Package Logger

This is the package for logging (based on zap).

Usage

// Initialise a logger first
logger.InitLogger()
// Use sugared logger
logger.Infof("Message: %s", msg)
// Don't forget to sync your logs when you log to file
defer logger.SyncAll()

Configuration

log:
  # development mode is more human-friendly: console encoding and not log to file
  development: true
  # THE FOLLOWING CONFIGS ONLY WORKS IN THE PRODUCTION MODE (development: false)
  # console or json
  encoding: console
  info:
    # whether log to stdout or not
    toStdout: true
    # whether log to file or not
    toFile: true
    # path and name of the INFO-level log file
    fileName: logs/core-info.log
    # maximum size of a single INFO-level log file (megabytes)
    maxSize: 10
    # how many days old INFO-level log files to retain
    maxAge: 7
    # how much old INFO-level log files to retain
    maxBackups: 10
  error:
    # whether log to stderr or not
    toStderr: true
    # whether log to file or not
    toFile: true
    # path and name of the ERROR-level log file
    fileName: logs/core-error.log
    # maximum size of a single ERROR-level log file (megabytes)
    maxSize: 10
    # how many days old ERROR-level log files to retain
    maxAge: 7
    # how much old ERROR-log files to retain
    maxBackups: 10

Supported log levels

From Log levels in Zap:

• DEBUG (-1):
  • Usage: logger.Debug() or zap.L().Debug().
  • For recording messages useful for debugging.
• INFO (0):
  • Usage: logger.Info() or zap.L().Info().
  • For messages describing normal application operations.
• WARN (1):
  • Usage: logger.Warn() or zap.L().Warn().
  • For recording messages indicating something unusual happened that may need attention before it escalates to a more severe issue.
• ERROR (2):
  • Usage: logger.Error() or zap.L().Error()
  • For recording unexpected error conditions in the program.
• DPANIC (3):
  • Usage: logger.DPanic() or zap.L().DPanic().
  • For recording severe error conditions in development. It behaves like PANIC in development and ERROR in production.
• PANIC (4):
  • Usage: logger.Panic() or zap.L().Panic().
  • Calls panic() after logging an error condition.
• FATAL (5):
  • Usage: logger.Fatal() or zap.L().Fatal().
  • Calls os.Exit(1) after logging an error condition.

Sugared logger

Sugared logger is more human-friendly and have a syntax similar to fmt.Print() and fmt.Printf().

// Use sugared logger
logger.Info("This is sugared logger!!!")
logger.Infof("Message: %s", msg)
logger.Errorf("Failed to do bar: %s", err.Error())

Supported formats:

• Print()-like (e.g., Info())
• Printf()-like (e.g., Infof())
• Println()-like (e.g., Infoln())
• Non-sugared logger style (use key-value pairs, e.g., Infow())

Non-sugared logger

Non-sugared logger has a better performance and is more suitable for performance sensitive scenario.

// Use non-sugared logger
zap.L().Info("This is message", zap.String("key", "value"))
zap.L().Error("Failed to do foo", zap.String("error", err.Error()))

Log format

The following two formats are supported:

• json (default): Default format that is better for parsing and processing. Example: {"level":"info","timestamp":"2024-03-15T12:29:34.657+1100","msg":"This is test logger: INFO level"}
• console: A more human-friendly format. Example: 2024-03-15T12:30:13.390+1100 INFO This is test logger: INFO level