integrating loguru with Flask and/or Celery loggers

I really like loguru and it simplifies my whole setup. I have a hacked custom Logger that I use in my code and Flask webapps. I’d like to replace it with loguru, since it basically does the things I hacked. What’s the procedure to properly integrate loguru into other apps that might have loggers based on the standard logger?

This is what I’ve done to integrate loguru into my code and Flask web app, and the error I am getting.

#Function to initialize a loguru logger
def initLog(logpath):
    ''' test loguru '''
    from loguru import logger
    import sys
    import logging

    logger.add(sys.stderr, format="{level} | {message} ",
               level="INFO", colorize=True, enqueue=True, backtrace=True)

    fmt = "{time:YYYY-MM-DD at HH:mm:ss} | {level} | {message} [{function} @ {file}]"
    logger.add("logtest.log", rotation='10 seconds', format=fmt,
               colorize=True, backtrace=True, enqueue=True)

    new_level = logger.level("IMPORTANT", no=25, color="<magenta>")

    # capture warnings
    showwarning_ = warnings.showwarning

    def showwarning(message, *args, **kwargs):
        logger.warning(message)
        showwarning_(message, *args, **kwargs)

    warnings.showwarning = showwarning

    # propagate to standard logger
    class PropagateHandler(logging.Handler):
        def emit(self, record):
            logging.getLogger(record.name).handle(record)

    logger.add(PropagateHandler(), format="{message}")

    return logger

Then I’m trying to integrate it into my Flask webapp

log = initLog('testlog.log')

app = Flask()
# add custom logger to the Flask app
app.logger.addHandler(log)

# use the Flask logger to log messages
app.logger.info('Loading Development Config!')

but I get the following error:

    app.logger.info('Loading Development Config!')
  File "/Users/Brian/anaconda3/lib/python3.7/logging/__init__.py", line 1375, in info
    Log 'msg % args' with severity 'INFO'.
  File "/Users/Brian/anaconda3/lib/python3.7/logging/__init__.py", line 1383, in info
    self._log(INFO, msg, args, **kwargs)
  File "/Users/Brian/anaconda3/lib/python3.7/logging/__init__.py", line 1519, in _log
    self.handle(record)
  File "/Users/Brian/anaconda3/lib/python3.7/logging/__init__.py", line 1529, in handle
    self.callHandlers(record)
  File "/Users/Brian/anaconda3/lib/python3.7/logging/__init__.py", line 1590, in callHandlers
    if record.levelno >= hdlr.level:
TypeError: '>=' not supported between instances of 'int' and 'method'

What’s the correct way to integrate loguru into other libraries like Flask or Celery?

[Delgan]

Hey @havok2063.

I think a confusion between “handler” and “logger” may explain the error you are seeing. The logger is a basic object to use when you want to write a log message. Handlers are objects which are in charge or receiving this log messages, and processing adequately: write it on stderr, write it in a file, send it by email, etc.

A logger has methods like .info(), .debug() and .warning() while an handler has methods like .write() and .emit().

So, when you use app.logger.addHandler(log), you pass a logger object while Flask expects an handler, hence bug occurs. 😄

Maybe you can try to use an InterceptHandler() instead, which should send Flask’s logs back to Loguru’s handlers:

class InterceptHandler(logging.Handler):
    def emit(self, record):
        # Retrieve context where the logging call occurred, this happens to be in the 6th frame upward
        logger_opt = logger.opt(depth=6, exception=record.exc_info)
        logger_opt.log(record.levelno, record.getMessage())
handler = InterceptHandler()
handler.setLevel(0)
app.logger.addHandler(handler)

[Delgan]

I’m closing this since it seems I answered the initial question. Be sure that I will work on the C implementation as soon as possible. 😄