new --add-only mode

README.md

@@ -8,10 +8,10 @@ chkbit is a tool that ensures the safety of your files by checking if their *dat
 
 - [How it works](#how-it-works)
 - [Installation](#installation)
-- [chkbit as a Go module](#chkbit-as-a-go-module)
 - [Usage](#usage)
 - [Repair](#repair)
 - [Ignore files](#ignore-files)
+- [chkbit as a Go module](#chkbit-as-a-go-module)
 - [FAQ](#faq)
 
 
@@ -50,7 +50,7 @@ Building from the source requires Go.
 - Either install it directly
 
 ```shell
-go install github.com/laktak/chkbit@latest
+go install github.com/laktak/chkbit/v5/cmd/chkbit@latest
 ```
 
 - or clone and build
@@ -58,22 +58,11 @@ go install github.com/laktak/chkbit@latest
 ```shell
 git clone https://github.com/laktak/chkbit
 chkbit/scripts/build
-# output is here:
-chkbit/chkbit
+# binary:
+ls -l chkbit/chkbit
 ```
 
 
-## chkbit as a Go module
-
-chkbit is can also be used in other Go programs.
-
-```
-go get github.com/laktak/chkbit
-```
-
-For more information see the linked documentation on [pkg.go.dev](https://pkg.go.dev/github.com/laktak/chkbit).
-
-
 ## Usage
 
 Run `chkbit -u PATH` to create/update the chkbit index.
@@ -95,13 +84,18 @@ Arguments:
 Flags:
   -h, --help                           Show context-sensitive help.
   -H, --tips                           Show tips.
-  -u, --update                         update indices (without this chkbit will verify files in readonly mode)
-      --show-ignored-only              only show ignored files
-      --algo="blake3"                  hash algorithm: md5, sha512, blake3 (default: blake3)
-  -f, --force                          force update of damaged items
-  -s, --skip-symlinks                  do not follow symlinks
+  -c, --check                          check mode: chkbit will verify files in readonly mode (default mode)
+  -u, --update                         update mode: add and update indices
+  -a, --add-only                       add mode: only add new files, do not check existing (quicker)
+  -i, --show-ignored-only              show-ignored mode: only show ignored files
+  -m, --show-missing                   show missing files/directories
+      --force                          force update of damaged items (advanced usage only)
+  -S, --skip-symlinks                  do not follow symlinks
+  -R, --no-recurse                     do not recurse into subdirectories
+  -D, --no-dir-in-index                do not track directories in the index
   -l, --log-file=STRING                write to a logfile if specified
       --log-verbose                    verbose logging
+      --algo="blake3"                  hash algorithm: md5, sha512, blake3 (default: blake3)
       --index-name=".chkbit"           filename where chkbit stores its hashes, needs to start with '.' (default: .chkbit)
       --ignore-name=".chkbitignore"    filename that chkbit reads its ignore list from, needs to start with '.' (default: .chkbitignore)
   -w, --workers=5                      number of workers to use (default: 5)
@@ -111,6 +105,27 @@ Flags:
   -V, --version                        show version information
 ```
 
+```
+$ chkbit -H
+
+.chkbitignore rules:
+  each line should contain exactly one name
+  you may use Unix shell-style wildcards (see README)
+  lines starting with '#' are skipped
+  lines starting with '/' are only applied to the current directory
+
+Status codes:
+  DMG: error, data damage detected
+  EIX: error, index damaged
+  old: warning, file replaced by an older version
+  new: new file
+  upd: file updated
+  ok : check ok
+  del: file/directory removed
+  ign: ignored (see .chkbitignore)
+  EXC: exception/panic
+```
+
 chkbit is set to use only 5 workers by default so it will not slow your system to a crawl. You can specify a higher number to make it a lot faster if the IO throughput can also keep up.
 
 
@@ -140,6 +155,17 @@ Add a `.chkbitignore` file containing the names of the files/directories you wis
 - hidden files (starting with a `.`) are ignored by default
 
 
+## chkbit as a Go module
+
+chkbit is can also be used in other Go programs.
+
+```
+go get github.com/laktak/chkbit/v5
+```
+
+For more information see the documentation on [pkg.go.dev](https://pkg.go.dev/github.com/laktak/chkbit/v5).
+
+
 ## FAQ
 
 ### Should I run `chkbit` on my whole drive?

cmd/chkbit/help.go

@@ -19,6 +19,7 @@ Status codes:
   new: new file
   upd: file updated
   ok : check ok
+  del: file/directory removed
   ign: ignored (see .chkbitignore)
   EXC: exception/panic
 `

cmd/chkbit/main.go

@@ -10,8 +10,8 @@ import (
 	"time"
 
 	"github.com/alecthomas/kong"
-	"github.com/laktak/chkbit"
-	"github.com/laktak/chkbit/cmd/chkbit/util"
+	"github.com/laktak/chkbit/v5"
+	"github.com/laktak/chkbit/v5/cmd/chkbit/util"
 	"github.com/laktak/lterm"
 )
 
@@ -44,13 +44,18 @@ var (
 var cli struct {
 	Paths           []string `arg:"" optional:"" name:"paths" help:"directories to check"`
 	Tips            bool     `short:"H" help:"Show tips."`
-	Update          bool     `short:"u" help:"update indices (without this chkbit will verify files in readonly mode)"`
-	ShowIgnoredOnly bool     `help:"only show ignored files"`
-	Algo            string   `default:"blake3" help:"hash algorithm: md5, sha512, blake3 (default: blake3)"`
-	Force           bool     `short:"f" help:"force update of damaged items"`
-	SkipSymlinks    bool     `short:"s" help:"do not follow symlinks"`
+	Check           bool     `short:"c" help:"check mode: chkbit will verify files in readonly mode (default mode)"`
+	Update          bool     `short:"u" help:"update mode: add and update indices"`
+	AddOnly         bool     `short:"a" help:"add mode: only add new files, do not check existing (quicker)"`
+	ShowIgnoredOnly bool     `short:"i" help:"show-ignored mode: only show ignored files"`
+	ShowMissing     bool     `short:"m" help:"show missing files/directories"`
+	Force           bool     `help:"force update of damaged items (advanced usage only)"`
+	SkipSymlinks    bool     `short:"S" help:"do not follow symlinks"`
+	NoRecurse       bool     `short:"R" help:"do not recurse into subdirectories"`
+	NoDirInIndex    bool     `short:"D" help:"do not track directories in the index"`
 	LogFile         string   `short:"l" help:"write to a logfile if specified"`
 	LogVerbose      bool     `help:"verbose logging"`
+	Algo            string   `default:"blake3" help:"hash algorithm: md5, sha512, blake3 (default: blake3)"`
 	IndexName       string   `default:".chkbit" help:"filename where chkbit stores its hashes, needs to start with '.' (default: .chkbit)"`
 	IgnoreName      string   `default:".chkbitignore" help:"filename that chkbit reads its ignore list from, needs to start with '.' (default: .chkbitignore)"`
 	Workers         int      `short:"w" default:"5" help:"number of workers to use (default: 5)"`
@@ -61,19 +66,16 @@ var cli struct {
 }
 
 type Main struct {
+	context    *chkbit.Context
 	dmgList    []string
 	errList    []string
-	numIdxUpd  int
-	numNew     int
-	numUpd     int
 	verbose    bool
 	logger     *log.Logger
 	logVerbose bool
 	progress   Progress
-	total      int
 	termWidth  int
-	fps        *RateCalc
-	bps        *RateCalc
+	fps        *util.RateCalc
+	bps        *util.RateCalc
 }
 
 func (m *Main) log(text string) {
@@ -82,44 +84,36 @@ func (m *Main) log(text string) {
 
 func (m *Main) logStatus(stat chkbit.Status, message string) bool {
 	if stat == chkbit.STATUS_UPDATE_INDEX {
-		m.numIdxUpd++
-	} else {
-		if stat == chkbit.STATUS_ERR_DMG {
-			m.total++
-			m.dmgList = append(m.dmgList, message)
-		} else if stat == chkbit.STATUS_PANIC {
-			m.errList = append(m.errList, message)
-		} else if stat == chkbit.STATUS_OK || stat == chkbit.STATUS_UPDATE || stat == chkbit.STATUS_NEW || stat == chkbit.STATUS_UP_WARN_OLD {
-			m.total++
-			if stat == chkbit.STATUS_UPDATE || stat == chkbit.STATUS_UP_WARN_OLD {
-				m.numUpd++
-			} else if stat == chkbit.STATUS_NEW {
-				m.numNew++
-			}
-		}
+		return false
+	}
 
-		if m.logVerbose || stat != chkbit.STATUS_OK && stat != chkbit.STATUS_IGNORE {
-			m.log(stat.String() + " " + message)
-		}
+	if stat == chkbit.STATUS_ERR_DMG {
+		m.dmgList = append(m.dmgList, message)
+	} else if stat == chkbit.STATUS_PANIC {
+		m.errList = append(m.errList, message)
+	}
 
-		if m.verbose || !stat.IsVerbose() {
-			col := ""
-			if stat.IsErrorOrWarning() {
-				col = termAlertFG
-			}
-			lterm.Printline(col, stat.String(), " ", message, lterm.Reset)
-			return true
+	if m.logVerbose || !stat.IsVerbose() {
+		m.log(stat.String() + " " + message)
+	}
+
+	if m.verbose || !stat.IsVerbose() {
+		col := ""
+		if stat.IsErrorOrWarning() {
+			col = termAlertFG
 		}
+		lterm.Printline(col, stat.String(), " ", message, lterm.Reset)
+		return true
 	}
 	return false
 }
 
-func (m *Main) showStatus(context *chkbit.Context) {
+func (m *Main) showStatus() {
 	last := time.Now().Add(-updateInterval)
 	stat := ""
 	for {
 		select {
-		case item := <-context.LogQueue:
+		case item := <-m.context.LogQueue:
 			if item == nil {
 				if m.progress == Fancy {
 					lterm.Printline("")
@@ -130,10 +124,10 @@ func (m *Main) showStatus(context *chkbit.Context) {
 				if m.progress == Fancy {
 					lterm.Write(termBG, termFG1, stat, lterm.ClearLine(0), lterm.Reset, "\r")
 				} else {
-					fmt.Print(m.total, "\r")
+					fmt.Print(m.context.NumTotal, "\r")
 				}
 			}
-		case perf := <-context.PerfQueue:
+		case perf := <-m.context.PerfQueue:
 			now := time.Now()
 			m.fps.Push(now, perf.NumFiles)
 			m.bps.Push(now, perf.NumBytes)
@@ -143,11 +137,11 @@ func (m *Main) showStatus(context *chkbit.Context) {
 					statF := fmt.Sprintf("%d files/s", m.fps.Last())
 					statB := fmt.Sprintf("%d MB/s", m.bps.Last()/sizeMB)
 					stat = "RW"
-					if !context.Update {
+					if !m.context.UpdateIndex {
 						stat = "RO"
 					}
 					stat = fmt.Sprintf("[%s:%d] %5d files $ %s %-13s $ %s %-13s",
-						stat, context.NumWorkers, m.total,
+						stat, m.context.NumWorkers, m.context.NumTotal,
 						util.Sparkline(m.fps.Stats), statF,
 						util.Sparkline(m.bps.Stats), statB)
 					stat = util.LeftTruncate(stat, m.termWidth-1)
@@ -155,38 +149,49 @@ func (m *Main) showStatus(context *chkbit.Context) {
 					stat = strings.Replace(stat, "$", termSepFG+termSep+termFG3, 1)
 					lterm.Write(termBG, termFG1, stat, lterm.ClearLine(0), lterm.Reset, "\r")
 				} else if m.progress == Plain {
-					fmt.Print(m.total, "\r")
+					fmt.Print(m.context.NumTotal, "\r")
 				}
 			}
 		}
 	}
 }
 
-func (m *Main) process() *chkbit.Context {
-	if cli.Update && cli.ShowIgnoredOnly {
-		fmt.Println("Error: use either --update or --show-ignored-only!")
-		return nil
+func (m *Main) process() bool {
+	// verify mode
+	var b01 = map[bool]int8{false: 0, true: 1}
+	if b01[cli.Check]+b01[cli.Update]+b01[cli.AddOnly]+b01[cli.ShowIgnoredOnly] > 1 {
+		fmt.Println("Error: can only run one mode at a time!")
+		os.Exit(1)
 	}
 
-	context, err := chkbit.NewContext(cli.Workers, cli.Force, cli.Update, cli.ShowIgnoredOnly, cli.Algo, cli.SkipSymlinks, cli.IndexName, cli.IgnoreName)
+	var err error
+	m.context, err = chkbit.NewContext(cli.Workers, cli.Algo, cli.IndexName, cli.IgnoreName)
 	if err != nil {
 		fmt.Println(err)
-		return nil
+		return false
 	}
+	m.context.ForceUpdateDmg = cli.Force
+	m.context.UpdateIndex = cli.Update || cli.AddOnly
+	m.context.AddOnly = cli.AddOnly
+	m.context.ShowIgnoredOnly = cli.ShowIgnoredOnly
+	m.context.ShowMissing = cli.ShowMissing
+	m.context.SkipSymlinks = cli.SkipSymlinks
+	m.context.SkipSubdirectories = cli.NoRecurse
+	m.context.TrackDirectories = !cli.NoDirInIndex
 
 	var wg sync.WaitGroup
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
-		m.showStatus(context)
+		m.showStatus()
 	}()
-	context.Start(cli.Paths)
+	m.context.Start(cli.Paths)
 	wg.Wait()
 
-	return context
+	return true
 }
 
-func (m *Main) printResult(context *chkbit.Context) {
+func (m *Main) printResult() {
 	cprint := func(col, text string) {
 		if m.progress != Quiet {
 			if m.progress == Fancy {
@@ -209,14 +214,14 @@ func (m *Main) printResult(context *chkbit.Context) {
 
 	if m.progress != Quiet {
 		mode := ""
-		if !context.Update {
+		if !m.context.UpdateIndex {
 			mode = " in readonly mode"
 		}
-		status := fmt.Sprintf("Processed %s%s.", util.LangNum1MutateSuffix(m.total, "file"), mode)
+		status := fmt.Sprintf("Processed %s%s.", util.LangNum1MutateSuffix(m.context.NumTotal, "file"), mode)
 		cprint(termOKFG, status)
 		m.log(status)
 
-		if m.progress == Fancy && m.total > 0 {
+		if m.progress == Fancy && m.context.NumTotal > 0 {
 			elapsed := time.Since(m.fps.Start)
 			elapsedS := elapsed.Seconds()
 			fmt.Println("-", elapsed.Truncate(time.Second), "elapsed")
@@ -224,17 +229,26 @@ func (m *Main) printResult(context *chkbit.Context) {
 			fmt.Printf("- %.2f MB/second\n", (float64(m.bps.Total)+float64(m.bps.Current))/float64(sizeMB)/elapsedS)
 		}
 
-		if context.Update {
-			if m.numIdxUpd > 0 {
-				cprint(termOKFG, fmt.Sprintf("- %s updated\n- %s added\n- %s updated",
-					util.LangNum1Choice(m.numIdxUpd, "directory was", "directories were"),
-					util.LangNum1Choice(m.numNew, "file hash was", "file hashes were"),
-					util.LangNum1Choice(m.numUpd, "file hash was", "file hashes were")))
+		del := ""
+		if m.context.UpdateIndex {
+			if m.context.NumIdxUpd > 0 {
+				if m.context.NumDel > 0 {
+					del = fmt.Sprintf("\n- %s been removed", util.LangNum1Choice(m.context.NumDel, "file/directory has", "files/directories have"))
+				}
+				cprint(termOKFG, fmt.Sprintf("- %s updated\n- %s added\n- %s updated%s",
+					util.LangNum1Choice(m.context.NumIdxUpd, "directory was", "directories were"),
+					util.LangNum1Choice(m.context.NumNew, "file hash was", "file hashes were"),
+					util.LangNum1Choice(m.context.NumUpd, "file hash was", "file hashes were"),
+					del))
 			}
-		} else if m.numNew+m.numUpd > 0 {
-			cprint(termAlertFG, fmt.Sprintf("No changes were made (specify -u to update):\n- %s would have been added and\n- %s would have been updated.",
-				util.LangNum1MutateSuffix(m.numNew, "file"),
-				util.LangNum1MutateSuffix(m.numUpd, "file")))
+		} else if m.context.NumNew+m.context.NumUpd+m.context.NumDel > 0 {
+			if m.context.NumDel > 0 {
+				del = fmt.Sprintf("\n- %s would have been removed", util.LangNum1Choice(m.context.NumDel, "file/directory", "files/directories"))
+			}
+			cprint(termAlertFG, fmt.Sprintf("No changes were made (specify -u to update):\n- %s would have been added\n- %s would have been updated%s",
+				util.LangNum1MutateSuffix(m.context.NumNew, "file"),
+				util.LangNum1MutateSuffix(m.context.NumUpd, "file"),
+				del))
 		}
 	}
 
@@ -310,9 +324,8 @@ func (m *Main) run() {
 
 	if len(cli.Paths) > 0 {
 		m.log("chkbit " + strings.Join(cli.Paths, ", "))
-		context := m.process()
-		if context != nil && !context.ShowIgnoredOnly {
-			m.printResult(context)
+		if m.process() && !m.context.ShowIgnoredOnly {
+			m.printResult()
 		}
 	} else {
 		fmt.Println("specify a path to check, see -h")
@@ -331,8 +344,8 @@ func main() {
 	m := &Main{
 		logger:    log.New(io.Discard, "", 0),
 		termWidth: termWidth,
-		fps:       NewRateCalc(time.Second, (termWidth-70)/2),
-		bps:       NewRateCalc(time.Second, (termWidth-70)/2),
+		fps:       util.NewRateCalc(time.Second, (termWidth-70)/2),
+		bps:       util.NewRateCalc(time.Second, (termWidth-70)/2),
 	}
 	m.run()
 }

cmd/chkbit/util/ratecalc.go

@@ -1,4 +1,4 @@
-package main
+package util
 
 import (
 	"time"

context.go

@@ -8,21 +8,32 @@ import (
 )
 
 type Context struct {
-	NumWorkers      int
-	Force           bool
-	Update          bool
-	ShowIgnoredOnly bool
-	HashAlgo        string
-	SkipSymlinks    bool
-	IndexFilename   string
-	IgnoreFilename  string
-	WorkQueue       chan *WorkItem
-	LogQueue        chan *LogEvent
-	PerfQueue       chan *PerfEvent
-	wg              sync.WaitGroup
+	NumWorkers         int
+	UpdateIndex        bool
+	AddOnly            bool
+	ShowIgnoredOnly    bool
+	ShowMissing        bool
+	ForceUpdateDmg     bool
+	HashAlgo           string
+	TrackDirectories   bool
+	SkipSymlinks       bool
+	SkipSubdirectories bool
+	IndexFilename      string
+	IgnoreFilename     string
+	WorkQueue          chan *WorkItem
+	LogQueue           chan *LogEvent
+	PerfQueue          chan *PerfEvent
+	wg                 sync.WaitGroup
+
+	mutex     sync.Mutex
+	NumTotal  int
+	NumIdxUpd int
+	NumNew    int
+	NumUpd    int
+	NumDel    int
 }
 
-func NewContext(numWorkers int, force bool, update bool, showIgnoredOnly bool, hashAlgo string, skipSymlinks bool, indexFilename string, ignoreFilename string) (*Context, error) {
+func NewContext(numWorkers int, hashAlgo string, indexFilename string, ignoreFilename string) (*Context, error) {
 	if indexFilename[0] != '.' {
 		return nil, errors.New("The index filename must start with a dot!")
 	}
@@ -33,21 +44,44 @@ func NewContext(numWorkers int, force bool, update bool, showIgnoredOnly bool, h
 		return nil, errors.New(hashAlgo + " is unknown.")
 	}
 	return &Context{
-		NumWorkers:      numWorkers,
-		Force:           force,
-		Update:          update,
-		ShowIgnoredOnly: showIgnoredOnly,
-		HashAlgo:        hashAlgo,
-		SkipSymlinks:    skipSymlinks,
-		IndexFilename:   indexFilename,
-		IgnoreFilename:  ignoreFilename,
-		WorkQueue:       make(chan *WorkItem, numWorkers*10),
-		LogQueue:        make(chan *LogEvent, numWorkers*100),
-		PerfQueue:       make(chan *PerfEvent, numWorkers*10),
+		NumWorkers:     numWorkers,
+		HashAlgo:       hashAlgo,
+		IndexFilename:  indexFilename,
+		IgnoreFilename: ignoreFilename,
+		WorkQueue:      make(chan *WorkItem, numWorkers*10),
+		LogQueue:       make(chan *LogEvent, numWorkers*100),
+		PerfQueue:      make(chan *PerfEvent, numWorkers*10),
 	}, nil
 }
 
 func (context *Context) log(stat Status, message string) {
+	context.mutex.Lock()
+	defer context.mutex.Unlock()
+	switch stat {
+	case STATUS_ERR_DMG:
+		context.NumTotal++
+	case STATUS_UPDATE_INDEX:
+		context.NumIdxUpd++
+	case STATUS_UP_WARN_OLD:
+		context.NumTotal++
+		context.NumUpd++
+	case STATUS_UPDATE:
+		context.NumTotal++
+		context.NumUpd++
+	case STATUS_NEW:
+		context.NumTotal++
+		context.NumNew++
+	case STATUS_OK:
+		if !context.AddOnly {
+			context.NumTotal++
+		}
+	case STATUS_MISSING:
+		context.NumDel++
+		//case STATUS_PANIC:
+		//case STATUS_ERR_IDX:
+		//case STATUS_IGNORE:
+	}
+
 	context.LogQueue <- &LogEvent{stat, message}
 }
 
@@ -63,8 +97,8 @@ func (context *Context) perfMonBytes(numBytes int64) {
 	context.PerfQueue <- &PerfEvent{0, numBytes}
 }
 
-func (context *Context) addWork(path string, filesToIndex []string, ignore *Ignore) {
-	context.WorkQueue <- &WorkItem{path, filesToIndex, ignore}
+func (context *Context) addWork(path string, filesToIndex []string, dirList []string, ignore *Ignore) {
+	context.WorkQueue <- &WorkItem{path, filesToIndex, dirList, ignore}
 }
 
 func (context *Context) endWork() {
@@ -76,12 +110,18 @@ func (context *Context) isChkbitFile(name string) bool {
 }
 
 func (context *Context) Start(pathList []string) {
+	context.NumTotal = 0
+	context.NumIdxUpd = 0
+	context.NumNew = 0
+	context.NumUpd = 0
+	context.NumDel = 0
+
 	var wg sync.WaitGroup
 	wg.Add(context.NumWorkers)
 	for i := 0; i < context.NumWorkers; i++ {
 		go func(id int) {
 			defer wg.Done()
-			context.RunWorker(id)
+			context.runWorker(id)
 		}(i)
 	}
 	go func() {
@@ -121,6 +161,11 @@ func (context *Context) scanDir(root string, parentIgnore *Ignore) {
 	var dirList []string
 	var filesToIndex []string
 
+	ignore, err := GetIgnore(context, root, parentIgnore)
+	if err != nil {
+		context.logErr(root+"/", err)
+	}
+
 	for _, file := range files {
 		path := filepath.Join(root, file.Name())
 		if file.Name()[0] == '.' {
@@ -130,24 +175,21 @@ func (context *Context) scanDir(root string, parentIgnore *Ignore) {
 			continue
 		}
 		if isDir(file, path) {
-			dirList = append(dirList, file.Name())
+			if !ignore.shouldIgnore(file.Name()) {
+				dirList = append(dirList, file.Name())
+			} else {
+				context.log(STATUS_IGNORE, file.Name()+"/")
+			}
 		} else if file.Type().IsRegular() {
 			filesToIndex = append(filesToIndex, file.Name())
 		}
 	}
 
-	ignore, err := GetIgnore(context, root, parentIgnore)
-	if err != nil {
-		context.logErr(root+"/", err)
-	}
+	context.addWork(root, filesToIndex, dirList, ignore)
 
-	context.addWork(root, filesToIndex, ignore)
-
-	for _, name := range dirList {
-		if !ignore.shouldIgnore(name) {
+	if !context.SkipSubdirectories {
+		for _, name := range dirList {
 			context.scanDir(filepath.Join(root, name), ignore)
-		} else {
-			context.log(STATUS_IGNORE, name+"/")
 		}
 	}
 }

go.mod

@@ -1,4 +1,4 @@
-module github.com/laktak/chkbit
+module github.com/laktak/chkbit/v5
 
 go 1.22.3
 

hashfile.go

@@ -50,7 +50,7 @@ func Hashfile(path string, hashAlgo string, perfMonBytes func(int64)) (string, e
 	return hex.EncodeToString(h.Sum(nil)), nil
 }
 
-func HashMd5(data []byte) string {
+func hashMd5(data []byte) string {
 	h := md5.New()
 	h.Write(data)
 	return hex.EncodeToString(h.Sum(nil))

index.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"os"
 	"path/filepath"
+	"slices"
 )
 
 const VERSION = 2 // index version
@@ -12,47 +13,54 @@ var (
 	algoMd5 = "md5"
 )
 
-type IdxInfo struct {
+type idxInfo struct {
 	ModTime    int64   `json:"mod"`
 	Algo       *string `json:"a,omitempty"`
 	Hash       *string `json:"h,omitempty"`
 	LegacyHash *string `json:"md5,omitempty"`
 }
 
-type IndexFile struct {
-	V       int             `json:"v"`
+type indexFile struct {
+	V int `json:"v"`
+	// IdxRaw -> map[string]idxInfo
 	IdxRaw  json.RawMessage `json:"idx"`
 	IdxHash string          `json:"idx_hash"`
+	// 2024-08 optional, list of subdirectories
+	Dir []string `json:"dirlist,omitempty"`
 }
 
-type IdxInfo1 struct {
+type idxInfo1 struct {
 	ModTime int64  `json:"mod"`
 	Hash    string `json:"md5"`
 }
 
-type IndexFile1 struct {
-	Data map[string]IdxInfo1 `json:"data"`
+type indexFile1 struct {
+	Data map[string]idxInfo1 `json:"data"`
 }
 
 type Index struct {
-	context  *Context
-	path     string
-	files    []string
-	cur      map[string]IdxInfo
-	new      map[string]IdxInfo
-	updates  []string
-	modified bool
-	readonly bool
+	context    *Context
+	path       string
+	files      []string
+	cur        map[string]idxInfo
+	new        map[string]idxInfo
+	curDirList []string
+	newDirList []string
+	modified   bool
+	readonly   bool
 }
 
-func NewIndex(context *Context, path string, files []string, readonly bool) *Index {
+func newIndex(context *Context, path string, files []string, dirList []string, readonly bool) *Index {
+	slices.Sort(dirList)
 	return &Index{
-		context:  context,
-		path:     path,
-		files:    files,
-		cur:      make(map[string]IdxInfo),
-		new:      make(map[string]IdxInfo),
-		readonly: readonly,
+		context:    context,
+		path:       path,
+		files:      files,
+		cur:        make(map[string]idxInfo),
+		new:        make(map[string]idxInfo),
+		curDirList: make([]string, 0),
+		newDirList: dirList,
+		readonly:   readonly,
 	}
 }
 
@@ -60,10 +68,6 @@ func (i *Index) getIndexFilepath() string {
 	return filepath.Join(i.path, i.context.IndexFilename)
 }
 
-func (i *Index) setMod(value bool) {
-	i.modified = value
-}
-
 func (i *Index) logFilePanic(name string, message string) {
 	i.context.log(STATUS_PANIC, filepath.Join(i.path, name)+": "+message)
 }
@@ -72,6 +76,10 @@ func (i *Index) logFile(stat Status, name string) {
 	i.context.log(stat, filepath.Join(i.path, name))
 }
 
+func (i *Index) logDir(stat Status, name string) {
+	i.context.log(stat, filepath.Join(i.path, name)+"/")
+}
+
 func (i *Index) calcHashes(ignore *Ignore) {
 	for _, name := range i.files {
 		if ignore != nil && ignore.shouldIgnore(name) {
@@ -80,13 +88,13 @@ func (i *Index) calcHashes(ignore *Ignore) {
 		}
 
 		var err error
-		var info *IdxInfo
+		var info *idxInfo
 		algo := i.context.HashAlgo
 		if val, ok := i.cur[name]; ok {
-			// existing
+			// existing file
 			if val.LegacyHash != nil {
 				// convert from py1 to new format
-				val = IdxInfo{
+				val = idxInfo{
 					ModTime: val.ModTime,
 					Algo:    &algoMd5,
 					Hash:    val.LegacyHash,
@@ -96,10 +104,15 @@ func (i *Index) calcHashes(ignore *Ignore) {
 			if val.Algo != nil {
 				algo = *val.Algo
 			}
-			info, err = i.calcFile(name, algo)
+			if i.context.AddOnly {
+				info = &val
+			} else {
+				info, err = i.calcFile(name, algo)
+			}
 		} else {
+			// new file
 			if i.readonly {
-				info = &IdxInfo{Algo: &algo}
+				info = &idxInfo{Algo: &algo}
 			} else {
 				info, err = i.calcFile(name, algo)
 			}
@@ -120,41 +133,70 @@ func (i *Index) showIgnoredOnly(ignore *Ignore) {
 	}
 }
 
-func (i *Index) checkFix(force bool) {
+func (i *Index) checkFix(forceUpdateDmg bool) {
 	for name, b := range i.new {
 		if a, ok := i.cur[name]; !ok {
 			i.logFile(STATUS_NEW, name)
-			i.setMod(true)
+			i.modified = true
 		} else {
 			amod := int64(a.ModTime)
 			bmod := int64(b.ModTime)
 			if a.Hash != nil && b.Hash != nil && *a.Hash == *b.Hash {
 				i.logFile(STATUS_OK, name)
 				if amod != bmod {
-					i.setMod(true)
+					i.modified = true
 				}
 				continue
 			}
 
 			if amod == bmod {
 				i.logFile(STATUS_ERR_DMG, name)
-				if !force {
+				if !forceUpdateDmg {
+					// keep DMG entry
 					i.new[name] = a
 				} else {
-					i.setMod(true)
+					i.modified = true
 				}
 			} else if amod < bmod {
 				i.logFile(STATUS_UPDATE, name)
-				i.setMod(true)
+				i.modified = true
 			} else if amod > bmod {
 				i.logFile(STATUS_UP_WARN_OLD, name)
-				i.setMod(true)
+				i.modified = true
 			}
 		}
 	}
+	// track missing
+	for name := range i.cur {
+		if _, ok := i.new[name]; !ok {
+			i.modified = true
+			if i.context.ShowMissing {
+				i.logFile(STATUS_MISSING, name)
+			}
+		}
+	}
+
+	// dirs
+	m := make(map[string]bool)
+	for _, n := range i.newDirList {
+		m[n] = true
+	}
+	for _, name := range i.curDirList {
+		if !m[name] {
+			i.modified = true
+			if i.context.ShowMissing {
+				i.logDir(STATUS_MISSING, name+"/")
+			}
+		}
+	}
+	if len(i.newDirList) != len(i.curDirList) {
+		// added
+		i.modified = true
+	}
+
 }
 
-func (i *Index) calcFile(name string, a string) (*IdxInfo, error) {
+func (i *Index) calcFile(name string, a string) (*idxInfo, error) {
 	path := filepath.Join(i.path, name)
 	info, _ := os.Stat(path)
 	mtime := int64(info.ModTime().UnixNano() / 1e6)
@@ -163,7 +205,7 @@ func (i *Index) calcFile(name string, a string) (*IdxInfo, error) {
 		return nil, err
 	}
 	i.context.perfMonFiles(1)
-	return &IdxInfo{
+	return &idxInfo{
 		ModTime: mtime,
 		Algo:    &a,
 		Hash:    &h,
@@ -180,10 +222,13 @@ func (i *Index) save() (bool, error) {
 		if err != nil {
 			return false, err
 		}
-		data := IndexFile{
+		data := indexFile{
 			V:       VERSION,
 			IdxRaw:  text,
-			IdxHash: HashMd5(text),
+			IdxHash: hashMd5(text),
+		}
+		if i.context.TrackDirectories {
+			data.Dir = i.newDirList
 		}
 
 		file, err := json.Marshal(data)
@@ -194,7 +239,7 @@ func (i *Index) save() (bool, error) {
 		if err != nil {
 			return false, err
 		}
-		i.setMod(false)
+		i.modified = false
 		return true, nil
 	} else {
 		return false, nil
@@ -208,12 +253,12 @@ func (i *Index) load() error {
 		}
 		return err
 	}
-	i.setMod(false)
+	i.modified = false
 	file, err := os.ReadFile(i.getIndexFilepath())
 	if err != nil {
 		return err
 	}
-	var data IndexFile
+	var data indexFile
 	err = json.Unmarshal(file, &data)
 	if err != nil {
 		return err
@@ -224,22 +269,22 @@ func (i *Index) load() error {
 			return err
 		}
 		text := data.IdxRaw
-		if data.IdxHash != HashMd5(text) {
-			// old versions may have save the JSON encoded with extra spaces
+		if data.IdxHash != hashMd5(text) {
+			// old versions may have saved the JSON encoded with extra spaces
 			text, _ = json.Marshal(data.IdxRaw)
 		} else {
 		}
-		if data.IdxHash != HashMd5(text) {
-			i.setMod(true)
+		if data.IdxHash != hashMd5(text) {
+			i.modified = true
 			i.logFile(STATUS_ERR_IDX, i.getIndexFilepath())
 		}
 	} else {
-		var data1 IndexFile1
+		var data1 indexFile1
 		json.Unmarshal(file, &data1)
 		if data1.Data != nil {
 			// convert from js to new format
 			for name, item := range data1.Data {
-				i.cur[name] = IdxInfo{
+				i.cur[name] = idxInfo{
 					ModTime: item.ModTime,
 					Algo:    &algoMd5,
 					Hash:    &item.Hash,
@@ -247,5 +292,12 @@ func (i *Index) load() error {
 			}
 		}
 	}
+
+	// dirs
+	if data.Dir != nil {
+		slices.Sort(data.Dir)
+		i.curDirList = data.Dir
+	}
+
 	return nil
 }

scripts/run_test.go

@@ -11,6 +11,8 @@ import (
 	"time"
 )
 
+// perform integration test using the compiled binary
+
 var testDir = "/tmp/chkbit"
 
 func getCmd() string {
@@ -25,6 +27,12 @@ func checkOut(t *testing.T, sout string, expected string) {
 	}
 }
 
+func checkNotOut(t *testing.T, sout string, notExpected string) {
+	if strings.Contains(sout, notExpected) {
+		t.Errorf("Did not expect '%s' in output, got '%s'\n", notExpected, sout)
+	}
+}
+
 // misc files
 
 var (
@@ -59,7 +67,12 @@ func setDate(filename string, r int) {
 	os.Chtimes(filename, date, date)
 }
 
-func genFile(dir string, a int) {
+func genFile(path string, size int) {
+	os.WriteFile(path, make([]byte, size), 0644)
+	setDate(path, size*size)
+}
+
+func genFiles(dir string, a int) {
 	os.MkdirAll(dir, 0755)
 	for i := 1; i <= 5; i++ {
 		size := a*i*wordIdx*100 + extIdx
@@ -70,9 +83,7 @@ func genFile(dir string, a int) {
 		}
 
 		file += "." + nextExt()
-		path := filepath.Join(dir, file)
-		os.WriteFile(path, make([]byte, size), 0644)
-		setDate(path, size*size)
+		genFile(filepath.Join(dir, file), size)
 	}
 }
 
@@ -81,11 +92,11 @@ func genDir(root string) {
 
 		for i := 1; i <= 5; i++ {
 			dir := filepath.Join(root, start, nextWord())
-			genFile(dir, 1)
+			genFiles(dir, 1)
 
 			if wordIdx%3 == 0 {
 				dir = filepath.Join(dir, nextWord())
-				genFile(dir, 1)
+				genFiles(dir, 1)
 			}
 		}
 	}
@@ -107,6 +118,8 @@ func setupMiscFiles() {
 
 	genDir(root)
 
+	os.MkdirAll(filepath.Join(root, "day/car/empty"), 0755)
+
 	rootPeople := filepath.Join(root, "people")
 	testPeople := filepath.Join(testDir, "people")
 
@@ -128,14 +141,125 @@ func TestRoot(t *testing.T) {
 
 	tool := getCmd()
 	root := filepath.Join(testDir, "root")
-	cmd := exec.Command(tool, "-u", root)
-	out, err := cmd.Output()
-	if err != nil {
-		t.Fatalf("cmd.Output() failed with '%s'\n", err)
-	}
-	sout := string(out)
-	checkOut(t, sout, "60 directories were updated")
-	checkOut(t, sout, "300 file hashes were added")
+
+	// update index, no recourse
+	t.Run("no-recourse", func(t *testing.T) {
+		cmd := exec.Command(tool, "-umR", filepath.Join(root, "day/office"))
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "Processed 5 files")
+		checkOut(t, sout, "- 1 directory was updated")
+		checkOut(t, sout, "- 5 file hashes were added")
+		checkOut(t, sout, "- 0 file hashes were updated")
+		checkNotOut(t, sout, "removed")
+	})
+
+	// update remaining index from root
+	t.Run("update-remaining", func(t *testing.T) {
+		cmd := exec.Command(tool, "-um", root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "Processed 300 files")
+		checkOut(t, sout, "- 66 directories were updated")
+		checkOut(t, sout, "- 295 file hashes were added")
+		checkOut(t, sout, "- 0 file hashes were updated")
+		checkNotOut(t, sout, "removed")
+	})
+
+	// delete files, check for missing
+	t.Run("delete", func(t *testing.T) {
+		os.RemoveAll(filepath.Join(root, "thing/change"))
+		os.Remove(filepath.Join(root, "time/hour/minute/body-information.csv"))
+
+		cmd := exec.Command(tool, "-m", root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "del /tmp/chkbit/root/thing/change/")
+		checkOut(t, sout, "2 files/directories would have been removed")
+	})
+
+	// do not report missing without -m
+	t.Run("no-missing", func(t *testing.T) {
+		cmd := exec.Command(tool, root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkNotOut(t, sout, "del ")
+		checkNotOut(t, sout, "removed")
+	})
+
+	// check for missing and update
+	t.Run("missing", func(t *testing.T) {
+		cmd := exec.Command(tool, "-um", root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "del /tmp/chkbit/root/thing/change/")
+		checkOut(t, sout, "2 files/directories have been removed")
+	})
+
+	// check again
+	t.Run("repeat", func(t *testing.T) {
+		for i := 0; i < 10; i++ {
+			cmd := exec.Command(tool, "-uv", root)
+			out, err := cmd.Output()
+			if err != nil {
+				t.Fatalf("failed with '%s'\n", err)
+			}
+			sout := string(out)
+			checkOut(t, sout, "Processed 289 files")
+			checkNotOut(t, sout, "removed")
+			checkNotOut(t, sout, "updated")
+			checkNotOut(t, sout, "added")
+		}
+	})
+
+	// add files only
+	t.Run("add-only", func(t *testing.T) {
+
+		genFiles(filepath.Join(root, "way/add"), 99)
+		genFile(filepath.Join(root, "time/add-file.txt"), 500)
+		// modify existing, will not be reported:
+		genFile(filepath.Join(root, "way/job/word-business.mp3"), 500)
+
+		cmd := exec.Command(tool, "-a", root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "Processed 6 files")
+		checkOut(t, sout, "- 3 directories were updated")
+		checkOut(t, sout, "- 6 file hashes were added")
+		checkOut(t, sout, "- 0 file hashes were updated")
+	})
+
+	// update remaining
+	t.Run("update-remaining-add", func(t *testing.T) {
+		cmd := exec.Command(tool, "-u", root)
+		out, err := cmd.Output()
+		if err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		}
+		sout := string(out)
+		checkOut(t, sout, "Processed 295 files")
+		checkOut(t, sout, "- 1 directory was updated")
+		checkOut(t, sout, "- 0 file hashes were added")
+		checkOut(t, sout, "- 1 file hash was updated")
+	})
 }
 
 func TestDMG(t *testing.T) {
@@ -161,50 +285,58 @@ func TestDMG(t *testing.T) {
 	t2, _ := time.Parse(time.RFC3339, "2022-02-01T12:00:00Z")
 	t3, _ := time.Parse(time.RFC3339, "2022-02-01T13:00:00Z")
 
-	// step1: create test and set the modified time"
-	os.WriteFile(testFile, []byte("foo1"), 0644)
-	os.Chtimes(testFile, t2, t2)
+	// create test and set the modified time"
+	t.Run("create", func(t *testing.T) {
+		os.WriteFile(testFile, []byte("foo1"), 0644)
+		os.Chtimes(testFile, t2, t2)
 
-	cmd := exec.Command(tool, "-u", ".")
-	if out, err := cmd.Output(); err != nil {
-		t.Fatalf("step1 failed with '%s'\n", err)
-	} else {
-		checkOut(t, string(out), "new test.txt")
-	}
-
-	// step2: update test with different content & old modified (expect 'old')"
-	os.WriteFile(testFile, []byte("foo2"), 0644)
-	os.Chtimes(testFile, t1, t1)
-
-	cmd = exec.Command(tool, "-u", ".")
-	if out, err := cmd.Output(); err != nil {
-		t.Fatalf("step2 failed with '%s'\n", err)
-	} else {
-		checkOut(t, string(out), "old test.txt")
-	}
-
-	// step3: update test & new modified (expect 'upd')"
-	os.WriteFile(testFile, []byte("foo3"), 0644)
-	os.Chtimes(testFile, t3, t3)
-
-	cmd = exec.Command(tool, "-u", ".")
-	if out, err := cmd.Output(); err != nil {
-		t.Fatalf("step3 failed with '%s'\n", err)
-	} else {
-		checkOut(t, string(out), "upd test.txt")
-	}
-
-	// step4: Now update test with the same modified to simulate damage (expect DMG)"
-	os.WriteFile(testFile, []byte("foo4"), 0644)
-	os.Chtimes(testFile, t3, t3)
-
-	cmd = exec.Command(tool, "-u", ".")
-	if out, err := cmd.Output(); err != nil {
-		if cmd.ProcessState.ExitCode() != 1 {
-			t.Fatalf("step4 expected to fail with exit code 1 vs %d!", cmd.ProcessState.ExitCode())
+		cmd := exec.Command(tool, "-u", ".")
+		if out, err := cmd.Output(); err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		} else {
+			checkOut(t, string(out), "new test.txt")
 		}
-		checkOut(t, string(out), "DMG test.txt")
-	} else {
-		t.Fatal("step4 expected to fail!")
-	}
+	})
+
+	// update test with different content & old modified (expect 'old')"
+	t.Run("expect-old", func(t *testing.T) {
+		os.WriteFile(testFile, []byte("foo2"), 0644)
+		os.Chtimes(testFile, t1, t1)
+
+		cmd := exec.Command(tool, "-u", ".")
+		if out, err := cmd.Output(); err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		} else {
+			checkOut(t, string(out), "old test.txt")
+		}
+	})
+
+	// update test & new modified (expect 'upd')"
+	t.Run("expect-upd", func(t *testing.T) {
+		os.WriteFile(testFile, []byte("foo3"), 0644)
+		os.Chtimes(testFile, t3, t3)
+
+		cmd := exec.Command(tool, "-u", ".")
+		if out, err := cmd.Output(); err != nil {
+			t.Fatalf("failed with '%s'\n", err)
+		} else {
+			checkOut(t, string(out), "upd test.txt")
+		}
+	})
+
+	// Now update test with the same modified to simulate damage (expect DMG)"
+	t.Run("expect-DMG", func(t *testing.T) {
+		os.WriteFile(testFile, []byte("foo4"), 0644)
+		os.Chtimes(testFile, t3, t3)
+
+		cmd := exec.Command(tool, "-u", ".")
+		if out, err := cmd.Output(); err != nil {
+			if cmd.ProcessState.ExitCode() != 1 {
+				t.Fatalf("expected to fail with exit code 1 vs %d!", cmd.ProcessState.ExitCode())
+			}
+			checkOut(t, string(out), "DMG test.txt")
+		} else {
+			t.Fatal("expected to fail!")
+		}
+	})
 }

scripts/tests

@@ -7,5 +7,5 @@ cd $script_dir/..
 # prep
 $script_dir/build
 
-go test -v ./cmd/chkbit/util
-go test -v ./scripts
+go test -v ./cmd/chkbit/util -count=1
+go test -v ./scripts -count=1

status.go

@@ -12,6 +12,7 @@ const (
 	STATUS_NEW          Status = "new"
 	STATUS_OK           Status = "ok "
 	STATUS_IGNORE       Status = "ign"
+	STATUS_MISSING      Status = "del"
 )
 
 func (s Status) String() string {

worker.go

@@ -3,17 +3,18 @@ package chkbit
 type WorkItem struct {
 	path         string
 	filesToIndex []string
+	dirList      []string
 	ignore       *Ignore
 }
 
-func (context *Context) RunWorker(id int) {
+func (context *Context) runWorker(id int) {
 	for {
 		item := <-context.WorkQueue
 		if item == nil {
 			break
 		}
 
-		index := NewIndex(context, item.path, item.filesToIndex, !context.Update)
+		index := newIndex(context, item.path, item.filesToIndex, item.dirList, !context.UpdateIndex)
 		err := index.load()
 		if err != nil {
 			context.log(STATUS_PANIC, index.getIndexFilepath()+": "+err.Error())
@@ -23,9 +24,9 @@ func (context *Context) RunWorker(id int) {
 			index.showIgnoredOnly(item.ignore)
 		} else {
 			index.calcHashes(item.ignore)
-			index.checkFix(context.Force)
+			index.checkFix(context.ForceUpdateDmg)
 
-			if context.Update {
+			if context.UpdateIndex {
 				if changed, err := index.save(); err != nil {
 					context.logErr(item.path, err)
 				} else if changed {