Finally, a full featured Javascript logging framework for the both browser and Node.

• Works in the browser and Node.
• Namespace support for fine-grained logging support at a module, feature, or method level.
• Log level support to enable prioritized log output.
• Dynamic configuration of log level mappings based on namespaces patterns.
• Pluggable middleware transform of log output (e.g. prefixing, colors, custom, etc.)
• Pluggable writers of log output (e.g. console, null, server-side, custom, etc.)
• Built in Typescript, compatible with Javascript.

Install Scribe using npm:

npm install --save @nascentdigital/scribe

Or yarn:

yarn add @nascentdigital/scribe

Getting started is really easy. Simply acquire a Log and start logging:

import {Scribe} from "@nascentdigtal/scribe";

Scribe.log.debug("A debug message");
// A debug message

Scribe.log.error("An error occurred - ", new Error("Oops!"));
// An error occurred - Error: Oops!
//   at ...

This example uses the global log to output a debug and error level log message. While this is a quick way to access and output log messages, the recommended approach is to use namespaces.

In modern Javascript we organize our code into modules which build up larger libraries and applications. Good logging should reflect the separation of concerns modelled by the application architecture.

Scribe uses namespaces to help developers group and control logging based on related functionality. The structure of a namespace is as follows:

Some examples of valid namespaces would be:

• "my-app"
• "my-app:mymodule"
• "my-app:mymodule/foo"
• "my-app:mymodule/foo/bar"
• "my-app:mymodule/foo/bar/variant_1"

A log can be obtained by accessing the global Scribe.log property:

const globalLog = Scribe.log;

Generally you'll want to fetch a log that is dedicated to a specific scope. You can obtain a specific logger via the Scribe.getLog() method. This method creates a log if it doesn't exist or fetches the cached log if it has already been created.

const logA = Scribe.getLog("moduleA");
const logB = Scribe.getLog("moduleB");
const logA2 = Scribe.getLog("moduleA");

logA.debug(logA === logA2); // true

In practice, you'll want to dedicate a log per module and scope the namespace to reflect that relationship.

const log = Scribe.getLog("mylib:foo");

export function foo(options) {

    if (!options) {
        log.error("foo() invoked without options");
        throw new Error("Missing 'options' argument.");
    }

    log.trace("calling foo with options: ", options);
}

const log = Scribe.getLog("mylib:bar");

export function bar(message) {
    log.debug("bar() was invoked with message: ", message);
}

Log levels provide a way of prioritizing the significance of a log message. The Log interface has log methods the correspond communicate the following intent around log messages:

The levels are cumulative in the sense that they stack. When trace level logging is enabled all other log level are automatically enabled as well, while debug enables all levels except trace. Outside of silent, the error level is the most strict level and will result in the other messages being suppressed.

Here's an example of how these

Most of the time you'll want to control which of the log levels actually get output to the console. The Scribe.setLogLevel() method give precise control of individual log levels by targeting namespaces. This method uses the * wildcard to help create glob expressions that can cover many different scenarios.

Filters can be set before or after a log is created, and is guaranteed to take precedence based on the most recently set filter that matches a namespace. The default level for logs is error if there is no overridden match.

// create a log early
const loginLog = Scribe.getLog("myapp:onboarding/login");

// change default level to info, make onboarding debug
Scribe.setLogLevel("*", "info");
Scribe.setLogLevel("myapp:onboarding*", "debug");

// show some logging
loginLog.trace("suppressed");
loginLog.debug("OUTPUT")
loginLog.error("OUTPUT");

// late binding of another log
const homeLog = Scribe.getLog("myapp:home");

// use the new log
homeLog.debug("suppressed");
homeLog.warn("OUTPUT");
homeLog.error("OUTPUT");

While debugging in the browser you might want to dynamically set the LogLevel to see what's going on. Scribe exposes itself through the globally in the browser through the window.nascentdigital namespace. This allows updates at any time:

> window.nascentdigital.Scribe.setLogLevel("*", "trace");