Minimal parametric JSON file store in Go

fishes.go

package main

import (
	"encoding/json"
	"io"
	"os"
	"path/filepath"
	"sync"
)

type (
	Name    string
	Habitat []string
)

type Fish struct {
	ID         string
	Scientific Name
	English    Name
	Swedish    Name
	Habitat    Habitat
}

func main() {
	if err := run("fishes", os.Stdout); err != nil {
		panic(err)
	}
}

func run(dir string, out io.Writer) error {
	cacheDir, err := os.UserCacheDir()
	if err != nil {
		return err
	}

	path := filepath.Join(cacheDir, dir)

	fishes, err := Store[Fish](path)
	if err != nil {
		return err
	}

	// Data grabbed from https://raw.githubusercontent.com/You-now-Who/dataset/refs/heads/main/List%20of%20fishes%20found%20in%20Sweden/List%20of%20fishes%20found%20in%20Sweden.csv
	fishes.Set("abborre", Fish{
		ID:         "abborre",
		Scientific: "Perca fluviatilis",
		English:    "European perch",
		Swedish:    "Abborre",
		Habitat: Habitat{
			"fresh",
			"brackish",
		},
	})

	fishes.Set("gädda", Fish{
		ID:         "gädda",
		Scientific: "Esox lucius",
		English:    "Pike",
		Swedish:    "Gädda",
		Habitat: Habitat{
			"fresh",
			"brackish",
		},
	})

	fishes.Set("makrill", Fish{
		ID:         "makrill",
		Scientific: "Scomber scombrus",
		English:    "Atlantic mackerel",
		Swedish:    "Makrill",
		Habitat: Habitat{
			"marine",
		},
	})

	a, err := fishes.Get("abborre")
	if err != nil {
		return err
	}

	g, err := fishes.Get("gädda")
	if err != nil {
		return err
	}

	m, err := fishes.Get("makrill")
	if err != nil {
		return err
	}

	return encodeJSON(out, []Fish{a, g, m})
}

type Disk[T any] struct {
	mu   sync.RWMutex
	path string
}

func Store[T any](path string) (*Disk[T], error) {
	perm := os.ModeDir | os.ModePerm

	if err := os.MkdirAll(path, perm); err != nil {
		return nil, err
	}

	return &Disk[T]{path: path}, nil
}

func (d *Disk[T]) Set(id string, v T) error {
	d.mu.Lock()
	defer d.mu.Unlock()

	f, err := os.Create(d.name(id))
	if err != nil {
		return err
	}
	defer f.Close()

	return encodeJSON(f, v)
}

func (d *Disk[T]) Get(id string) (v T, err error) {
	d.mu.RLock()
	defer d.mu.RUnlock()

	f, err := os.Open(d.name(id))
	if err != nil {
		return v, err
	}
	defer f.Close()

	return v, json.NewDecoder(f).Decode(&v)
}

func (d *Disk[T]) name(id string) string {
	return filepath.Join(d.path, id+".json")
}

func encodeJSON(w io.Writer, v any) error {
	enc := json.NewEncoder(w)
	enc.SetIndent("", "  ")

	return enc.Encode(v)
}

And yet another variant with All and Vals methods:

package main

import (
	"encoding/json"
	"io"
	"os"
	"path/filepath"
	"strings"
	"sync"
	"time"
)

type (
	Name    string
	Habitat []string
)

type Fish struct {
	Scientific Name
	English    Name
	Swedish    Name
	Habitat    Habitat
}

func main() {
	if err := run("fishes", os.Stdout); err != nil {
		panic(err)
	}
}

func run(dir string, out io.Writer) error {
	cacheDir, err := os.UserCacheDir()
	if err != nil {
		return err
	}

	path := filepath.Join(cacheDir, dir)

	fishes, err := Store[Fish](path)
	if err != nil {
		return err
	}

	// Data grabbed from https://raw.githubusercontent.com/You-now-Who/dataset/refs/heads/main/List%20of%20fishes%20found%20in%20Sweden/List%20of%20fishes%20found%20in%20Sweden.csv
	fishes.Set("abborre", Fish{
		Scientific: "Perca fluviatilis",
		English:    "European perch",
		Swedish:    "Abborre",
		Habitat: Habitat{
			"fresh",
			"brackish",
		},
	})

	fishes.Set("gädda", Fish{
		Scientific: "Esox lucius",
		English:    "Pike",
		Swedish:    "Gädda",
		Habitat: Habitat{
			"fresh",
			"brackish",
		},
	})

	fishes.Set("makrill", Fish{
		Scientific: "Scomber scombrus",
		English:    "Atlantic mackerel",
		Swedish:    "Makrill",
		Habitat: Habitat{
			"marine",
		},
	})

	list, err := fishes.All()
	if err != nil {
		return err
	}

	return encodeJSON(out, list)
}

type Value[Data any] struct {
	ID        string    `json:"id"`
	CreatedAt time.Time `json:"created_at"`
	Data      Data      `json:"data"`
}

type Disk[Data any] struct {
	mu   sync.RWMutex
	path string
	now  func() time.Time
}

func Store[Data any](path string, options ...func(*Disk[Data])) (*Disk[Data], error) {
	perm := os.ModeDir | os.ModePerm

	if err := os.MkdirAll(path, perm); err != nil {
		return nil, err
	}

	d := &Disk[Data]{
		path: path,
		now:  time.Now,
	}

	for _, o := range options {
		o(d)
	}

	return d, nil
}

func (d *Disk[Data]) Set(id string, data Data) error {
	d.mu.Lock()
	defer d.mu.Unlock()

	f, err := os.Create(d.name(id))
	if err != nil {
		return err
	}
	defer f.Close()

	return encodeJSON(f, Value[Data]{
		ID:        id,
		Data:      data,
		CreatedAt: d.now(),
	})
}

func (d *Disk[Data]) Get(id string) (Data, error) {
	v, err := d.Val(id)
	return v.Data, err
}

func (d *Disk[Data]) All() ([]Data, error) {
	vs, err := d.Vals()
	if err != nil {
		return nil, err
	}

	ds := []Data{}

	for _, v := range vs {
		ds = append(ds, v.Data)
	}

	return ds, nil
}

func (d *Disk[Data]) Vals() ([]Value[Data], error) {
	d.mu.RLock()
	defer d.mu.RUnlock()

	vs := []Value[Data]{}

	des, err := os.ReadDir(d.path)
	if err != nil {
		return nil, err
	}

	for _, de := range des {
		if v, err := d.Val(de.Name()); err == nil {
			vs = append(vs, v)
		}
	}

	return vs, nil
}

func (d *Disk[Data]) Val(id string) (Value[Data], error) {
	d.mu.RLock()
	defer d.mu.RUnlock()

	var v Value[Data]

	f, err := os.Open(d.name(id))
	if err != nil {
		return v, err
	}
	defer f.Close()

	return v, json.NewDecoder(f).Decode(&v)
}

func (d *Disk[Data]) name(id string) string {
	ext := ".json"

	return filepath.Join(d.path, strings.TrimSuffix(id, ext)+ext)
}

func encodeJSON(w io.Writer, v any) error {
	enc := json.NewEncoder(w)
	enc.SetIndent("", "  ")

	return enc.Encode(v)
}

Go Playground: https://go.dev/play/p/b1eteN_FBh5

Now with support for All and Vals methods.

I asked deepseek-r1:14b to describe what the latest version does:

The provided Go program is designed to store and retrieve fish data locally in a 
structured manner. Here's a step-by-step explanation:

1. **Setup and Data Structure**:
   - The program defines a `Fish` struct to hold information about different fish 
species, including their scientific, English, and Swedish names, along with their 
habitats.

2. **Storage Mechanism**:
   - It uses a key-value store approach where each fish entry is stored as a 
separate JSON file in a cache directory (`$HOME/.cache/fishes` by default).
   - Each entry is uniquely identified by an ID (e.g., "abborre").

3. **Storing Fish Data**:
   - The program adds three fish entries to the store using their respective IDs and 
detailed information.

4. **Retrieving All Entries**:
   - After storing the entries, it retrieves all stored fish data, which includes 
reading each JSON file in the cache directory.

5. **Output**:
   - Finally, it prints all retrieved fish data as a single JSON array to standard 
output.

This program is useful for applications needing local persistence of fish data, 
demonstrating how to use Go's file system and JSON encoding/decoding capabilities 
effectively.