@tv2media/logger

MediaTech Logger

The framework aims to facilitate the logging needs of the developers in TV2 Media Technology, as well as creating an uniform logging strategy across Typescript and Javascript projects.

Install & build

The package can be installed by:

$ yarn add @tv2media/logger

To build from source:

$ git clone https://github.com/tv2/mediatech-logger.git
$ cd mediatech-logger
$ yarn && yarn build

Usage - Simple

import { createDefaultLogger } from '@tv2media/logger'

const logger = createDefaultLogger()

logger.info('Server started.')
logger.error('Request failed.')
logger.data(new Error('Some dangerous error!')).error('Request failed.')

Usage - Advanced

import {
  Logger,              // The base class Logger for custom configuration.
  createDefaultLogger, // Returns an instance of one of the following loggers based upon NODE_ENV.
  ProductionLogger,    // Logger used in production.
  StagingLogger,       // Logger used in staging.
  DevelopmentLogger,   // Logger used in development.
  LocalLogger,         // Logger used in local development.
  
  Level,               // The severity of the 
  Format,              // Formatting type of the log
  Vault,               // Where to store logs
} from '@tv2media/logger'

const logger = new Logger({
  level: Level.Info,                // .Error | .Warn | .Info | .Debug | .Trace
  format: {
    kind: Format.Custom,            // .Plaintext | .JSON | .Custom
    format: (log, options) => {     // Only used for .Custom, and is custom format.
      let out = '[' + log.level + ']'
      if (options.timestamp) {
        out += '[' + new Date().toString() + ']'
      }
      return out + ' ' + data.message
    },
    timestamp: true,                // Whether or not to include timestamp.
    depth: 3n,                      // Depth to traverse in objects. Default is -1n (full depth).
  },
  vault: {                          // Where to store logs
    kind: Vault.Console,            // .Console
  },
})

logger.data('some-data')                 // Adds the key-value pair { "data": "some-data" } to a new log context.
logger.tag('some-tag')                   // Adds the key-value pair { "tag": "some-tag" } to a new log context.
logger.error('Server failed.')           // Stores a log context with severity level of 'error'.
logger.warn('No response from client.')  // Stores a log context with severity level of 'warn'.
logger.info('Server started at ip:port') // Stores a log context with severity level of 'info'.
logger.debug({ ip: '0.0.0.0' })          // Stores a log context with severity level of 'debug'.
logger.trace('some trace here')          // Stores a log context with severity level of 'trace'.

logger.tag('testing').info('test message')
logger.info('message', { tag: 'testing', otherMeta: 'meta' }) // Each of the severity level methods takes an optional argument, with extra attributes for the log context.

Environment Variables

NODE_ENV

The createDefaultLogger is using the environment variable NODE_ENV to determine which log level and format which will be used. The current setup is the following.

NODE_ENV=production            # fomat = JSON, log level = warn
NODE_ENV=stage | staging       # format = JSON, log level = info
NODE_ENV=develop | development # fomat = PLAINTEXT, log level = debug
NODE_ENV=local                 # fomat = PLAINTEXT, log level = trace
NODE_ENV="any other value"     # fomat = PLAINTEXT, log level = trace

LOG_LEVEL

Setting the environment variable LOG_LEVEL overrides the log level from the NODE_ENV setup, this can come in handy when you need to enable e.g. debugging logs in a production enviorment.

function getLevel(): Level | undefined {
    switch (process.env.LOG_LEVEL?.toLowerCase()) {
        case 'error': return Level.Error
        case 'warn': return Level.Warn
        case 'info': return Level.Info
        case 'debug': return Level.Debug
        case 'trace': return Level.Trace
        default: return undefined
    }
}