go-annotate

High-Performance Go Code Instrumentation

---

🎯 About

go-annotate is a high-performance Go source code instrumentation tool that automatically transforms Go programs by injecting logging calls around function entries and exits. It reports function parameters and return values, enabling external tools to monitor runtime behavior with minimal overhead.

---

✨ Features

• AST-based instrumentation: Safe and precise Go source code transformation
• Multiple output formats: JSON, CBOR, text, and debug formats
• Socket and file logging: Real-time network streaming or local file output
• Memory optimized: Object pools and buffer reuse for ~60% allocation reduction
• Non-blocking design: Prevents deadlocks in concurrent applications
• Selective instrumentation: Target exported functions only or use custom filters
• Academic integration: Generate monitoring rules for SpecMon

---

📦 Installation

Prerequisites

• Go 1.22+
• Git

Steps

1. Install from Go modules:

   go install github.com/specmon/go-annotate@latest

2. Or build from source:

   git clone https://github.com/specmon/go-annotate.git
   cd go-annotate
   go build -o go-annotate

---

🚀 Usage

Basic Instrumentation

1. Instrument your Go code:

   go-annotate -import "github.com/specmon/go-annotate/log" -w main.go

2. Configure logging:

   export GO_ANNOTATE_LOG_TARGET="/path/to/output.log"  # File mode
   # OR
   export GO_ANNOTATE_LOG_TARGET="localhost:8080"       # Socket mode
   export GO_ANNOTATE_LOG_FORMAT="json"                 # json, cbor, text, debug

3. Run your instrumented program:

   go run main.go

Configuration

• GO_ANNOTATE_LOG_TARGET - Required. Log destination (auto-detects mode):
  • File: /path/to/logfile.log
  • TCP Socket: localhost:8080
  • Unix Socket: /tmp/socket.sock
• GO_ANNOTATE_LOG_FORMAT - Log format: json (default), cbor, text, debug

Command Line Options

go-annotate [options] <source-files>

Options:
  -import string     Import path for the log package (required)
  -w                 Write changes back to source files (default: print to stdout)
  -exported          Only instrument exported functions
  -package           Include package name prefix in function calls
  -returns           Show function return values
  -timing            Include timing information (implies -returns)
  -generate string   Generate monitoring rules file

---

📊 Examples

Basic Function Tracing

# Instrument main.go with function entry/exit logging
go-annotate -import "github.com/specmon/go-annotate/log" -w main.go

# Run with file logging
export GO_ANNOTATE_LOG_TARGET="trace.log"
export GO_ANNOTATE_LOG_FORMAT="json"
go run main.go

Real-time Network Streaming

# Terminal 1: Start log receiver
go run test/test_socket_server.go

# Terminal 2: Run instrumented program
export GO_ANNOTATE_LOG_TARGET="localhost:8080"
export GO_ANNOTATE_LOG_FORMAT="json"
go run main.go

SpecMon Monitoring Rules

# Generate monitoring rules
go-annotate -import "github.com/specmon/go-annotate/log" \
        -generate "rules.thy" \
        -w main.go

Output Formats

JSON Format

{
  "time": 1704067200000000000,
  "event": {
    "name": "pair",
    "type": "function",
    "args": [
      {"name": "main_Add_Enter", "type": "function", "args": [...]},
      {"name": "pair", "type": "function", "args": [...]}
    ]
  }
}

Text Format

main_Add_Enter(1, 5, 10)
main_Add_Leave(1, 5, 10) = (15)

CBOR Format

Binary format optimized for performance and network transmission.

---

📈 Performance

Recent optimizations deliver significant performance improvements:

• ~60% allocation reduction through object pooling
• Non-blocking channels prevent application deadlocks
• Optimized serialization with buffer reuse
• Smart buffering handles network delays gracefully

---

🏗️ Architecture

Core Components

• AST Parser: Safe Go source code transformation
• Logger: High-performance event collection with multiple outputs
• Network Layer: Robust socket handling with reconnection
• Memory Management: Pool-based allocation for hot paths

Use Cases

• Performance Profiling: Trace function calls and timing
• Security Analysis: Generate traces to detect suspicious function invocations
• Runtime Monitoring: Generate execution traces for external validation
• Debugging: Detailed program flow analysis
• Academic Research: Generate traces for formal verification and analysis tools

---

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

---

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

📚 Citation

If you use go-annotate in academic research, please cite:

@software{go-annotate,
  title = {go-annotate: High-Performance Go Code Instrumentation},
  author = {Morio, Kevin},
  year = {2025},
  url = {https://github.com/specmon/go-annotate}
}