bench-memory-alloc/

Memory Allocation Benchmarks

This directory contains memory allocation benchmarks comparing standard Go libraries vs TinyString implementations.

Structure

memory-bench/
├── standard/           # Standard library implementation
│   ├── main.go        # Main program with standard library operations  
│   ├── main_test.go   # Benchmark tests for standard library
│   └── go.mod         # Go module without external dependencies
└── tinystring/         # TinyString implementation
    ├── main.go        # Main program with TinyString operations
    ├── main_test.go   # Benchmark tests for TinyString (including pointer optimization)
    └── go.mod         # Go module with TinyString dependency

Benchmarks Included

String Processing

• Standard: Uses strings.ToLower, strings.ReplaceAll, strings.Fields, etc.
• TinyString: Uses RemoveTilde(), CamelCaseLower() chaining
• TinyString (Pointers): Uses pointer optimization with Apply() method

Number Processing

• Standard: Uses fmt.Sprintf, strings.Split for number formatting
• TinyString: Uses RoundDecimals(), FormatNumber() chaining

Mixed Operations

• Standard: Combines string and number operations using standard library
• TinyString: Uses unified TinyString API for all data types

Running Benchmarks

# Run all memory benchmarks and update README
../scripts/memory-benchmark.sh

# Run only standard library benchmarks
cd standard && go test -bench=. -benchmem

# Run only TinyString benchmarks  
cd tinystring && go test -bench=. -benchmem

# Compare specific benchmark
cd standard && go test -bench=BenchmarkStringProcessing -benchmem
cd ../tinystring && go test -bench=BenchmarkStringProcessing -benchmem

Benchmark Output Format

Go benchmark output format:

BenchmarkName-N    iterations   ns/op   bytes/op   allocs/op

Where:

• iterations: Number of benchmark iterations
• ns/op: Nanoseconds per operation
• bytes/op: Bytes allocated per operation
• allocs/op: Number of heap allocations per operation

Integration

The memory benchmark tool (../memory-tool/main.go) automatically:

1. Runs benchmarks in both directories
2. Parses benchmark output
3. Calculates improvement percentages
4. Updates the main README.md with real data

Expected Results

TinyString typically shows:

• 20-35% less memory allocation per operation
• 30-45% fewer heap allocations per operation
• Similar or better execution time performance
• Significant improvement with pointer optimization

These improvements come from:

• Manual implementations avoiding standard library overhead
• Pointer optimization reducing string copies
• Chainable API reducing intermediate allocations
• Unified type conversion system