p>

        // TODO not sure that it should assume this, maybe the calls to addMetadata should pre-merge, then artifact
        // replaces?
        AbstractRepositoryMetadata repoMetadata = (AbstractRepositoryMetadata) metadata;
        this.metadata.merge(repoMetadata.getMetadata());
    }

    public void merge(ArtifactMetadata metadata) {
        // TODO not sure that it should assume this, maybe the calls to addMetadata should pre-merge, then artifact

                    LifecyclePhaseNotFoundException, LifecycleNotFoundException, PluginVersionResolutionException {
        // Maybe do something ?
    }

    public MavenExecutionPlan calculateExecutionPlan(
            MavenSession session, MavenProject project, List<Task> tasks, boolean setup)

  project
- nice little sample projects that could be run in the core as well as integration tests

All Possible Errors
- invalid lifecycle phase (maybe same as bad CLI param, though you were talking about embedder too)
- <module> specified is not found
- malformed settings
- malformed POM
- local repository not writable
- remote repositories not available

 * </p>
 * <strong>Note:</strong> This interface is part of work in progress and can be changed or removed without notice.
 */
public interface Builder {
    //
    // Be nice to whittle this down to Session, maybe add task segments to the session. The session really is
    // the place to store reactor related information.
    //
    void build(
            MavenSession session,
            ReactorContext reactorContext,

                                }

                                if (artifact.getVersion() == null) {
                                    // set the recommended version
                                    // TODO maybe its better to just pass the range through to retrieval and use a
                                    // transformation?
                                    ArtifactVersion version;

                session1, ProjectDependencyGraphStub.A, taskSegment2.getTasks());
        assertEquals(3, executionPlan2.size());
    }

    // Maybe also make one with LifeCycleTasks

    public static LifecycleExecutionPlanCalculator createExecutionPlaceCalculator(
            MojoDescriptorCreator mojoDescriptorCreator) throws ComponentLookupException {

     * instance construction.
     *
     * @param session the Maven session
     * @throws MavenExecutionException in case of issue
     */
    // TODO This is too early for build extensions, so maybe just remove it?
    public void afterSessionStart(MavenSession session) throws MavenExecutionException {
        // do nothing
    }

    /**
     * Invoked after all projects were built.
     *

    private final List<ProjectSegment> items;

    public ProjectBuildList(List<ProjectSegment> items) {
        this.items = Collections.unmodifiableList(items);
    }

    // TODO Optimize; or maybe just rewrite the whole way aggregating mojos are being run.
    /**
     * Returns aProjectBuildList that contains only items for the specified taskSegment
     * @param taskSegment the requested task segment

            <multiplicity>*</multiplicity>
          </association>
        </field>
      </fields>
      <comment>We could probably have a specific element for a dev mailing list for things like CI,
        and maybe even a specific element for the user and scm mailing lists. Then leave the more
        lose structure for any other type of mailing list.</comment>
      <codeSegments>
        <codeSegment>