Merge pull request #29 from thushan/report-output

Refactor App Run and tweaking reporting

go.mod

@@ -6,8 +6,8 @@ require (
 	github.com/alphadose/haxmap v1.3.0
 	github.com/cespare/xxhash v1.1.0
 	github.com/dustin/go-humanize v1.0.1
-	github.com/pterm/pterm v0.12.70
-	github.com/spaolacci/murmur3 v0.0.0-20180118202830-f09979ecbc72
+	github.com/pterm/pterm v0.12.71
+	github.com/spaolacci/murmur3 v1.1.0
 	github.com/spf13/cobra v1.8.0
 	github.com/thediveo/enumflag/v2 v2.0.5
 	golang.org/x/term v0.14.0
@@ -26,7 +26,7 @@ require (
 	github.com/rivo/uniseg v0.4.4 // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
 	github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e // indirect
-	golang.org/x/exp v0.0.0-20231006140011-7918f672742d // indirect
+	golang.org/x/exp v0.0.0-20231110203233-9a3e6036ecaa // indirect
 	golang.org/x/sys v0.14.0 // indirect
 	golang.org/x/text v0.14.0 // indirect
 )

go.sum

@@ -70,8 +70,8 @@ github.com/pterm/pterm v0.12.31/go.mod h1:32ZAWZVXD7ZfG0s8qqHXePte42kdz8ECtRyEej
 github.com/pterm/pterm v0.12.33/go.mod h1:x+h2uL+n7CP/rel9+bImHD5lF3nM9vJj80k9ybiiTTE=
 github.com/pterm/pterm v0.12.36/go.mod h1:NjiL09hFhT/vWjQHSj1athJpx6H8cjpHXNAK5bUw8T8=
 github.com/pterm/pterm v0.12.40/go.mod h1:ffwPLwlbXxP+rxT0GsgDTzS3y3rmpAO1NMjUkGTYf8s=
-github.com/pterm/pterm v0.12.70 h1:8W0oBICz0xXvUeB8v9Pcfr2wNtsm7zfSb+FJzIbFB5w=
-github.com/pterm/pterm v0.12.70/go.mod h1:SUAcoZjRt+yjPWlWba+/Fd8zJJ2lSXBQWf0Z0HbFiIQ=
+github.com/pterm/pterm v0.12.71 h1:KcEJ98EiVCbzDkFbktJ2gMlr4pn8IzyGb9bwK6ffkuA=
+github.com/pterm/pterm v0.12.71/go.mod h1:SUAcoZjRt+yjPWlWba+/Fd8zJJ2lSXBQWf0Z0HbFiIQ=
 github.com/rivo/uniseg v0.2.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
 github.com/rivo/uniseg v0.4.4 h1:8TfxU8dW6PdqD27gjM8MVNuicgxIjxpm4K7x4jp8sis=
 github.com/rivo/uniseg v0.4.4/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
@@ -80,8 +80,9 @@ github.com/sergi/go-diff v1.2.0 h1:XU+rvMAioB0UC3q1MFrIQy4Vo5/4VsRDQQXHsEya6xQ=
 github.com/sergi/go-diff v1.2.0/go.mod h1:STckp+ISIX8hZLjrqAeVduY0gWCT9IjLuqbuNXdaHfM=
 github.com/sirupsen/logrus v1.9.3 h1:dueUQJ1C2q9oE3F7wvmSGAaVtTmUizReu6fjN8uqzbQ=
 github.com/sirupsen/logrus v1.9.3/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
-github.com/spaolacci/murmur3 v0.0.0-20180118202830-f09979ecbc72 h1:qLC7fQah7D6K1B0ujays3HV9gkFtllcxhzImRR7ArPQ=
 github.com/spaolacci/murmur3 v0.0.0-20180118202830-f09979ecbc72/go.mod h1:JwIasOWyU6f++ZhiEuf87xNszmSA2myDM2Kzu9HwQUA=
+github.com/spaolacci/murmur3 v1.1.0 h1:7c1g84S4BPRrfL5Xrdp6fOJ206sU9y293DDHaoy0bLI=
+github.com/spaolacci/murmur3 v1.1.0/go.mod h1:JwIasOWyU6f++ZhiEuf87xNszmSA2myDM2Kzu9HwQUA=
 github.com/spf13/cobra v1.8.0 h1:7aJaZx1B85qltLMc546zn58BxxfZdR/W22ej9CFoEf0=
 github.com/spf13/cobra v1.8.0/go.mod h1:WXLWApfZ71AjXPya3WOlMsY9yMs7YeiHhFVlvLyhcho=
 github.com/spf13/pflag v1.0.5 h1:iy+VFUOCP1a+8yFto/drg2CJ5u0yRoB7fZw3DKv/JXA=
@@ -102,8 +103,8 @@ github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e/go.mod h1:RbqR21r5mrJu
 github.com/yuin/goldmark v1.4.13/go.mod h1:6yULJ656Px+3vBD8DxQVa3kxgyrAnzto9xy5taEt/CY=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20210921155107-089bfa567519/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
-golang.org/x/exp v0.0.0-20231006140011-7918f672742d h1:jtJma62tbqLibJ5sFQz8bKtEM8rJBtfilJ2qTU199MI=
-golang.org/x/exp v0.0.0-20231006140011-7918f672742d/go.mod h1:ldy0pHrwJyGW56pPQzzkH36rKxoZW1tw7ZJpeKx+hdo=
+golang.org/x/exp v0.0.0-20231110203233-9a3e6036ecaa h1:FRnLl4eNAQl8hwxVVC17teOw8kdjVDVAiFMtgUdTSRQ=
+golang.org/x/exp v0.0.0-20231110203233-9a3e6036ecaa/go.mod h1:zk2irFbV9DP96SEBUUAy67IdHUaZuSnrz1n472HUCLE=
 golang.org/x/mod v0.6.0-dev.0.20220419223038-86c51ed26bb4/go.mod h1:jJ57K6gSWd91VN4djpZkiMVwK6gcyfeH4XE8wZrZaV4=
 golang.org/x/mod v0.8.0/go.mod h1:iBbtSCu2XBx23ZKBPSOrRkjjQPZFPuis4dIYUhu/chs=
 golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=

internal/cli/cli.go

@@ -21,7 +21,6 @@ var (
 		Use:          "smash [flags] [locations-to-smash]",
 		Short:        "Find duplicates fast!",
 		Long:         "",
-		Version:      smash.Version,
 		SilenceUsage: true,
 		RunE:         runE,
 	}
@@ -42,8 +41,10 @@ func init() {
 	flags.IntVarP(&af.MaxWorkers, "max-workers", "w", bestMaxWorkers(), "Maximum workers to utilise when smashing.")
 	flags.BoolVarP(&af.DisableSlicing, "disable-slicing", "", false, "Disable slicing (hashes full file).")
 	flags.BoolVarP(&af.IgnoreEmptyFiles, "ignore-emptyfiles", "", false, "Ignore & don't report on empty/zero byte files.")
+	flags.StringVarP(&af.OutputFile, "output-file", "o", "", "Export as JSON")
 	flags.BoolVarP(&af.Silent, "silent", "q", false, "Run in silent mode.")
 	flags.BoolVarP(&af.Verbose, "verbose", "", false, "Run in verbose mode.")
+	flags.BoolVarP(&af.ShowVersion, "version", "v", false, "Show version information.")
 }
 func bestMaxWorkers() int {
 	cpus := runtime.NumCPU()

internal/report/output.go

@@ -0,0 +1,6 @@
+package report
+
+type ReportOutput struct {
+	Summary    *RunSummary
+	Duplicates []SmashFile
+}

internal/smash/app.go

@@ -7,10 +7,10 @@ import (
 	"sync/atomic"
 	"time"
 
-	"github.com/thushan/smash/internal/report"
-
 	"golang.org/x/term"
 
+	"github.com/thushan/smash/internal/report"
+
 	"github.com/pterm/pterm"
 	"github.com/thushan/smash/internal/theme"
 
@@ -25,6 +25,8 @@ import (
 type App struct {
 	Flags     *Flags
 	Session   *AppSession
+	Runtime   *AppRuntime
+	Summary   *report.RunSummary
 	Args      []string
 	Locations []string
 }
@@ -35,48 +37,62 @@ type AppSession struct {
 	StartTime int64
 	EndTime   int64
 }
+type AppRuntime struct {
+	Slicer        *slicer.Slicer
+	SlicerOptions *slicer.SlicerOptions
+	IndexerConfig *indexer.IndexerConfig
+	Files         chan indexer.FileFS
+}
 
 func (app *App) Run() error {
 
-	var emptyFiles []report.SmashFile
+	if !app.Flags.Silent {
+		PrintVersionInfo(app.Flags.ShowVersion)
+		if app.Flags.ShowVersion {
+			return nil
+		}
+		app.printConfiguration()
+	}
 
-	session := AppSession{
+	app.Session = &AppSession{
 		Dupes:     haxmap.New[string, []report.SmashFile](),
 		Fails:     haxmap.New[string, error](),
-		Empty:     &emptyFiles,
+		Empty:     &[]report.SmashFile{},
 		StartTime: time.Now().UnixMilli(),
 		EndTime:   -1,
 	}
-	app.Session = &session
-	app.setMaxThreads()
 
-	locations := app.Locations
-	excludeDirs := app.Flags.ExcludeDir
-	excludeFiles := app.Flags.ExcludeFile
-	disableSlicing := app.Flags.DisableSlicing
-
-	updateTicker := int64(1000)
-
-	if !term.IsTerminal(int(os.Stdout.Fd())) {
-		pterm.DisableColor()
-		pterm.DisableStyling()
+	sl := slicer.New(algorithms.Algorithm(app.Flags.Algorithm))
+	wk := indexer.NewConfigured(app.Flags.ExcludeDir, app.Flags.ExcludeFile)
+	slo := slicer.SlicerOptions{
+		DisableSlicing:       app.Flags.DisableSlicing,
+		DisableMeta:          false, // TODO: Flag this
+		DisableFileDetection: false, // TODO: Flag this
 	}
 
-	if !app.Flags.Silent {
-		PrintVersionInfo(false)
-		app.printConfiguration()
+	app.Runtime = &AppRuntime{
+		Slicer:        &sl,
+		SlicerOptions: &slo,
+		IndexerConfig: wk,
+		Files:         make(chan indexer.FileFS),
 	}
 
-	files := make(chan indexer.FileFS)
+	app.setMaxThreads()
+	app.checkTerminal()
 
-	sl := slicer.New(algorithms.Algorithm(app.Flags.Algorithm))
-	wk := indexer.NewConfigured(excludeDirs, excludeFiles)
+	return app.Exec()
+}
+func (app *App) Exec() error {
 
-	slo := slicer.SlicerOptions{
-		DisableSlicing:       disableSlicing,
-		DisableMeta:          false, // TODO: Flag this
-		DisableFileDetection: false, // TODO: Flag this
-	}
+	session := app.Session
+
+	wk := app.Runtime.IndexerConfig
+	sl := app.Runtime.Slicer
+	slo := app.Runtime.SlicerOptions
+
+	files := app.Runtime.Files
+	locations := app.Locations
+	isVerbose := app.Flags.Verbose && !app.Flags.Silent
 
 	pap := theme.MultiWriter()
 	psi, _ := theme.IndexingSpinner().WithWriter(pap.NewWriter()).Start("Indexing locations...")
@@ -92,7 +108,7 @@ func (app *App) Run() error {
 			err := wk.WalkDirectory(os.DirFS(location), location, files)
 
 			if err != nil {
-				if app.Flags.Verbose {
+				if isVerbose {
 					theme.WarnSkipWithContext(location, err)
 				}
 				_, _ = session.Fails.GetOrSet(location, err)
@@ -101,6 +117,7 @@ func (app *App) Run() error {
 	}()
 
 	totalFiles := int64(0)
+	updateTicker := int64(1000)
 
 	pss, _ := theme.SmashingSpinner().WithWriter(pap.NewWriter()).Start("Finding duplicates...")
 
@@ -119,11 +136,11 @@ func (app *App) Run() error {
 				}
 
 				startTime := time.Now().UnixMilli()
-				stats, err := sl.SliceFS(file.FileSystem, file.Path, &slo)
+				stats, err := sl.SliceFS(file.FileSystem, file.Path, slo)
 				elapsedMs := time.Now().UnixMilli() - startTime
 
 				if err != nil {
-					if app.Flags.Verbose {
+					if isVerbose {
 						theme.WarnSkipWithContext(file.FullName, err)
 					}
 					_, _ = session.Fails.GetOrSet(sf, err)
@@ -136,10 +153,22 @@ func (app *App) Run() error {
 	wg.Wait()
 
 	pss.Success("Finding duplicates...Done!")
+
+	psr, _ := theme.FinaliseSpinner().WithWriter(pap.NewWriter()).Start("Finding smash hits...")
+	app.generateRunSummary(totalFiles)
+	psr.Success("Finding smash hits...Done!")
+
 	pap.Stop()
 
-	summary := app.generateRunSummary(totalFiles)
-	report.PrintRunSummary(summary, app.Flags.IgnoreEmptyFiles)
+	app.PrintRunAnalysis(app.Flags.IgnoreEmptyFiles)
+	report.PrintRunSummary(*app.Summary, app.Flags.IgnoreEmptyFiles)
 
 	return nil
 }
+
+func (app *App) checkTerminal() {
+	if !term.IsTerminal(int(os.Stdout.Fd())) {
+		pterm.DisableColor()
+		pterm.DisableStyling()
+	}
+}

internal/smash/formatter.go

@@ -20,8 +20,53 @@ func (app *App) printVerbose(message ...any) {
 	}
 }
 
+func (app *App) PrintRunAnalysis(ignoreEmptyFiles bool) {
+	duplicates := app.Session.Dupes
+	emptyFiles := *app.Session.Empty
+
+	totalDuplicates := app.Summary.DuplicateFiles
+
+	theme.StyleHeading.Println("---| Duplicates (", totalDuplicates, ")")
+
+	if duplicates.Len() == 0 {
+		theme.Println(theme.ColourSuccess("No duplicates found :-)"))
+	} else {
+
+		duplicates.ForEach(func(hash string, files []report.SmashFile) bool {
+			duplicateFiles := len(files) - 1
+			if duplicateFiles != 0 {
+				root := files[0]
+				dupes := files[1:]
+				theme.Println(theme.ColourFilename(root.Filename), " ", theme.ColourFileSize(root.FileSizeF), " ", theme.ColourHash(root.Hash))
+				printSmashHits(dupes)
+			}
+			return true
+		})
+
+	}
+
+	if !ignoreEmptyFiles && len(emptyFiles) != 0 {
+		theme.StyleHeading.Println("---| Empty Files (", len(emptyFiles), ")")
+		printSmashHits(emptyFiles)
+	}
+
+}
+
+func printSmashHits(files []report.SmashFile) {
+	lastIndex := len(files) - 1
+	for index, file := range files {
+		var subTree string
+		if index < lastIndex {
+			subTree = TreeNextChild
+		} else {
+			subTree = TreeLastChild
+		}
+		theme.Println(theme.ColourFolderHierarchy(subTree), theme.ColourFilenameA(file.Filename))
+	}
+}
+
 // generateRunSummary Generates the smash hits of duplicates and returns the total size of duplicates.
-func (app *App) generateRunSummary(totalFiles int64) report.RunSummary {
+func (app *App) generateRunSummary(totalFiles int64) {
 	session := *app.Session
 	duplicates := session.Dupes
 	emptyFiles := *session.Empty
@@ -32,7 +77,6 @@ func (app *App) generateRunSummary(totalFiles int64) report.RunSummary {
 	totalFailFileCount := int64(session.Fails.Len())
 	totalEmptyFileCount := int64(len(emptyFiles))
 
-	theme.StyleHeading.Println("---| Duplicates")
 	duplicates.ForEach(func(hash string, files []report.SmashFile) bool {
 		duplicateFiles := len(files) - 1
 		if duplicateFiles == 0 {
@@ -41,24 +85,13 @@ func (app *App) generateRunSummary(totalFiles int64) report.RunSummary {
 		} else {
 			root := files[0]
 			dupes := files[1:]
-			theme.Println(theme.ColourFilename(root.Filename), " ", theme.ColourFileSize(humanize.Bytes(root.FileSize)), " ", theme.ColourHash(root.Hash))
-			printSmashHits(dupes)
 			totalDuplicates += len(dupes)
 			totalDuplicateSize += root.FileSize * uint64(duplicateFiles)
 		}
 		return true
 	})
 
-	if totalDuplicates == 0 {
-		theme.Println(theme.ColourSuccess("No duplicates found :-)"))
-	}
-
-	if !app.Flags.IgnoreEmptyFiles && totalEmptyFileCount != 0 {
-		theme.StyleHeading.Println("---| Empty Files (", totalEmptyFileCount, ")")
-		printSmashHits(emptyFiles)
-	}
-
-	return report.RunSummary{
+	summary := report.RunSummary{
 		TotalFiles:         totalFiles,
 		TotalFileErrors:    totalFailFileCount,
 		UniqueFiles:        totalUniqueFiles,
@@ -68,17 +101,5 @@ func (app *App) generateRunSummary(totalFiles int64) report.RunSummary {
 		DuplicateFileSizeF: humanize.Bytes(totalDuplicateSize),
 		ElapsedTime:        time.Now().UnixMilli() - app.Session.StartTime,
 	}
-}
-
-func printSmashHits(files []report.SmashFile) {
-	lastIndex := len(files) - 1
-	for index, file := range files {
-		var subTree string
-		if index < lastIndex {
-			subTree = TreeNextChild
-		} else {
-			subTree = TreeLastChild
-		}
-		theme.Println(theme.ColourFolderHierarchy(subTree), theme.ColourFilenameA(file.Filename))
-	}
+	app.Summary = &summary
 }

internal/smash/structs.go

@@ -1,6 +1,7 @@
 package smash
 
 type Flags struct {
+	OutputFile       string   `yaml:"output"`
 	Base             []string `yaml:"base"`
 	ExcludeDir       []string `yaml:"exclude-dir"`
 	ExcludeFile      []string `yaml:"exclude-file"`
@@ -9,6 +10,7 @@ type Flags struct {
 	MaxWorkers       int      `yaml:"max-workers"`
 	DisableSlicing   bool     `yaml:"disable-slicing"`
 	IgnoreEmptyFiles bool     `yaml:"ignore-emptyfiles"`
+	ShowVersion      bool     `yaml:"show-version"`
 	Silent           bool     `yaml:"silent"`
 	Verbose          bool     `yaml:"verbose"`
 }