log

Log Package

A lightweight, production-ready JSON logger SDK for Go that wraps the industry-standard zap logger with context-aware logging capabilities.

Features

• JSON Formatting: All logs are formatted as JSON for easy parsing and analysis
• Context-Aware: Automatically extracts and includes request IDs, trace IDs, and user IDs from context
• Multiple Log Levels: Support for debug, info, warn, error, and fatal levels
• Production Ready: Built on go.uber.org/zap, one of the fastest and most reliable loggers in Go
• Global & Instance Loggers: Use global functions or create logger instances
• Easy Context Management: Helper functions to add and retrieve values from context

Installation

go get github.com/adityayuga/go-sdk

Usage

Basic Initialization

import (
    "context"
    "github.com/adityayuga/go-sdk/log"
)

// Initialize the global logger
err := log.Init("info") // or "debug", "warn", "error"
if err != nil {
    panic(err)
}
defer log.GetLogger().Sync()

ctx := context.Background()

// Log messages with fields
log.Info(ctx, "user login", "username", "john", "email", "john@example.com")
log.Warn(ctx, "slow query", "duration_ms", 1500)
log.Error(ctx, "database error", "error", err)

Using Context for Trace Information

// Add trace information to context
ctx := context.Background()
ctx = log.WithRequestID(ctx, "req-12345")
ctx = log.WithTraceID(ctx, "trace-67890")
ctx = log.WithUserID(ctx, "user-456")

// These values are automatically included in all logs from this context
log.Info(ctx, "processing payment", "amount", 99.99)
// Output: {"level":"info","ts":"2026-02-13T10:00:00.000Z","msg":"processing payment","request_id":"req-12345","trace_id":"trace-67890","user_id":"user-456","amount":99.99}

Creating Logger Instances

// Create multiple logger instances with different configurations
logger, err := log.NewLogger("debug")
if err != nil {
    panic(err)
}

ctx := context.Background()
logger.Info(ctx, "instance logger message", "key", "value")
logger.Sync()

Available Methods

Global Functions

• log.Info(ctx, message, fields...) - Info level
• log.Warn(ctx, message, fields...) - Warning level
• log.Error(ctx, message, fields...) - Error level
• log.Debug(ctx, message, fields...) - Debug level
• log.Fatal(ctx, message, fields...) - Fatal level (exits process)

Logger Instance Methods

• logger.Info(ctx, message, fields...)
• logger.Warn(ctx, message, fields...)
• logger.Error(ctx, message, fields...)
• logger.Debug(ctx, message, fields...)
• logger.Fatal(ctx, message, fields...)
• logger.Sync() - Flushes buffered logs

Context Helper Functions

• log.WithRequestID(ctx, id) - Add request ID
• log.WithTraceID(ctx, id) - Add trace ID
• log.WithUserID(ctx, id) - Add user ID
• log.WithLogContext(ctx, requestID, traceID, userID) - Add all at once
• log.GetRequestID(ctx) - Retrieve request ID
• log.GetTraceID(ctx) - Retrieve trace ID
• log.GetUserID(ctx) - Retrieve user ID

Development Mode

For development environments with pretty-printed logs:

err := log.InitDevelopment()
if err != nil {
    panic(err)
}

Log Output Example

{
  "level": "error",
  "ts": "2026-02-13T10:00:00.000Z",
  "msg": "failed to fetch user",
  "request_id": "req-12345",
  "trace_id": "trace-67890",
  "user_id": "user-456",
  "user_id": 123,
  "error": "connection timeout"
}

Best Practices

1. Always defer Sync(): Call logger.Sync() before exiting to ensure all logs are flushed
2. Use Context: Pass context through your request handlers to maintain trace information
3. Structured Logging: Use key-value pairs instead of string formatting for better searchability
4. Log Levels: Use appropriate levels (debug for development, info for general, error for problems)
5. Avoid Sensitive Data: Don't log passwords, tokens, or personal information

Performance

Built on zap, this logger is optimized for performance:

• Minimal allocations
• Efficient JSON encoding
• Designed for high-throughput services

License

See LICENSE file in the repository