debugging information with `--config=dbg`, as issues occur on Linux if
there is too much debug info (see [this GitHub
issue](https://github.com/tensorflow/tensorflow/issues/48919) for context). If
you want to debug a kernel, you can compile specific files with `-g` using the
`--per_file_copt` bazel option. For example, if you want to debug the Identity
op, which are in files starting with `identity_op`, you can run

```bash

      - name: Notify Translations
        run: uv run ./scripts/notify_translations.py
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          NUMBER: ${{ github.event.inputs.number || null }}

                } else {
                    message += " from local repository";
                }
                getLogger().debug(message);
            } else {
                // Locally installed file is newer, don't use the resolved version
                getLogger().debug(artifact.getArtifactId() + ": using locally installed snapshot");
            }
        }
        return version;
    }

class Settings(BaseSettings):
    github_repository: str
    token: SecretStr
    debug: bool | None = False
    config: dict[str, LabelSettings] | Literal[""] = default_config


settings = Settings()
if settings.debug:
    logging.basicConfig(level=logging.DEBUG)
else:
    logging.basicConfig(level=logging.INFO)
logging.debug(f"Using config: {settings.model_dump_json()}")
g = Github(settings.token.get_secret_value())

development.here = false

# The title of environment (e.g. local or integration or production)
environment.title = Production

# Does it enable the Framework internal debug? (true only when emergency)
framework.debug = false

# one day: 86400000, three days: 259200000, five days: 432000000, one week: 604800000, one year: 31556926000
# special script :: absolute mode: $(2014/07/10), relative mode: addDay(3).addMonth(4)

log4j.appender.STDOUT.ImmediateFlush=true
log4j.appender.STDOUT.layout=org.apache.log4j.PatternLayout
log4j.appender.STDOUT.layout.ConversionPattern=%-5p %d [%t] %m%n

 * system is initialized.
 *
 * <p>This package includes:</p>
 * <ul>
 *   <li>Basic logging interfaces and implementations</li>
 *   <li>Support for different logging levels (DEBUG, INFO, WARN, ERROR)</li>
 *   <li>Accumulating loggers for deferred output</li>
 * </ul>
 *
 * @since 4.0.0
 */

 * implementation.
 *
 * @since 3.1.0
 */
public interface Slf4jConfiguration {
    /**
     * Level
     */
    enum Level {
        DEBUG,
        INFO,
        ERROR
    }

    /**
     * Set root logging level.
     *
     * @param level the level
     */
    void setRootLoggerLevel(Level level);

    /**

---

例如，在 Visual Studio Code 中，你可以：

* 前往 "Debug" 面板
* 點選 "Add configuration..."
* 選擇 "Python"
* 使用選項 "`Python: Current File (Integrated Terminal)`" 啟動偵錯器

接著它會用你的 **FastAPI** 程式碼啟動伺服器、在你的中斷點停下等。

可能會長這樣：

<img src="/img/tutorial/debugging/image01.png">

---

如果你使用 PyCharm，你可以：

* 開啟 "Run" 選單
* 選擇 "Debug..."
* 會出現一個情境選單
* 選擇要偵錯的檔案（此例為 `main.py`）

	"${WORK_DIR}/mc" ready minio/

	if ! ps -p ${pid} 1>&2 >/dev/null; then
		echo "server1 log:"
		cat "${WORK_DIR}/server1.log"
		echo "FAILED"
		purge "$WORK_DIR"
		exit 1
	fi

	if ! ./s3-check-md5 \
		-debug \
		-versions \
		-access-key minio \
		-secret-key minio123 \
		-endpoint "http://127.0.0.1:${start_port}/" 2>&1 | grep INTACT; then
		echo "server1 log:"
		cat "${WORK_DIR}/server1.log"
		echo "FAILED"