Duplicate image detector

dupimgs.go

package main

// Duplicate image detector.
//
// Usage:
// ./dupimgs -dir .
//
// This will recursively search the given directory for files with
// {.jpg, .jpeg} extensions (currently hardcoded). For each file with
// such an extension, the MD5 hash of the file's contents will be calculated.
// After traversing all sub-directories and matching files, the paths of files
// with identical hashes will be printed to STDOUT.

import (
	"crypto/md5"
	"flag"
	"fmt"
	"io"
	"io/fs"
	"os"
	"path/filepath"
	"strings"
)

var args = struct {
	dir string
}{
	dir: ".",
}

func main() {
	flag.StringVar(&args.dir, "dir", args.dir, "Directory to scan")
	flag.Parse()

	hashes := make(map[string][]string, 1024)
	h := md5.New()

	ignoredExts := make(map[string]int)

	err := filepath.WalkDir(args.dir, func(path string, d fs.DirEntry, err error) error {
		if err != nil {
			return err
		}
		if d.IsDir() {
			return nil
		}
		ext := strings.ToLower(filepath.Ext(d.Name()))
		switch ext {
		case ".jpg", ".jpeg":
			// OK
		default:
			ignoredExts[ext]++
			return nil
		}

		fd, err := os.Open(path)
		if err != nil {
			return err
		}
		defer fd.Close()

		h.Reset()
		if _, err := io.Copy(h, fd); err != nil {
			return err
		}
		hash := fmt.Sprintf("%x", h.Sum(nil))

		hashes[hash] = append(hashes[hash], path)

		return nil
	})
	if err != nil {
		bail(err)
	}

	fmt.Printf("Inspected %d images\n\n", len(hashes))

	var dups int
	for hash, paths := range hashes {
		if len(paths) < 2 {
			continue
		}
		dups++

		fmt.Printf("%d photos with hash %q:\n", len(paths), hash)
		for _, p := range paths {
			fmt.Printf(" - %s\n", p)
		}
		fmt.Println("===")
	}

	if len(ignoredExts) > 0 {
		fmt.Println("The following file extensions were ignored:")
		for ext, count := range ignoredExts {
			fmt.Printf("- %-15s %d\n", ext, count)
		}
	}

	if dups > 0 {
		fmt.Printf("%d sets of duplicates found\n", dups)
	} else {
		fmt.Println("No duplicates found")
	}
}

func bail(err error) {
	fmt.Fprintf(os.Stderr, "Error: %s\n", err.Error())
	os.Exit(1)
}