* resources) are stored and retrieved. There are two primary types of repositories:
 * {@linkplain LocalRepository local repositories} and
 * {@linkplain RemoteRepository remote repositories}.</p>
 *
 * <h2>Repository Resolution Process</h2>
 *
 * <p>When resolving dependencies, Maven follows this order:</p><ol>
 * <li>Check Local Repository: Maven first checks if the artifact is available in the local repository.</li>

                    Process process = builder.start();
                    Thread out = new Thread(new StreamGobbler(process.getInputStream(), shellContext.writer));
                    Thread err = new Thread(new StreamGobbler(process.getErrorStream(), shellContext.logger::error));
                    out.start();
                    err.start();
                    int exitCode = process.waitFor();
                    out.join();

### TODO

In subsequent releases we are planning to extend this to provide things like

- Reload() of MinIO server arguments without fully restarting the process.

- Expanding 1 node at a time by automating the process of creating a new pool
  and decommissioning to provide a functionality that smaller deployments
  care about.

	// is obviously garbage (Unix time == 0), then ignore modtimes
	// and don't process the If-Modified-Since header.
	if objInfo.ModTime.IsZero() || objInfo.ModTime.Equal(time.Unix(0, 0)) {
		return false
	}

	// If top level is a delete marker proceed to upload.
	if objInfo.DeleteMarker {
		return false
	}

[19515] [INFO] Booting worker with pid: 19515
[19511] [INFO] Started server process [19511]
[19511] [INFO] Waiting for application startup.
[19511] [INFO] Application startup complete.
[19513] [INFO] Started server process [19513]
[19513] [INFO] Waiting for application startup.
[19513] [INFO] Application startup complete.
[19514] [INFO] Started server process [19514]
[19514] [INFO] Waiting for application startup.

  // Process inputs.
  std::vector<tensorflow::OutputTensor> input_tensors;
  std::unordered_map<const Node*, std::vector<int>> input_nodes;
  status->status = tensorflow::ProcessInputs(fn_body, fn_name, ninputs, inputs,
                                             &input_tensors, &input_nodes);
  if (TF_GetCode(status) != TF_OK) return nullptr;

  // Process outputs.

        // simple checks first to short circuit processing below.
        if (WILDCARD.equals(pattern) || pattern.equals(originalId)) {
            result = true;
        } else {
            // process the list
            String[] repos = pattern.split(",");
            for (String repo : repos) {
                repo = repo.trim();
                // see if this is a negative match

# Extending OpenAPI { #extending-openapi }

There are some cases where you might need to modify the generated OpenAPI schema.

In this section you will see how.

## The normal process { #the-normal-process }

The normal (default) process, is as follows.

A `FastAPI` application (instance) has an `.openapi()` method that is expected to return the OpenAPI schema.

// restartProcess starts a new process passing it the active fd's. It
// doesn't fork, but starts a new process using the same environment and
// arguments as when it was originally started. This allows for a newly
// deployed binary to be started. It returns the pid of the newly started
// process when successful.
func restartProcess() error {
	if runtime.GOOS == globalWindowsOSName {

          currentTask = coordinatorTask
          taskRunner.notifyAll()
        }
      }

    // Let the coordinator process its interruption.
    runTasks()
  }

  /** Ask a single task to proceed. Used by the test thread only. */
  fun runNextTask() {
    taskRunner.assertLockNotHeld()

    taskRunner.withLock {
      val contextSwitchCountBefore = contextSwitchCount