Logging
Ape provides a logger and uses it to show messages throughout the execution of its modules.
Every CLI command comes with the logger in Ape, even custom user scripts (unless they change the behavior of --verbosity
).
The following log levels are available with Ape:
Log Level |
Numeric Value |
Purpose |
Color |
---|---|---|---|
DEBUG |
10 |
Debug stuff |
Blue |
INFO |
20 |
General information |
Blue |
SUCCESS |
21 |
To mark a successful operation |
Green |
WARNING |
30 |
Indicates a potential issue |
Yellow |
ERROR |
40 |
An error occurred |
Red |
Note
SUCCESS
is a non-standard verbosity level custom to the framework.
It is shown during INFO
but not shown if set to WARNING
or above.
CLI Logging
If you are running into issues and wish to see more information logged, you likely want to run your command with --verbosity DEBUG
or -v debug
:
ape --verbosity DEBUG my_cmd # long form
ape -v debug my_cmd # short form
This will output HTTP requests and anything else with a DEBUG
logging verbosity in Ape.
Alternatively, you may wish to log less and show important logs, such as ERROR
logs.
To do this, use the ERROR
verbosity:
ape my_cmd -v ERROR
NOTE: You can put the verbosity flag anywhere in your CLI command for most commands.
To disable logging completely, you can use keyword “DISABLE” or “NONE” as the --verbosity
value:
ape my_cmd -v NONE
Now, Ape won’t log at all. This can be useful if you need the output of the CLI for something else.
Python Logging
You can also import and use the logger in your own Python scripts or commands:
from ape.logging import logger, LogLevel
def main():
logger.info("You have entered `main()`.")
logger.set_level(LogLevel.WARNING)
You can also use the .at_level()
context-manager to temporarily change the log-level:
from ape.logging import logger, LogLevel
def main():
with logger.at_level(LogLevel.WARNING):
# An operation where you want to ensure WARN-level logs appear.
pass
You can also disable the logger in Python:
from ape.logging import logger, LogLevel
def main():
logger.disable() # Turns off logging entirely.
with logger.disabled():
# Turns off logging in a context - useful for capturing stdout.
pass