StriveLogger 1.1.0

A flexible logging abstraction layer

Strive Logger

What is the Strive Logger?

The Strive Logger is a logging abstraction layer which is used in our company's python projects. We find that it works well for us, perhaps it will also work for you.

How does it work?

The package provides a class named "StriveLogger" which has class methods for "debug", "info", "warn" and "error" level logging. A dictionary of extra information to be logged can optionally be supplied to all methods, and an exception or error string can also optionally be provided to the error method.

Where does it log?

The actual logging is handled by a logger implementation which must be supplied once during your application startup, by calling the "initialize" method.

from strivelogger import StriveLogger, ConsoleLogger


StriveLogger.initialize(logger=ConsoleLogger())

Logger Implementations

You may choose to create your own logger implementation, for example to wrap up a structured logger like Google's StackDriver, but some simple implementations have been provided as part of this project as well.

To create your own implementation, create a class which inherits from LoggerImplementation and implement the required methods:

from strivelogger.logger_implementation import LoggerImplementation

class MyLogger(LoggerImplementation):
    def debug(self, message: str, trace_id: Optional[str], extra: Dict = None) -> None:
        # TODO: log somewhere
        pass

    def info(cls, message: str, trace_id: Optional[str], extra: Dict = None) -> None:
        # TODO: log somewhere
        pass

    def warn(cls, message: str, trace_id: Optional[str], extra: Dict = None) -> None:
        # TODO: log somewhere
        pass

    def error(
        cls,
        message: str,
        trace_id: Optional[str],
        exc_info: Optional[Union[BaseException, str]],
        extra: Optional[Dict],
    ) -> None:
        # TODO: log somewhere
        pass

Provided Implementations

ConsoleLogger

from strivelogger.logger_implementations.console_logger import ConsoleLogger

Prints all log messages to the console.

NullLogger

from strivelogger.logger_implementations.null_logger import NullLogger

Throws away all log requests. This implementation is used prior to initialization in order to ensure that the StriveLogger doesn't throw null pointer exceptions if not initialized.

PythonLogger

from strivelogger.logger_implementations.python_logger import PythonLogger

Wraps an instance of a python logger. You can supply a name in order to create or use an existing instance, or simply use the root logger. This logger can log simple strings, or it can log as JSON.

UvicornLogger

from strivelogger.logger_implementations.uvicorn_logger import UvicornLogger

A specialized implementation of the PythonLogger - this logger will attach to the python logger instance which uvicorn creates. If JSON logging is enabled, this logger will also attach and reformat uvicorn.access logging calls.

Useful if using the StriveLogger in a FastAPI project.

Tracing

The StriveLogger can also supply trace ids to log entries, if a TracerImplementation is supplied during initialization.

No tracer implementations are included in this library currently - you may implement your own by inheriting from TracerImplementation and implementing the "get_trace_id" method:

from strivelogger.tracer_implementation import TracerImplementation


class MyTracerImplementation(TracerImplementation):
    def get_trace_id(self) -> str:
        # TODO: find or set a trace ID and return it
        raise NotImplementedError()

Example usage

from strivelogger import StriveLogger, UvicornLogger


StriveLogger.initialize(logger=UvicornLogger(level="DEBUG", enable_json=True))
StriveLogger.info("Logging initialized", extra={"timestamp": datetime.now().isoformat()})

Contributing

This is an open source project, and thus contributions to this project are welcome - please feel free to create a new issue if you encounter any problems, or submit a pull request.

If submitting a pull request, please ensure the following standards are met:

1. Code files must be formatted (run ./lint.sh prior to submitting)

2. Tests must pass (run ./test.sh). New test cases to validate your changes are highly recommended.

3. Implementations must not add any project dependencies. If you wish to create a logger or tracer implementation which depends on another package, consider creating a new library which references this one, and submitting a PR which documents your new implementation here.

Additional information

This package has no dependencies on any other packages.

Developed by:

© 2022 Strive Business Solutions