Meadow.Logging
Many good logging libraries already exist in the .NET development community, so why Meadow.Logging
?
The primary driver is that Meadow
is designed to run on microcontrollers, where storage and memory resources are limited. Most existing libraries provide a wide array of capabilties and features, but at the cost of size. Meadow.Logging
is intended to be very, very lean, only exposing the minimal set of what might be needed by a typical Meadow application, but still completely usable in larger, non-Meadow application scenarios.
Contents
Logger
The Logger
class has two functions:
- Providing the methods an Application would typically call in order to add a log entry.
- Holding a collection of
ILogProvider
instances that will react to those methods.
ILogProvider
The ILogProvider
interface provides the methods that the Logger
will call when an Application makes a specific Logging call that meets the currently set Loglevel.
Two common implementations are inluded in the compiled library:
ConsoleLogger
outputs log information toSystem.Console
DebugLogger
outputs log information toSystem.Diagnostics.Debug
Since a primary goal of Meadow.Logging
is to be small, some additional implementations (e.g. UDP broadcast logging or SQLIte logging) will likely be provided in source form until linking is fully supported by Meadow.
Usage
The general usage pattern is:
- Create a
Logger
instance - Register one or more
ILogProvider
implementations - Access the logger instance using a static, dependency injection, or whatever your preferred mechanism
Basic Usage
A simple example of Console
output would look like this:
static void Main()
{
var logger = new Logger(new ConsoleLogProvider());
var App = new App(logger);
app.DoStuff();
}
class App
{
Logger Logger { get; init; }
App(Logger logger)
{
Logger = logger;
}
public void DoStuff()
{
while(true)
{
Logger.Debug("tick");
Thread.Sleep(1000);
}
}
}
ILogProvider
s
Using Multiple The Logger
allows you to add 0:N ILogProvider
instances, and all of them get called when ap application makes a logging call. For example, your application may want to output to both the Console and a UDP broadcast. You can achieve this by simply registering multiple providers:
var logger = new Logger();
logger.AddProvider(new ConsoleLogProvider());
logger.AddProvider(new UdpLogProvider());
...
logger.Info("This message will get delivered to all Providers");
Conditional Output Using LogLevel
Logger-wide output can be controlled using the LogLevel
property. Log output will occur for any level at or above the currently set level.
var logger = new Logger(new ConsoleLogProvider());
...
logger.Loglevel = LogLevel.Debug;
logger.Debug("This debug message will get delivered to all Providers");
logger.Warn("This warning message will get delivered to all Providers");
...
logger.Loglevel = LogLevel.Info;
logger.Debug("This debug message will *not* get delivered to any Providers");
logger.Warn("This warning message will still get delivered to all Providers");
Conditional Output Using Runtime Checks
Occasionally you may want to keep a given log level, but output information based on a runtime check. All of the logging outputs support this as well.
var logger = new Logger(new ConsoleLogProvider());
bool logCondition = true;
...
logger.Debug("This debug message will get delivered to all Providers");
logger.Debug(logCondition, "This debug message will also get delivered to all Providers");
bool logCondition = false;
logger.Debug("This debug message will also get delivered to all Providers");
logger.Debug(logCondition, "This debug message will *not* get delivered to any Providers");