* profiles from the project's POM and all its parent POMs as well as from external sources like the
     * {@code settings.xml}. The profile identifiers are grouped by the identifier of their source, e.g.
     * {@code <groupId>:<artifactId>:<version>} for a POM profile or {@code external} for profiles from the
     * {@code settings.xml}.
     *

                        .map(Object::toString)
                        .toList();
                if (!Objects.equals(oldRepos, newRepos)) {
                    logger.debug("Merging repositories from " + model.getId() + "\n"
                            + newRepos.stream().map(s -> "    " + s).collect(Collectors.joining("\n")));
                }
            }

            // phase 1: get file Models from the reactor.
            List<InterimResult> interimResults = build(projectIndex, pomFiles, new LinkedHashSet<>(), true, recursive);

            ClassLoader oldContextClassLoader = Thread.currentThread().getContextClassLoader();

            try {
                // Phase 2: get effective models from the reactor

    // them into a resolver, create the expression to extract the data to validate the Model, and the URI
    // to validate the properties. We also need a way to navigate from the Tex specification documents to
    // the test in question and vice versa. A little Eclipse plugin would do the trick.
    public void testThatExecutionsWithoutIdsAreMergedAndTheChildWins() throws Exception {