standard-log is a powerful and extensible logging library.
There are two ways to use standard-log: global or standalone.
Standalone is the preferred way.
You call createStandardLog()
to get an isolated log system.
import { createStandardLog } from 'standard-log'
const standardLog = createStandarLog(/* options */)
const log = standardLog.getLogger('my-app')
log.error('error message')
log.warn('warn message')
log.info('info message')
// by default debug() will not be printed
// because production mode defaults log level to `info`
log.debug('debug message')
// If the message is consume resource to generage,
// you can use `log.on()` so that it will be called only
// if the level is met.
log.on(logLevels.trace, () => prettify(someValue))
(above is logged with standard-log-color
)
You can configure a logger by doing getLogger(id, options)
:
options.level: number
:
Log level of this logger.
options.writeTo: string | RegExp | ((reporterId: string) => boolean)
:
Only log to specific reporter(s).
standard-log
log level defaults to logLevels.info
.
It comes with many log levels out of the box:
log.emergency('msg')
log.alert('msg')
log.critical('msg')
log.error('msg')
log.warn('msg')
log.notice('msg')
log.info('msg')
log.debug('msg')
log.trace('msg')
log.planck('msg')
When sending logs to console, they are mapped to info
, warn
, error
, and debug
based on severity.
You can also add your own custom levels:
import { createStandardLog } from 'standard-log'
const standardLog = createStandardLog({
customLevels: {
'important': logLevel.warn + 1,
'silly': logLevel.debug + 1
}
})
const log = standardLog.getLogger('custom')
log.important('this is an important message')
log.silly('oh silly')
Besides printing the logs to console, you can use different reporters to save the logs in memory, file, service, or others.
import { createStandardLog, createConsoleLogReporter, createMemoryLogReporter } from 'standard-log'
createStandardLog({
reporters: [createConsoleLogReporter(), createMemoryLogReporter()]
})
Some reporters allow you to format the logs and/or filter them. Using the console log reporter as an example:
import { createConsoleLogReporter } from 'standard-log'
createConsoleLogReporter({
formatter: (entry) => [...],
filter: (entry) => entry.args.every(arg => arg !== 'secret')
})
Here are some additional reporters:
standard-log-file
(TODO)You can temporarily suppress logs by suppressLogs()
.
This allows you to disable downstream logs to reduce log noises.
import { getLogger, suppressLogs } from 'standard-log'
const log = getLogger('some logger')
suppressLogs(() => log.info('not logged'), log)
During test,
you should use createStandardLogForTest
which includes a MemoryReporter
to capture the logs.
import { createStandardLogForTest } from 'standard-log/testing'
test('your test', () => {
const standardLog = createStandardLogForTest()
yourApp.standardLog = standardLog
// do your thing...
const messages = standardLog.reporter.getLogMessage() // or getLogMessageWithLevel()
// validate the message if you want to
})
You can also call getLogger()
to get a logger and log away:
import { getLogger } from 'standard-log'
const log = getLogger('my-lib')
log.info('log away')
It internally creates a global instance of standard-log
.
To configure this global instance, use configGlobal()
:
import { configGlobal } from 'standard-log'
configGlobal({ logLevel: logLevels.info, reporters: [ ... ] })
This global instance does not support custom levels.
And it will emit a warning when configGlobal()
is called more than once.
Typically, configGlobal()
should be called by the application, and it happens only once.
But in micro-app situation, the library can be shared, and each application can call configGlobal()
In general, using this global instance should be avoided. It is the main driving force for 9.0.
Generated using TypeDoc