pygogo: a Python logger with superpowers
Index
Introduction | Requirements | Motivation | Usage | Installation | Project Structure | Design Principles | Structured Logging | Formatters | Handlers | Scripts | Contributing | License
Introduction
pygogo is a Python logging library and command-line interface with super powers.
pygogo leverages the standard Python logging module under the hood, so there's
no need to learn yet-another logging library. The default implementation sends
all messages to stdout, and any messages at level WARNING or above also to stderr.
With pygogo, you can
- Log via different handlers depending on the event severity
- Format log messages as plain text, csv, json, and more..
- Send logs to stdout, stderr, file, email, sockets, and more..
- Inter-operate with the standard python logging module
- and much more...
Requirements
pygogo has been tested and is known to work on Python 3.7, 3.8, and 3.9; and PyPy3.7.
Motivation
The standard logging module is great, but requires a ton of boilerplate before you can do anything really interesting with it. I designed pygogo to provide many useful logging use-cases out of the box. A reimplementation of Using LoggerAdapters to impart contextual information is shown below:
import pygogo as gogo
logger = gogo.Gogo(__name__).get_structured_logger(connid='1234')
logger.info('log message')
# Prints the following to stdout
{"message": "log message", "connid": "1234"}Usage
pygogo is intended to be used either directly as a Python library or from the terminal via the command-line interface.
Library
Examples
Hello World
from pygogo import logger
logger.debug('hello world')
logger.error('hello error')
# Prints the following to `stdout`
hello world
hello error
# Prints the following to `stderr`
hello errorLog based debugging
import pygogo as gogo
def main(verbose=False):
logger = gogo.Gogo(__name__, verbose=verbose).logger
logger.debug('I will log to `stdout` only if `verbose` is True')
logger.info('I will log to `stdout` always')
logger.warning('I will log to both `stdout` and `stderr` always')Disabled dual logging
import pygogo as gogo
logger = gogo.Gogo(monolog=True).logger
logger.debug('debug message')
logger.info('info message')
logger.warning('warning message')
logger.error('error message')
logger.critical('critical message')
# Prints the following to `stdout.log` (all messages at level `INFO` or below):
debug message
info message
# Prints the following to `stderr` (messages at level `WARNING` or above):
warning message
error message
critical messageCustom formatter [1]
import logging
import pygogo as gogo
log_format = '%(asctime)s - %(name)s - %(levelname)s - %(message)s'
formatter = logging.Formatter(log_format)
logger = gogo.Gogo(
'examples.fmt',
low_hdlr=gogo.handlers.file_hdlr('custom_fmt.log'),
low_formatter=formatter,
high_level='error',
high_formatter=formatter).logger
# Now let's log something!
logger.debug('debug message')
logger.info('info message')
logger.warn('warn message')
logger.error('error message')
logger.critical('critical message')
# Prints the following to `custom_fmt.log` (all messages):
2015-12-18 18:51:30,416 - examples.fmt.base - DEBUG - debug message
2015-12-18 18:51:30,416 - examples.fmt.base - INFO - info message
2015-12-18 18:51:30,416 - examples.fmt.base - WARNING - warn message
2015-12-18 18:51:30,416 - examples.fmt.base - ERROR - error message
2015-12-18 18:51:30,416 - examples.fmt.base - CRITICAL - critical message
# Prints the following to `stderr` (messages at level `ERROR` or above):
2015-12-18 18:51:30,416 - examples.fmt.base - ERROR - error message
2015-12-18 18:51:30,416 - examples.fmt.base - CRITICAL - critical messageStructured logging [2]
import pygogo as gogo
formatter = gogo.formatters.structured_formatter
kwargs = {'low_level': 'info', 'low_formatter': formatter}
logger = gogo.Gogo('examples.structured', **kwargs).logger
extra = {'set_value': set([1, 2, 3]), 'snowman': '☃'}
logger.info('log message', extra=extra) # doctest: +ELLIPSIS
# Prints the following to `stdout`:
{"snowman": "\u2603", "name": "examples.structured.base", "level": "INFO", "message": "log message", "time": "2015-12-18 18:52:39", "msecs": 58.973073959350586, "set_value": [1, 2, 3]}Using Filters to impart contextual information [3]
import logging
import pygogo as gogo
levels = ('DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL')
log_frmt = (
'%(asctime)-4s %(name)-5s %(levelname)-8s IP: %(ip)-15s User: '
'%(user)-8s %(message)s')
formatter = logging.Formatter(log_frmt)
going = gogo.Gogo('a', low_formatter=formatter)
a1 = going.get_logger('b.c', ip='123.231.231.123', user='fred')
a2 = going.get_logger('e.f', ip='192.168.0.1', user='sheila')
# Now let's log something!
a1.debug('A debug message')
a1.info('An info %s', 'message')
for level in [getattr(logging, l) for l in levels]:
name = logging.getLevelName(level)
a2.log(level, 'A %s msg', name)
# Prints the following to `stdout` (all messages):
2015-12-19 10:12:24,479 a.b.c DEBUG IP: 123.231.231.123 User: fred A debug message
2015-12-19 10:12:24,479 a.b.c INFO IP: 123.231.231.123 User: fred An info message
2015-12-19 10:12:24,479 a.e.f DEBUG IP: 192.168.0.1 User: sheila A DEBUG msg
2015-12-19 10:12:24,479 a.e.f INFO IP: 192.168.0.1 User: sheila AN INFO msg
2015-12-19 10:12:24,479 a.e.f WARNING IP: 192.168.0.1 User: sheila A WARNING msg
2015-12-19 10:12:24,479 a.e.f ERROR IP: 192.168.0.1 User: sheila AN ERROR msg
2015-12-19 10:12:24,479 a.e.f CRITICAL IP: 192.168.0.1 User: sheila A CRITICAL msg
# Prints the following to `stderr` (messages at level `WARNING` or above):
2015-12-19 10:12:24,479 a.e.f WARNING IP: 192.168.0.1 User: sheila A WARNING msg
2015-12-19 10:12:24,479 a.e.f ERROR IP: 192.168.0.1 User: sheila AN ERROR msg
2015-12-19 10:12:24,479 a.e.f CRITICAL IP: 192.168.0.1 User: sheila A CRITICAL msgMultiple loggers [4]
import pygogo as gogo
going = gogo.Gogo(
'examples.lggrs',
low_hdlr=gogo.handlers.file_hdlr('multi_lggrs.log'),
low_formatter=gogo.formatters.fixed_formatter,
high_level='info',
high_formatter=gogo.formatters.console_formatter)
root = going.logger
logger1 = going.get_logger('area1')
logger2 = going.get_logger('area2')
# Now let's log something!
root.info('Jackdaws love my big sphinx.')
logger1.debug('Quick zephyrs blow, daft Jim.')
logger1.info('How daft jumping zebras vex.')
logger2.warning('Jail zesty vixen who grabbed pay.')
logger2.error('The five boxing wizards jump.')
# Prints the following to `multi_lggrs.log` (all messages):
2015-12-18 17:21:37.417 examples.lggrs.base INFO Jackdaws love my big sphinx.
2015-12-18 17:21:37.417 examples.lggrs.area1 DEBUG Quick zephyrs blow, daft Jim.
2015-12-18 17:21:37.417 examples.lggrs.area1 INFO How daft jumping zebras vex.
2015-12-18 17:21:37.417 examples.lggrs.area2 WARNING Jail zesty vixen who grabbed pay.
2015-12-18 17:21:37.417 examples.lggrs.area2 ERROR The five boxing wizards jump.
# Prints the following to `stderr` (messages at level `INFO` or above):
examples.lggrs.base: INFO Jackdaws love my big sphinx.
examples.lggrs.area1: INFO How daft jumping zebras vex.
examples.lggrs.area2: WARNING Jail zesty vixen who grabbed pay.
examples.lggrs.area2: ERROR The five boxing wizards jump.Notes
| [1] | https://docs.python.org/2/howto/logging-cookbook.html#multiple-handlers-and-formatters |
| [2] | https://docs.python.org/2/howto/logging-cookbook.html#implementing-structured-logging |
| [3] | https://docs.python.org/2/howto/logging-cookbook.html#using-filters-to-impart-contextual-information |
| [4] | https://docs.python.org/2/howto/logging-cookbook.html#logging-to-multiple-destinations |
Command-line Interface
Examples
Basic Usage
gogo [options] <message>show help
gogo -hCLI usage
usage: gogo [options] <message>
description: Logs a given message
- positional arguments:
- message The message to log (defaults to reading from stdin).
- optional arguments:
-h, --help show this help message and exit -l LEVEL, --msg-level LEVEL The level to log the message (default: info). Must be one of: critical, error, warning, info, debug. -n NAME, --name NAME The logger name (default: pygogo) -D HANDLER, --high-hdlr HANDLER The high pass log handler (default: stderr). Must be one of: buffered, email, file, fileobj, socket, stderr, stdout, syslog, webhook. -d HANDLER, --low-hdlr HANDLER The low pass log handler (default: stdout). Must be one of: buffered, email, file, fileobj, socket, stderr, stdout, syslog, webhook. -L LEVEL, --high-level LEVEL Min level to log to the high pass handler (default: warning). Must be one of: buffered, email, file, fileobj, socket, stderr, stdout, syslog, webhook. -e LEVEL, --low-level LEVEL Min level to log to the low pass handler (default: debug). Must be one of: buffered, email, file, fileobj, socket, stderr, stdout, syslog, webhook. -F FORMAT, --high-format FORMAT High pass handler log format (default: basic). Must be one of: basic, bom, console, csv, fixed, json, structured. -o FORMAT, --low-format FORMAT Low pass handler log format (default: basic). Must be one of: basic, bom, console, csv, fixed, json, structured. -m, --monolog Log high level events only to high pass handler. -f FILENAME, --filename FILENAME The filename to log to. Required for the follow handlers: file. -s SUBJECT, --subject SUBJECT The log subject (default: You've got mail). Used in the follow handlers: email. -u URL, --url URL The log url. Required for the follow handlers: webhook. -H HOST, --host HOST The host. Used in the follow handlers: socket and syslog. -p NUM, --port NUM The port number. Used in the follow handlers: socket and syslog. -t, --tcp Use TCP instead of UDP. Used in the follow handlers: socket and syslog. -g, --get Use a GET request instead of POST. Used in the follow handlers: webhook. -v, --version Show version and exit. -V, --verbose Increase output verbosity.
Hello World
gogo 'hello world'Log based debugging
gogo 'default info level will log to `stdout`'
gogo --level=debug "debug won't log"
gogo --level=debug -V 'verbose will log to `stdout`'
gogo --level=info 'info will log to `stdout`'
gogo --level=warning 'warning will log to both `stdout` and `stderr`'
# Prints the following to `stdout`:
default info level will log to `stdout`
verbose will log to `stdout`
info will log to `stdout`
warning will log to both `stdout` and `stderr`
# Prints the following to `stderr`:
warning will log to both `stdout` and `stderr`Disable dual logging
gogo --level=debug -V 'debug message'
gogo --level=info 'info message'
gogo --level=warning -m 'warning message'
gogo --level=error -m 'error message'
gogo --level=critical -m 'critical message'
# Prints the following to `stdout.log` (all messages at level `INFO` or below):
debug message
info message
# Prints the following to `stderr` (messages at level `WARNING` or above):
warning message
error message
critical messageStructured logging
gogo --low-format=json 'log message'
# Prints the following to `stdout`:
{"time": "2015-12-19 11:26:53.776", "name": "pygogo.runner", "level": "INFO", "message": "log message"}Alternate handler
gogo --low-hdlr=file 'log message'
# Prints the following to `pygogo.log` in the current dir (assuming the current dir is named `pygogo`):
{"time": "2015-12-19 11:26:53.776", "name": "pygogo.runner", "level": "INFO", "message": "log message"}Installation
(You are using a virtualenv, right?)
At the command line, install pygogo using either pip (recommended)
pip install pygogoor easy_install
easy_install pygogoPlease see the installation doc for more details.
Project Structure
┌── bin
│ └── gogo
├── docs
│ ├── AUTHORS.rst
│ ├── CHANGES.rst
│ ├── INSTALLATION.rst
│ └── TODO.rst
├── helpers
│ ├── check-stage
│ ├── clean
│ ├── pippy
│ ├── srcdist
│ └── wheel
├── pygogo
│ ├── __init__.py
│ ├── formatters.py
│ ├── handlers.py
│ ├── main.py
│ └── utils.py
├── tests
│ ├── __init__.py
│ ├── standard.rc
│ ├── test.py
│ └── test_main.py
├── CONTRIBUTING.rst
├── LICENSE
├── MANIFEST.in
├── Makefile
├── README.rst
├── dev-requirements.txt
├── examples.py
├── manage.py
├── py2-requirements.txt
├── setup.cfg
├── setup.py
└── tox.iniDesign Principles
- the built-in
loggingmodule isn't broken so don't reinvent the wheel - prefer functions over objects
- keep the API as simple as possible
Structured Logging
There are severals ways to get structured (machine readable) log messages using pygogo. Each method makes a different customization/complexity trade-off which is outlined below:
Setup
The following methods make use of these variables.
import pygogo as gogo
kwargs = {'contextual': True}
extra = {'additional': True}Methods
basic structured logger
The simplest to use. Useful if you don’t need message metadata, i.e., log level, log name, and log time.
logger = gogo.Gogo('basic').get_structured_logger('base', **kwargs)
logger.debug('message', extra=extra)
# Prints the following to `stdout`:
{"additional": true, "contextual": true, "message": "message"}structured formatter
Requires an additional step of specifying a formatter. Useful if you need message metadata, i.e., log level, log name, and log time.
formatter = gogo.formatters.structured_formatter
logger = gogo.Gogo('struct', low_formatter=formatter).get_logger(**kwargs)
logger.debug('message', extra=extra)
# Prints the following to `stdout`:
{"additional": true, "contextual": true, "level": "DEBUG", "message": "message", "msecs": 760.5140209197998, "name": "struct.base", "time": "2015-12-19 14:25:58"}JSON formatter
Requires an additional step of specifying a formatter. Useful if you require
millisecond precision in the date. If you are ok with having the milliseconds
in a separate field, consider the structured formatter since it supports
the extra keyword and contextual information.
formatter = gogo.formatters.json_formatter
logger = gogo.Gogo('json', low_formatter=formatter).get_logger(**kwargs)
logger.debug('message', extra=extra)
# Prints the following to `stdout`:
{"level": "DEBUG", "message": "message", "name": "json.base", "time": "2015-12-19 14:25:58.760"}
# Note that both `extra` and `kwargs` were ignoredcustom logger
The most complex and customizable. Useful if you need a custom
log or date format not provided by the above methods. However, even though this
method supports the extra keyword when logging, it is static (unlike the
structured logger or structured formatter). This is because the log
format must be specified at the time of the log's creation and therefore can't
adapt to log messages with differing extra parameters.
logfmt = (
'{"time": "%(asctime)s.%(msecs)d", "name": "%(name)s", "level":'
' "%(levelname)s", "message": "%(message)s", '
'"contextual": "%(contextual)s", "additional": "%(additional)s"}')
fmtr = logging.Formatter(logfmt, datefmt=gogo.formatters.DATEFMT)
logger = gogo.Gogo('custom', low_formatter=fmtr).get_logger(**kwargs)
logger.debug('message', extra=extra)
# Prints the following to `stdout`:
{"additional": "True", "contextual": "True", "level": "DEBUG", "message": "message", "name": "custom.logger", "time": "2015-12-19 14:25:58.760"}Summary
The following table can help make sense of the different methods:
| structured logger | structured formatter | json formatter | custom logger | |
|---|---|---|---|---|
| contextual information | ✔ | ✔ | ✔ | |
extra param support |
✔ | ✔ | ✔ | |
dynamic extra support |
✔ | ✔ | ||
| message metadata | ✔ | ✔ | ✔ | |
| available via the command line | ✔ | ✔ | ||
msecs field |
✔ | |||
| milliseconds in time field | ✔ | ✔ | ||
| custom date format | ✔ | |||
| custom log format | ✔ |
Formatters
pygogo has several builtin formatters and also supports any logging.Formatter
instance.
Examples
builtin CSV format in python
import pygogo as gogo
formatter = gogo.formatters.csv_formatter
gogo.Gogo('csv', low_formatter=formatter).logger.debug('message')
# Prints the following to `stdout`:
2015-12-19 17:03:48.99,csv.base,DEBUG,"message"logging.Formatter instance in python
import logging
import pygogo as gogo
datefmt = gogo.formatters.DATEFMT
formatter = logging.Formatter(gogo.formatters.CSV_FORMAT, datefmt=datefmt)
gogo.Gogo('csv', low_formatter=formatter).get_logger('custom').debug('message')
# Prints the following to `stdout`:
2015-12-19 17:03:48.99,csv.custom,DEBUG,"message"builtin CSV format via CLI
gogo --low-format=csv 'message'
# Prints the following to `stdout`:
2015-12-19 15:51:32.16,pygogo.runner,INFO,"message"Summary
The following table can help make sense of the different builtin formatters:
| name | message |
|---|---|
| basic | message |
| bom | message |
| console | name: INFO message |
| csv | 2015-12-19 15:51:32.16,name,INFO,"message" |
| fixed | 2015-12-19 15:51:32.16 name INFO message |
| json | {"level": "INFO", "message": "message", "name": "name", "time": "2015-12-19 15:51:32.16"} |
| structured | {"level": "INFO", "message": "message", "msecs": 16.5140209197998, "name": "name", "time": "2015-12-19 15:51:32"} |
Handlers
pygogo has several builtin handlers and also supports any instance from the
logging.handlers module.
Examples
builtin stdout handler in python
import pygogo as gogo
hdlr = gogo.handlers.stdout_hdlr()
gogo.Gogo('stdout', low_hdlr=hdlr).logger.debug('message')
# Prints 'message' to `stdout`logging.StreamHandler instance in python
import logging
import sys
import pygogo as gogo
hdlr = logging.StreamHandler(sys.stdout)
gogo.Gogo('stdout', low_hdlr=hdlr).get_logger('custom').debug('message')
# Prints 'message' to `stdout`builtin CSV format via CLI
gogo --low-hdlr=stdout 'message'
# Prints 'message' to `stdout`Summary
The following table can help make sense of the different builtin handlers:
| name | description |
|---|---|
| buffered | Holds log in memory until it reaches its capacity, or it logs a message with a level at or above the flush level |
| Emails log to a given email address | |
| file | Writes log to a given filename |
| fileobj | Writes log to a given file-like object |
| socket | Writes log to a given network socket |
| stderr | Writes log to standard error |
| stdout | Writes log to standard output |
| syslog | Writes log to syslog |
| webhook | POSTs log to a url |
Scripts
pygogo comes with a built in task manager manage.py
Setup
pip install -r dev-requirements.txtExamples
Run python linter and nose tests
manage lint
manage testContributing
Please mimic the coding style/conventions used in this repo. If you add new classes or functions, please add the appropriate doc blocks with examples. Also, make sure the python linter and nose tests pass.
Please see the contributing doc for more details.
License
pygogo is distributed under the MIT License.
