Stuber

High-performance gRPC stub server with advanced matching capabilities.

Performance Results

V1 (Old) vs V2 (New) Comparison

Benchmark	V1 (Old)	V2 (New)	Improvement
Found	1,568 ns/op	294.1 ns/op	5.3x faster
NotFound	70.0 ns/op	61.0 ns/op	15% faster
Multiple Stubs	15,635 ns/op	2,022 ns/op	7.7x faster

Memory Usage

Benchmark	V1 (Old) Memory	V2 (New) Memory	Improvement
Found	608 B/op	136 B/op	78% less
Multiple Stubs	7,959 B/op	616 B/op	92% less
Stream	160 B/op	136 B/op	15% less

Allocations

Benchmark	V1 (Old) Allocs	V2 (New) Allocs	Improvement
Found	27 allocs/op	7 allocs/op	74% less
Multiple Stubs	262 allocs/op	12 allocs/op	95% less
Stream	8 allocs/op	7 allocs/op	13% less

Parallel Processing Results

Stub Count	Sequential	Parallel	Improvement
50	7,023 ns/op	7,060 ns/op	1% slower
100	13,967 ns/op	15,604 ns/op	12% slower
200	30,658 ns/op	26,657 ns/op	13% faster
500	72,135 ns/op	53,036 ns/op	26% faster

Note: Parallel processing activates at 100+ stubs. Below 200 stubs, sequential is faster due to overhead.

Key Features

• Exact matching with field-level precision
• Array order ignoring with ignoreArrayOrder: true
• Regex pattern matching with caching
• Priority-based selection with specificity scoring
• Stream support: Unary, Server, Client, Bidirectional
• LRU caching for string hashes and regex patterns
• Memory pools for zero-allocation operations

Usage

package main

import (
    "github.com/gripmock/stuber"
    "github.com/bavix/features"
)

func main() {
    toggles := features.New()
    stuber := stuber.NewBudgerigar(toggles)
    
    stub := &stuber.Stub{
        Service: "example.service",
        Method:  "ExampleMethod",
        Input: stuber.InputData{
            Equals: map[string]any{"id": "123"},
        },
        Output: stuber.Output{
            Data: map[string]any{"result": "success"},
        },
    }
    
    stuber.PutMany(stub)
}

API Reference

type Stub struct {
    ID       uuid.UUID
    Service  string
    Method   string
    Priority int
    Input    InputData
    Output   Output
    Stream   []InputData
}

type InputData struct {
    Equals          map[string]any
    Contains        map[string]any
    Matches         map[string]any
    IgnoreArrayOrder bool
}

Testing

go test -v
go test -bench=. -benchmem

Performance Notes

• V2 (New): Use for best performance in most cases
• Parallel processing: Only beneficial for 200+ stubs
• Memory: V2 (New) significantly more memory efficient
• Allocations: V2 (New) reduces allocations by 74-95%

License

MIT License