Fixed Length Formatting for Combination of Fields in Python logging modules

Question

I am using the Python logging module in my application. Here is my current format.

LOGGER_FORMAT = '%(asctime)s %(levelname)-5s [%(threadName)-10s] %(name)s.%(module)s.%(funcName)s : %(message)s'
logging.basicConfig(level=logging.INFO, format=LOGGER_FORMAT)

And a sample output:

2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.util.get_login_info : Log line 1
2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.cache.call : Log line 2
2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.web.parse_response : Log line 3

What I am looking for is that I get the message displayed from the fixed position instead of varying start points.

I am able to configure the fixed length for each individual attribute, but I wanted to configure a fixed length for the entire name.module.funcName part, so that my output is like below:

2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.util.get_login_info : Log line 1
2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.cache.call          : Log line 2
2018-03-14 11:35:27,257 INFO  [Thread-2  ] my_app.web.parse_response  : Log line 3

Is it possible, and how can I configure this?

Answer 1

Your best bet would be to use a logging.Filter such as:

class MyFilter(logging.Filter):
    def filter(self, record):
        record.expandedFuncName = '%s.%s.%s' % (record.name, record.module, record.funcName)
        return True

and then use a format string something like

%(asctime)s %(levelname)-5s [%(threadName)-10s] %(expandedFuncName)-20.20s : %(message)s

You'd need to ensure that an instance of this filter was applied to every handler which uses a formatter with expandedFuncName in the format string. You can't do this with basicConfig , so I'd recommend using the dictConfig logging configuration API instead.

Note that the logger name will generally include the module name if you use the convention logger = logging.getLogger(__name__) as recommended in the logging documentation, though this won't apply exactly in the case of the __main__ module.