logger

Loguru-based project-specific logger. Consolidates multiprocessing/multithreading logs into a single log file.

Customize its behaviors with these environment variables:

• logging file path: <PACKAGE_NAME>_LOG_PATH=”/tmp/logs/your_log.log”

• logging level (for stderr only, default=INFO): <PACKAGE_NAME>_LOG=”TRACE” * If <PACKAGE_NAME>_LOG=”TRACE”, set <PACKAGE_NAME>_FILE_LOG=”TRACE” as well

  unless <PACKAGE_NAME>_FILE_LOG is explicitly set.

• logging level (for file sink only, default=DEBUG): <PACKAGE_NAME>_FILE_LOG=”TRACE”

Usage:

• Copy this file into your package (e.g., <package_name>/utils/logger.py)

• Add the following lines to your __init__.py:

  ```python3 from <package_name>.utils.logger import Logger

  LOGGER = Logger() ```

• In anywhere of your package, import and use as follows:

  ```python3 from <package_name> import LOGGER

  LOGGER.info(“Hello, world!”) ```

Quick start reference:

• Logging methods:

  LOGGER.trace(“This is a {} from class {!r} __repr__()”, “trace message”, Class)

  .debug() .info() .success() .warning() .error() .critical() .log(level, “This is {:.4f}”, np.pi)

• Log ERROR while also capturing exception:

  try:

  …

  except Exception:

  LOGGER.exception(“This is an {}”, “exception message”)

• Available logging levels = [

  “TRACE” (5), “DEBUG” (10), “INFO” (20), “SUCCESS” (25), “WARNING” (30), “ERROR” (40), “CRITICAL” (50)

  ]

version 0.1.3

Written by Kolin Guo

class real_robot.utils.logger.Logger

Bases: object

A logger for packages that need to use loguru without disrupting the global config. All logs of this package should use an instance of this Logger.

catch(*args, **kwargs)

Delegate the catch context manager to the internal logger.

cleanup()

Remove all handlers added by this library. Restore previous filter functions for stdout/stderr handlers.

static new_filter_fn(record: dict, old_filter_fn: Callable | None) → bool

New log filter function for existing handlers