Logging Cookbook

Release 2.7.6

Guido van Rossum Fred L. Drake, Jr., editor

January 10, 2014 Python Software Foundation Email: [email protected]

Contents
1 2 3 4 5 6 Using logging in multiple modules Multiple handlers and formatters Logging to multiple destinations Conguration server example Sending and receiving logging events across a network ii iii iv v vi

Adding contextual information to your logging output viii 6.1 Using LoggerAdapters to impart contextual information . . . . . . . . . . . . . . . . . . . . . . . . . viii Using objects other than dicts to pass contextual information . . . . . . . . . . . . . . . . . . . . . . ix 6.2 Using Filters to impart contextual information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Logging to a single le from multiple processes Using le rotation An example dictionary-based conguration xi xi xii xiii xiii

7 8 9

10 Inserting a BOM into messages sent to a SysLogHandler 11 Implementing structured logging

Author Vinay Sajip <vinay_sajip at red-dove dot com> This page contains a number of recipes related to logging, which have been found useful in the past.

1 Using logging in multiple modules

Multiple calls to logging.getLogger(someLogger) return a reference to the same logger object. This is true not only within the same module, but also across modules as long as it is in the same Python interpreter process. It is true for references to the same object; additionally, application code can dene and congure a parent logger in one module and create (but not congure) a child logger in a separate module, and all logger calls to the child will pass up to the parent. Here is a main module: import logging import auxiliary_module # create logger with spam_application logger = logging.getLogger(spam_application) logger.setLevel(logging.DEBUG) # create file handler which logs even debug messages fh = logging.FileHandler(spam.log) fh.setLevel(logging.DEBUG) # create console handler with a higher log level ch = logging.StreamHandler() ch.setLevel(logging.ERROR) # create formatter and add it to the handlers formatter = logging.Formatter(%(asctime)s - %(name)s - %(levelname)s - %(message)s) fh.setFormatter(formatter) ch.setFormatter(formatter) # add the handlers to the logger logger.addHandler(fh) logger.addHandler(ch) logger.info(creating an instance of auxiliary_module.Auxiliary) a = auxiliary_module.Auxiliary() logger.info(created an instance of auxiliary_module.Auxiliary) logger.info(calling auxiliary_module.Auxiliary.do_something) a.do_something() logger.info(finished auxiliary_module.Auxiliary.do_something) logger.info(calling auxiliary_module.some_function()) auxiliary_module.some_function() logger.info(done with auxiliary_module.some_function()) Here is the auxiliary module: import logging # create logger module_logger = logging.getLogger(spam_application.auxiliary) class Auxiliary: def __init__(self): self.logger = logging.getLogger(spam_application.auxiliary.Auxiliary) self.logger.info(creating an instance of Auxiliary) def do_something(self): self.logger.info(doing something) a = 1 + 1 self.logger.info(done doing something) def some_function():

module_logger.info(received a call to "some_function") The output looks like this: 2005-03-23 23:47:11,663 - spam_application - INFO creating an instance of auxiliary_module.Auxiliary 2005-03-23 23:47:11,665 - spam_application.auxiliary.Auxiliary - INFO creating an instance of Auxiliary 2005-03-23 23:47:11,665 - spam_application - INFO created an instance of auxiliary_module.Auxiliary 2005-03-23 23:47:11,668 - spam_application - INFO calling auxiliary_module.Auxiliary.do_something 2005-03-23 23:47:11,668 - spam_application.auxiliary.Auxiliary - INFO doing something 2005-03-23 23:47:11,669 - spam_application.auxiliary.Auxiliary - INFO done doing something 2005-03-23 23:47:11,670 - spam_application - INFO finished auxiliary_module.Auxiliary.do_something 2005-03-23 23:47:11,671 - spam_application - INFO calling auxiliary_module.some_function() 2005-03-23 23:47:11,672 - spam_application.auxiliary - INFO received a call to some_function 2005-03-23 23:47:11,673 - spam_application - INFO done with auxiliary_module.some_function()

2 Multiple handlers and formatters

Loggers are plain Python objects. The addHandler() method has no minimum or maximum quota for the number of handlers you may add. Sometimes it will be benecial for an application to log all messages of all severities to a text le while simultaneously logging errors or above to the console. To set this up, simply congure the appropriate handlers. The logging calls in the application code will remain unchanged. Here is a slight modication to the previous simple module-based conguration example: import logging logger = logging.getLogger(simple_example) logger.setLevel(logging.DEBUG) # create file handler which logs even debug messages fh = logging.FileHandler(spam.log) fh.setLevel(logging.DEBUG) # create console handler with a higher log level ch = logging.StreamHandler() ch.setLevel(logging.ERROR) # create formatter and add it to the handlers formatter = logging.Formatter(%(asctime)s - %(name)s - %(levelname)s - %(message)s) ch.setFormatter(formatter) fh.setFormatter(formatter) # add the handlers to logger logger.addHandler(ch) logger.addHandler(fh) # application code logger.debug(debug message) logger.info(info message)

logger.warn(warn message) logger.error(error message) logger.critical(critical message) Notice that the application code does not care about multiple handlers. All that changed was the addition and conguration of a new handler named fh. The ability to create new handlers with higher- or lower-severity lters can be very helpful when writing and testing an application. Instead of using many print statements for debugging, use logger.debug: Unlike the print statements, which you will have to delete or comment out later, the logger.debug statements can remain intact in the source code and remain dormant until you need them again. At that time, the only change that needs to happen is to modify the severity level of the logger and/or handler to debug.

3 Logging to multiple destinations

Lets say you want to log to console and le with different message formats and in differing circumstances. Say you want to log messages with levels of DEBUG and higher to le, and those messages at level INFO and higher to the console. Lets also assume that the le should contain timestamps, but the console messages should not. Heres how you can achieve this: import logging # set up logging to file - see previous section for more details logging.basicConfig(level=logging.DEBUG, format=%(asctime)s %(name)-12s %(levelname)-8s %(message)s, datefmt=%m-%d %H:%M, filename=/temp/myapp.log, filemode=w) # define a Handler which writes INFO messages or higher to the sys.stderr console = logging.StreamHandler() console.setLevel(logging.INFO) # set a format which is simpler for console use formatter = logging.Formatter(%(name)-12s: %(levelname)-8s %(message)s) # tell the handler to use this format console.setFormatter(formatter) # add the handler to the root logger logging.getLogger().addHandler(console) # Now, we can log to the root logger, or any other logger. First the root... logging.info(Jackdaws love my big sphinx of quartz.) # Now, define a couple of other loggers which might represent areas in your # application: logger1 = logging.getLogger(myapp.area1) logger2 = logging.getLogger(myapp.area2) logger1.debug(Quick zephyrs blow, vexing daft Jim.) logger1.info(How quickly daft jumping zebras vex.) logger2.warning(Jail zesty vixen who grabbed pay from quack.) logger2.error(The five boxing wizards jump quickly.) When you run this, on the console you will see

root myapp.area1 myapp.area2 myapp.area2

: : : :

INFO INFO WARNING ERROR

Jackdaws love my big sphinx of quartz. How quickly daft jumping zebras vex. Jail zesty vixen who grabbed pay from quack. The five boxing wizards jump quickly.

and in the le you will see something like 10-22 10-22 10-22 10-22 10-22 22:19 22:19 22:19 22:19 22:19 root myapp.area1 myapp.area1 myapp.area2 myapp.area2 INFO DEBUG INFO WARNING ERROR Jackdaws love my big sphinx of quartz. Quick zephyrs blow, vexing daft Jim. How quickly daft jumping zebras vex. Jail zesty vixen who grabbed pay from quack. The five boxing wizards jump quickly.

As you can see, the DEBUG message only shows up in the le. The other messages are sent to both destinations. This example uses console and le handlers, but you can use any number and combination of handlers you choose.

4 Conguration server example

Here is an example of a module using the logging conguration server: import import import import logging logging.config time os

# read initial config file logging.config.fileConfig(logging.conf) # create and start listener on port 9999 t = logging.config.listen(9999) t.start() logger = logging.getLogger(simpleExample) try: # loop through logging calls to see the difference # new configurations make, until Ctrl+C is pressed while True: logger.debug(debug message) logger.info(info message) logger.warn(warn message) logger.error(error message) logger.critical(critical message) time.sleep(5) except KeyboardInterrupt: # cleanup logging.config.stopListening() t.join() And here is a script that takes a lename and sends that le to the server, properly preceded with the binary-encoded length, as the new logging conguration: #!/usr/bin/env python import socket, sys, struct

with open(sys.argv[1], rb) as f: data_to_send = f.read() HOST = localhost PORT = 9999 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) print(connecting...) s.connect((HOST, PORT)) print(sending config...) s.send(struct.pack(>L, len(data_to_send))) s.send(data_to_send) s.close() print(complete)

5 Sending and receiving logging events across a network

Lets say you want to send logging events across a network, and handle them at the receiving end. A simple way of doing this is attaching a SocketHandler instance to the root logger at the sending end: import logging, logging.handlers rootLogger = logging.getLogger() rootLogger.setLevel(logging.DEBUG) socketHandler = logging.handlers.SocketHandler(localhost, logging.handlers.DEFAULT_TCP_LOGGING_PORT) # dont bother with a formatter, since a socket handler sends the event as # an unformatted pickle rootLogger.addHandler(socketHandler) # Now, we can log to the root logger, or any other logger. First the root... logging.info(Jackdaws love my big sphinx of quartz.) # Now, define a couple of other loggers which might represent areas in your # application: logger1 = logging.getLogger(myapp.area1) logger2 = logging.getLogger(myapp.area2) logger1.debug(Quick zephyrs blow, vexing daft Jim.) logger1.info(How quickly daft jumping zebras vex.) logger2.warning(Jail zesty vixen who grabbed pay from quack.) logger2.error(The five boxing wizards jump quickly.) At the receiving end, you can set up a receiver using the SocketServer module. Here is a basic working example: import import import import import pickle logging logging.handlers SocketServer struct

class LogRecordStreamHandler(SocketServer.StreamRequestHandler): """Handler for a streaming logging request.

This basically logs the record using whatever logging policy is configured locally. """ def handle(self): """ Handle multiple requests - each expected to be a 4-byte length, followed by the LogRecord in pickle format. Logs the record according to whatever policy is configured locally. """ while True: chunk = self.connection.recv(4) if len(chunk) < 4: break slen = struct.unpack(>L, chunk)[0] chunk = self.connection.recv(slen) while len(chunk) < slen: chunk = chunk + self.connection.recv(slen - len(chunk)) obj = self.unPickle(chunk) record = logging.makeLogRecord(obj) self.handleLogRecord(record) def unPickle(self, data): return pickle.loads(data) def handleLogRecord(self, record): # if a name is specified, we use the named logger rather than the one # implied by the record. if self.server.logname is not None: name = self.server.logname else: name = record.name logger = logging.getLogger(name) # N.B. EVERY record gets logged. This is because Logger.handle # is normally called AFTER logger-level filtering. If you want # to do filtering, do it at the client end to save wasting # cycles and network bandwidth! logger.handle(record) class LogRecordSocketReceiver(SocketServer.ThreadingTCPServer): """ Simple TCP socket-based logging receiver suitable for testing. """ allow_reuse_address = 1 def __init__(self, host=localhost, port=logging.handlers.DEFAULT_TCP_LOGGING_PORT, handler=LogRecordStreamHandler): SocketServer.ThreadingTCPServer.__init__(self, (host, port), handler) self.abort = 0 self.timeout = 1 self.logname = None

def serve_until_stopped(self): import select abort = 0 while not abort: rd, wr, ex = select.select([self.socket.fileno()], [], [], self.timeout) if rd: self.handle_request() abort = self.abort def main(): logging.basicConfig( format=%(relativeCreated)5d %(name)-15s %(levelname)-8s %(message)s) tcpserver = LogRecordSocketReceiver() print(About to start TCP server...) tcpserver.serve_until_stopped() if __name__ == __main__: main() First run the server, and then the client. On the client side, nothing is printed on the console; on the server side, you should see something like: About 59 59 69 69 69 to start TCP server... root INFO myapp.area1 DEBUG myapp.area1 INFO myapp.area2 WARNING myapp.area2 ERROR Jackdaws love my big sphinx of quartz. Quick zephyrs blow, vexing daft Jim. How quickly daft jumping zebras vex. Jail zesty vixen who grabbed pay from quack. The five boxing wizards jump quickly.

Note that there are some security issues with pickle in some scenarios. If these affect you, you can use an alternative serialization scheme by overriding the makePickle() method and implementing your alternative there, as well as adapting the above script to use your alternative serialization.

6 Adding contextual information to your logging output

Sometimes you want logging output to contain contextual information in addition to the parameters passed to the logging call. For example, in a networked application, it may be desirable to log client-specic information in the log (e.g. remote clients username, or IP address). Although you could use the extra parameter to achieve this, its not always convenient to pass the information in this way. While it might be tempting to create Logger instances on a per-connection basis, this is not a good idea because these instances are not garbage collected. While this is not a problem in practice, when the number of Logger instances is dependent on the level of granularity you want to use in logging an application, it could be hard to manage if the number of Logger instances becomes effectively unbounded.

6.1 Using LoggerAdapters to impart contextual information

An easy way in which you can pass contextual information to be output along with logging event information is to use the LoggerAdapter class. This class is designed to look like a Logger, so that you can call debug(), info(), warning(), error(), exception(), critical() and log(). These methods have the same signatures as their counterparts in Logger, so you can use the two types of instances interchangeably.

When you create an instance of LoggerAdapter, you pass it a Logger instance and a dict-like object which contains your contextual information. When you call one of the logging methods on an instance of LoggerAdapter, it delegates the call to the underlying instance of Logger passed to its constructor, and arranges to pass the contextual information in the delegated call. Heres a snippet from the code of LoggerAdapter: def debug(self, msg, *args, **kwargs): """ Delegate a debug call to the underlying logger, after adding contextual information from this adapter instance. """ msg, kwargs = self.process(msg, kwargs) self.logger.debug(msg, *args, **kwargs) The process() method of LoggerAdapter is where the contextual information is added to the logging output. Its passed the message and keyword arguments of the logging call, and it passes back (potentially) modied versions of these to use in the call to the underlying logger. The default implementation of this method leaves the message alone, but inserts an extra key in the keyword argument whose value is the dict-like object passed to the constructor. Of course, if you had passed an extra keyword argument in the call to the adapter, it will be silently overwritten. The advantage of using extra is that the values in the dict-like object are merged into the LogRecord instances __dict__, allowing you to use customized strings with your Formatter instances which know about the keys of the dict-like object. If you need a different method, e.g. if you want to prepend or append the contextual information to the message string, you just need to subclass LoggerAdapter and override process() to do what you need. Here is a simple example: class CustomAdapter(logging.LoggerAdapter): """ This example adapter expects the passed in dict-like object to have a connid key, whose value in brackets is prepended to the log message. """ def process(self, msg, kwargs): return [%s] %s % (self.extra[connid], msg), kwargs which you can use like this: logger = logging.getLogger(__name__) adapter = CustomAdapter(logger, {connid: some_conn_id}) Then any events that you log to the adapter will have the value of some_conn_id prepended to the log messages. Using objects other than dicts to pass contextual information You dont need to pass an actual dict to a LoggerAdapter - you could pass an instance of a class which implements __getitem__ and __iter__ so that it looks like a dict to logging. This would be useful if you want to generate values dynamically (whereas the values in a dict would be constant).

6.2 Using Filters to impart contextual information

You can also add contextual information to log output using a user-dened Filter. Filter instances are allowed to modify the LogRecords passed to them, including adding additional attributes which can then be output using a suitable format string, or if needed a custom Formatter. For example in a web application, the request being processed (or at least, the interesting parts of it) can be stored in a threadlocal (threading.local) variable, and then accessed from a Filter to add, say, information from the request - say, the remote IP address and remote users username - to the LogRecord, using the attribute names ip

and user as in the LoggerAdapter example above. In that case, the same format string can be used to get similar output to that shown above. Heres an example script: import logging from random import choice class ContextFilter(logging.Filter): """ This is a filter which injects contextual information into the log. Rather than use actual contextual information, we just use random data in this demo. """ USERS = [jim, fred, sheila] IPS = [123.231.231.123, 127.0.0.1, 192.168.0.1] def filter(self, record): record.ip = choice(ContextFilter.IPS) record.user = choice(ContextFilter.USERS) return True

if __name__ == __main__: levels = (logging.DEBUG, logging.INFO, logging.WARNING, logging.ERROR, logging.CRITICAL) logging.basicConfig(level=logging.DEBUG, format=%(asctime)-15s %(name)-5s %(levelname)-8s IP: %(ip)-15s User a1 = logging.getLogger(a.b.c) a2 = logging.getLogger(d.e.f) f = ContextFilter() a1.addFilter(f) a2.addFilter(f) a1.debug(A debug message) a1.info(An info message with %s, some parameters) for x in range(10): lvl = choice(levels) lvlname = logging.getLevelName(lvl) a2.log(lvl, A message at %s level with %d %s, lvlname, 2, parameters) which, when run, produces something like: 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 2010-09-06 22:38:15,292 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,300 22:38:15,301 22:38:15,301 22:38:15,301 a.b.c a.b.c d.e.f d.e.f d.e.f d.e.f d.e.f d.e.f d.e.f d.e.f d.e.f d.e.f DEBUG INFO CRITICAL ERROR DEBUG ERROR CRITICAL CRITICAL DEBUG ERROR DEBUG INFO IP: IP: IP: IP: IP: IP: IP: IP: IP: IP: IP: IP: 123.231.231.123 192.168.0.1 127.0.0.1 127.0.0.1 127.0.0.1 123.231.231.123 192.168.0.1 127.0.0.1 192.168.0.1 127.0.0.1 123.231.231.123 123.231.231.123 User: User: User: User: User: User: User: User: User: User: User: User: fred sheila sheila jim sheila fred jim sheila jim sheila fred fred

A debug message An info message w A message at CRIT A message at ERRO A message at DEBU A message at ERRO A message at CRIT A message at CRIT A message at DEBU A message at ERRO A message at DEBU A message at INFO

7 Logging to a single le from multiple processes

Although logging is thread-safe, and logging to a single le from multiple threads in a single process is supported, logging to a single le from multiple processes is not supported, because there is no standard way to serialize access to a single le across multiple processes in Python. If you need to log to a single le from multiple processes, one way of doing this is to have all the processes log to a SocketHandler, and have a separate process which implements a socket server which reads from the socket and logs to le. (If you prefer, you can dedicate one thread in one of the existing processes to perform this function.) This section documents this approach in more detail and includes a working socket receiver which can be used as a starting point for you to adapt in your own applications. If you are using a recent version of Python which includes the multiprocessing module, you could write your own handler which uses the Lock class from this module to serialize access to the le from your processes. The existing FileHandler and subclasses do not make use of multiprocessing at present, though they may do so in the future. Note that at present, the multiprocessing module does not provide working lock functionality on all platforms (see http://bugs.python.org/issue3770).

8 Using le rotation
Sometimes you want to let a log le grow to a certain size, then open a new le and log to that. You may want to keep a certain number of these les, and when that many les have been created, rotate the les so that the number of les and the size of the les both remain bounded. For this usage pattern, the logging package provides a RotatingFileHandler: import glob import logging import logging.handlers LOG_FILENAME = logging_rotatingfile_example.out # Set up a specific logger with our desired output level my_logger = logging.getLogger(MyLogger) my_logger.setLevel(logging.DEBUG) # Add the log message handler to the logger handler = logging.handlers.RotatingFileHandler( LOG_FILENAME, maxBytes=20, backupCount=5) my_logger.addHandler(handler) # Log some messages for i in range(20): my_logger.debug(i = %d % i) # See what files are created logfiles = glob.glob(%s* % LOG_FILENAME) for filename in logfiles: print(filename) The result should be 6 separate les, each with part of the log history for the application: logging_rotatingfile_example.out logging_rotatingfile_example.out.1 logging_rotatingfile_example.out.2

logging_rotatingfile_example.out.3 logging_rotatingfile_example.out.4 logging_rotatingfile_example.out.5 The most current le is always logging_rotatingfile_example.out, and each time it reaches the size limit it is renamed with the sufx .1. Each of the existing backup les is renamed to increment the sufx (.1 becomes .2, etc.) and the .6 le is erased. Obviously this example sets the log length much too small as an extreme example. You would want to set maxBytes to an appropriate value.

9 An example dictionary-based conguration

Below is an example of a logging conguration dictionary - its taken from the documentation on the Django project. This dictionary is passed to dictConfig() to put the conguration into effect:

LOGGING = { version: 1, disable_existing_loggers: True, formatters: { verbose: { format: %(levelname)s %(asctime)s %(module)s %(process)d %(thread)d %(messag }, simple: { format: %(levelname)s %(message)s }, }, filters: { special: { (): project.logging.SpecialFilter, foo: bar, } }, handlers: { null: { level:DEBUG, class:django.utils.log.NullHandler, }, console:{ level:DEBUG, class:logging.StreamHandler, formatter: simple }, mail_admins: { level: ERROR, class: django.utils.log.AdminEmailHandler, filters: [special] } }, loggers: { django: { handlers:[null], propagate: True, level:INFO,

}, django.request: { handlers: [mail_admins], level: ERROR, propagate: False, }, myproject.custom: { handlers: [console, mail_admins], level: INFO, filters: [special] } } } For more information about this conguration, you can see the relevant section of the Django documentation.

10 Inserting a BOM into messages sent to a SysLogHandler

RFC 5424 requires that a Unicode message be sent to a syslog daemon as a set of bytes which have the following structure: an optional pure-ASCII component, followed by a UTF-8 Byte Order Mark (BOM), followed by Unicode encoded using UTF-8. (See the relevant section of the specication.) In Python 2.6 and 2.7, code was added to SysLogHandler to insert a BOM into the message, but unfortunately, it was implemented incorrectly, with the BOM appearing at the beginning of the message and hence not allowing any pure-ASCII component to appear before it. As this behaviour is broken, the incorrect BOM insertion code is being removed from Python 2.7.4 and later. However, it is not being replaced, and if you want to produce RFC 5424-compliant messages which include a BOM, an optional pure-ASCII sequence before it and arbitrary Unicode after it, encoded using UTF-8, then you need to do the following: 1. Attach a Formatter instance to your SysLogHandler instance, with a format string such as: uASCII section\ufeffUnicode section The Unicode code point u\ufeff, when encoded using UTF-8, will be encoded as a UTF-8 BOM the byte-string \xef\xbb\xbf. 2. Replace the ASCII section with whatever placeholders you like, but make sure that the data that appears in there after substitution is always ASCII (that way, it will remain unchanged after UTF-8 encoding). 3. Replace the Unicode section with whatever placeholders you like; if the data which appears there after substitution contains characters outside the ASCII range, thats ne it will be encoded using UTF-8. If the formatted message is Unicode, it will be encoded using UTF-8 encoding by SysLogHandler. If you follow the above rules, you should be able to produce RFC 5424-compliant messages. If you dont, logging may not complain, but your messages will not be RFC 5424-compliant, and your syslog daemon may complain.

11 Implementing structured logging

Although most logging messages are intended for reading by humans, and thus not readily machine-parseable, there might be cirumstances where you want to output messages in a structured format which is capable of being parsed by a program (without needing complex regular expressions to parse the log message). This is straightforward to achieve using the logging package. There are a number of ways in which this could be achieved, but the following is a simple approach which uses JSON to serialise the event in a machine-parseable manner:

import json import logging class StructuredMessage(object): def __init__(self, message, **kwargs): self.message = message self.kwargs = kwargs def __str__(self): return %s >>> %s % (self.message, json.dumps(self.kwargs)) _ = StructuredMessage # optional, to improve readability

logging.basicConfig(level=logging.INFO, format=%(message)s) logging.info(_(message 1, foo=bar, bar=baz, num=123, fnum=123.456)) If the above script is run, it prints: message 1 >>> {"fnum": 123.456, "num": 123, "bar": "baz", "foo": "bar"} Note that the order of items might be different according to the version of Python used. If you need more specialised processing, you can use a custom JSON encoder, as in the following complete example: from __future__ import unicode_literals import json import logging # This next bit is to ensure the script runs unchanged on 2.x and 3.x try: unicode except NameError: unicode = str class Encoder(json.JSONEncoder): def default(self, o): if isinstance(o, set): return tuple(o) elif isinstance(o, unicode): return o.encode(unicode_escape).decode(ascii) return super(Encoder, self).default(o) class StructuredMessage(object): def __init__(self, message, **kwargs): self.message = message self.kwargs = kwargs def __str__(self): s = Encoder().encode(self.kwargs) return %s >>> %s % (self.message, s) _ = StructuredMessage # optional, to improve readability

def main(): logging.basicConfig(level=logging.INFO, format=%(message)s)

logging.info(_(message 1, set_value=set([1, 2, 3]), snowman=\u2603)) if __name__ == __main__: main() When the above script is run, it prints: message 1 >>> {"snowman": "\u2603", "set_value": [1, 2, 3]} Note that the order of items might be different according to the version of Python used.