logsaw

logsaw is a simple JSON file-based logger for node.

Installation

npm

npm install logsaw

GitHub

npm install https://github.com/martinrue/logsaw/tarball/master

Configuration

To change the default config, call configure:

var logger = require('logsaw');
 
logger.configure({
  console: true,
  verbose: true,
  directory: __dirname + '/logs'
});

Options

Key	Default Value	Description
console	true if not production	display logs on console
verbose	false	display complete log object on console
debug	false	save debug logs to log file
generators	false	allow log functions to be yielded from ES6 generator
directory	process.cwd() + 'logs'	the directory to write log files to (created if new)

API

logsaw exposes 4 functions: debug(), info(), warn() and error(). All functions take a single object argument which may contain message, context and data fields. You may specify a type field in calls to info() to create custom log types. Calls to debug() are saved to the log file only if { debug: true } is set in config.

Example

var logger = require('logsaw');
 
// creates a debug log (saved to the log file if debug is set to true in config)
logger.debug({ message: 'some debug message', data: { /*...*/ } });
 
// creates an info log
logger.info({ message: 'user signed up' });
 
// creates a custom user-signup log
logger.info({ type: 'user-signup', message: 'user signed up' });
 
// creates a warning log with a context
logger.warn({ message: 'account locked', context: '[e-mail]' });
 
// creates an error log with associated err data
logger.error({ message: 'something went wrong', data: err });

Callback

If { generators: false } is set (default), an optional callback can be supplied as the last argument to the 4 logging functions to be notified once the underlying write operation completes.

Generators

If { generators: true } is set, you can optionally yield any of the 4 log functions from within Co, or similar.

Logs

Filename

Log files are created based on the current date (e.g. 2014-07-22.log) and written to the config directory.

Format

Each log file consists of a line-delimited JSON entry of the form:

{
  "type":"error",
  "message":"something went wrong",
  "time":"2014-07-22T19:30:00.630Z",
  "data":"{ err: 'error info' }",
  "context":"[e-mail]"
}

Context

The context field of a log entry is intended to hold a scalar value that can be used to group sets of logs together.

You could set it to a user's email address to associate a series of logs with a user, you could set it to a request ID to associate a series of logs with an HTTP request, or you could set it to some other meaningful value.

License

MIT