feat: added option to follow symlinks (#206)

* feat: follow symlinks in the analyzer

* feat: set followSymlinks

* added follow-symlinks parameter

* test for following symlinks

* test broken symlink

* refactor

* better doc

* test SetFollowSymlinks

* refactor

* readme updated

10 files changed, 169 insertions, 6 deletions

diff --git a/README.md b/README.md
index 5114342..ac0bbc8 100644
--- a/README.md
+++ b/README.md
@@ -74,6 +74,7 @@ Flags:
       --config-file string            Read config from file (default is $HOME/.gdu.yaml)
   -g, --const-gc                      Enable memory garbage collection during analysis with constant level set by GOGC
       --enable-profiling              Enable collection of profiling data and provide it on http://localhost:6060/debug/pprof/
+  -L, --follow-symlinks               Follow symlinks for files, i.e. show the size of the file to which symlink points to (symlinks to directories are not followed)
   -h, --help                          help for gdu
   -i, --ignore-dirs strings           Absolute paths to ignore (separated by comma) (default [/proc,/dev,/sys,/run])
   -I, --ignore-dirs-pattern strings   Absolute path patterns to ignore (separated by comma)
diff --git a/cmd/gdu/app/app.go b/cmd/gdu/app/app.go
index bd6440f..63afcd8 100644
--- a/cmd/gdu/app/app.go
+++ b/cmd/gdu/app/app.go
@@ -34,6 +34,7 @@ type UI interface {
 	SetIgnoreDirPatterns(paths []string) error
 	SetIgnoreFromFile(ignoreFile string) error
 	SetIgnoreHidden(value bool)
+	SetFollowSymlinks(value bool)
 	StartUILoop() error
 }
 
@@ -57,6 +58,7 @@ type Flags struct {
 	NoProgress        bool     `yaml:"no-progress"`
 	NoCross           bool     `yaml:"no-cross"`
 	NoHidden          bool     `yaml:"no-hidden"`
+	FollowSymlinks    bool     `yaml:"follow-symlinks"`
 	Profiling         bool     `yaml:"profiling"`
 	ConstGC           bool     `yaml:"const-gc"`
 	Summarize         bool     `yaml:"summarize"`
@@ -203,10 +205,7 @@ func (a *App) createUI() (UI, error) {
 			a.Flags.ConstGC,
 			a.Flags.UseSIPrefix,
 		)
-		return ui, nil
-	}
-
-	if a.Flags.NonInteractive || !a.Istty {
+	} else if a.Flags.NonInteractive || !a.Istty {
 		ui = stdout.CreateStdoutUI(
 			a.Writer,
 			!a.Flags.NoColor && a.Istty,
@@ -261,6 +260,11 @@ func (a *App) createUI() (UI, error) {
 		}
 		tview.Styles.BorderColor = tcell.ColorDefault
 	}
+
+	if a.Flags.FollowSymlinks {
+		ui.SetFollowSymlinks(true)
+	}
+
 	return ui, nil
 }
 
diff --git a/cmd/gdu/app/app_test.go b/cmd/gdu/app/app_test.go
index 746550e..947f453 100644
--- a/cmd/gdu/app/app_test.go
+++ b/cmd/gdu/app/app_test.go
@@ -47,6 +47,21 @@ func TestAnalyzePath(t *testing.T) {
 	assert.Nil(t, err)
 }
 
+func TestFollowSymlinks(t *testing.T) {
+	fin := testdir.CreateTestDir()
+	defer fin()
+
+	out, err := runApp(
+		&Flags{LogFile: "/dev/null", FollowSymlinks: true},
+		[]string{"test_dir"},
+		false,
+		testdev.DevicesInfoGetterMock{},
+	)
+
+	assert.Contains(t, out, "nested")
+	assert.Nil(t, err)
+}
+
 func TestAnalyzePathProfiling(t *testing.T) {
 	fin := testdir.CreateTestDir()
 	defer fin()
diff --git a/cmd/gdu/main.go b/cmd/gdu/main.go
index b0c9471..5ac5e19 100644
--- a/cmd/gdu/main.go
+++ b/cmd/gdu/main.go
@@ -49,6 +49,10 @@ func init() {
 	flags.StringSliceVarP(&af.IgnoreDirPatterns, "ignore-dirs-pattern", "I", []string{}, "Absolute path patterns to ignore (separated by comma)")
 	flags.StringVarP(&af.IgnoreFromFile, "ignore-from", "X", "", "Read absolute path patterns to ignore from file")
 	flags.BoolVarP(&af.NoHidden, "no-hidden", "H", false, "Ignore hidden directories (beginning with dot)")
+	flags.BoolVarP(
+		&af.FollowSymlinks, "follow-symlinks", "L", false,
+		"Follow symlinks for files, i.e. show the size of the file to which symlink points to (symlinks to directories are not followed)",
+	)
 	flags.BoolVarP(&af.NoCross, "no-cross", "x", false, "Do not cross filesystem boundaries")
 	flags.BoolVarP(&af.ConstGC, "const-gc", "g", false, "Enable memory garbage collection during analysis with constant level set by GOGC")
 	flags.BoolVar(&af.Profiling, "enable-profiling", false, "Enable collection of profiling data and provide it on http://localhost:6060/debug/pprof/")
diff --git a/internal/common/analyze.go b/internal/common/analyze.go
index 152bd5e..bb08b3c 100644
--- a/internal/common/analyze.go
+++ b/internal/common/analyze.go
@@ -15,6 +15,7 @@ type ShouldDirBeIgnored func(name, path string) bool
 // Analyzer is type for dir analyzing function
 type Analyzer interface {
 	AnalyzeDir(path string, ignore ShouldDirBeIgnored, constGC bool) fs.Item
+	SetFollowSymlinks(bool)
 	GetProgressChan() chan CurrentProgress
 	GetDone() SignalGroup
 	ResetProgress()
diff --git a/internal/common/ui.go b/internal/common/ui.go
index 0553bba..d094a67 100644
--- a/internal/common/ui.go
+++ b/internal/common/ui.go
@@ -19,6 +19,11 @@ type UI struct {
 	ConstGC               bool
 }
 
+// SetFollowSymlinks sets whether symlinks to files should be followed
+func (ui *UI) SetFollowSymlinks(v bool) {
+	ui.Analyzer.SetFollowSymlinks(v)
+}
+
 // binary multiplies prefixes (IEC)
 const (
 	_          = iota
diff --git a/internal/common/ui_test.go b/internal/common/ui_test.go
index 3a6c28a..e669a61 100644
--- a/internal/common/ui_test.go
+++ b/internal/common/ui_test.go
@@ -3,6 +3,7 @@ package common
 import (
 	"testing"
 
+	"github.com/dundee/gdu/v5/pkg/fs"
 	"github.com/stretchr/testify/assert"
 )
 
@@ -10,3 +11,43 @@ func TestFormatNumber(t *testing.T) {
 	res := FormatNumber(1234567890)
 	assert.Equal(t, "1,234,567,890", res)
 }
+
+func TestSetFollowSymlinks(t *testing.T) {
+	ui := UI{
+		Analyzer: &MockedAnalyzer{},
+	}
+	ui.SetFollowSymlinks(true)
+
+	assert.Equal(t, true, ui.Analyzer.(*MockedAnalyzer).FollowSymlinks)
+}
+
+type MockedAnalyzer struct {
+	FollowSymlinks bool
+}
+
+// AnalyzeDir returns dir with files with different size exponents
+func (a *MockedAnalyzer) AnalyzeDir(
+	path string, ignore ShouldDirBeIgnored, enableGC bool,
+) fs.Item {
+	return nil
+}
+
+// GetProgressChan returns always Done
+func (a *MockedAnalyzer) GetProgressChan() chan CurrentProgress {
+	return make(chan CurrentProgress)
+}
+
+// GetDone returns always Done
+func (a *MockedAnalyzer) GetDone() SignalGroup {
+	c := make(SignalGroup)
+	defer c.Broadcast()
+	return c
+}
+
+// ResetProgress does nothing
+func (a *MockedAnalyzer) ResetProgress() {}
+
+// SetFollowSymlinks does nothing
+func (a *MockedAnalyzer) SetFollowSymlinks(v bool) {
+	a.FollowSymlinks = v
+}
diff --git a/internal/testanalyze/analyze.go b/internal/testanalyze/analyze.go
index cd61cc7..89ca4cf 100644
--- a/internal/testanalyze/analyze.go
+++ b/internal/testanalyze/analyze.go
@@ -70,7 +70,7 @@ func (a *MockedAnalyzer) GetProgressChan() chan common.CurrentProgress {
 	return make(chan common.CurrentProgress)
 }
 
-// GetDoneChan returns always Done
+// GetDone returns always Done
 func (a *MockedAnalyzer) GetDone() common.SignalGroup {
 	c := make(common.SignalGroup)
 	defer c.Broadcast()
@@ -80,6 +80,9 @@ func (a *MockedAnalyzer) GetDone() common.SignalGroup {
 // ResetProgress does nothing
 func (a *MockedAnalyzer) ResetProgress() {}
 
+// SetFollowSymlinks does nothing
+func (a *MockedAnalyzer) SetFollowSymlinks(v bool) {}
+
 // RemoveItemFromDirWithErr returns error
 func RemoveItemFromDirWithErr(dir fs.Item, file fs.Item) error {
 	return errors.New("Failed")
diff --git a/pkg/analyze/dir.go b/pkg/analyze/dir.go
index 08c0b43..eedce4d 100644
--- a/pkg/analyze/dir.go
+++ b/pkg/analyze/dir.go
@@ -22,6 +22,7 @@ type ParallelAnalyzer struct {
 	doneChan         common.SignalGroup
 	wait             *WaitGroup
 	ignoreDir        common.ShouldDirBeIgnored
+	followSymlinks   bool
 }
 
 // CreateAnalyzer returns Analyzer
@@ -39,12 +40,17 @@ func CreateAnalyzer() *ParallelAnalyzer {
 	}
 }
 
+// SetFollowSymlinks sets whether symlink to files should be followed
+func (a *ParallelAnalyzer) SetFollowSymlinks(v bool) {
+	a.followSymlinks = v
+}
+
 // GetProgressChan returns channel for getting progress
 func (a *ParallelAnalyzer) GetProgressChan() chan common.CurrentProgress {
 	return a.progressOutChan
 }
 
-// GetDoneChan returns channel for checking when analysis is done
+// GetDone returns channel for checking when analysis is done
 func (a *ParallelAnalyzer) GetDone() common.SignalGroup {
 	return a.doneChan
 }
@@ -130,8 +136,18 @@ func (a *ParallelAnalyzer) processDir(path string) *Dir {
 			info, err = f.Info()
 			if err != nil {
 				log.Print(err.Error())
+				dir.Flag = '!'
 				continue
 			}
+			if a.followSymlinks && info.Mode()&os.ModeSymlink != 0 {
+				err = followSymlink(entryPath, &info)
+				if err != nil {
+					log.Print(err.Error())
+					dir.Flag = '!'
+					continue
+				}
+			}
+
 			file = &File{
 				Name:   name,
 				Flag:   getFlag(info),
@@ -204,3 +220,18 @@ func getFlag(f os.FileInfo) rune {
 		return ' '
 	}
 }
+
+func followSymlink(path string, f *os.FileInfo) error {
+	target, err := filepath.EvalSymlinks(path)
+	if err != nil {
+		return err
+	}
+	tInfo, err := os.Lstat(target)
+	if err != nil {
+		return err
+	}
+	if !tInfo.IsDir() {
+		*f = tInfo
+	}
+	return nil
+}
diff --git a/pkg/analyze/dir_test.go b/pkg/analyze/dir_test.go
index 61ac0d8..89f6f86 100644
--- a/pkg/analyze/dir_test.go
+++ b/pkg/analyze/dir_test.go
@@ -133,6 +133,64 @@ func TestHardlink(t *testing.T) {
 	assert.Equal(t, 'H', dir.Files[0].(*Dir).Files[1].GetFlag())
 }
 
+func TestFollowSymlink(t *testing.T) {
+	fin := testdir.CreateTestDir()
+	defer fin()
+
+	err := os.Mkdir("test_dir/empty", 0644)
+	assert.Nil(t, err)
+
+	err = os.Symlink("./file2", "test_dir/nested/file3")
+	assert.Nil(t, err)
+
+	analyzer := CreateAnalyzer()
+	analyzer.SetFollowSymlinks(true)
+	dir := analyzer.AnalyzeDir(
+		"test_dir", func(_, _ string) bool { return false }, false,
+	).(*Dir)
+	analyzer.GetDone().Wait()
+	dir.UpdateStats(make(fs.HardLinkedItems))
+
+	sort.Sort(sort.Reverse(dir.Files))
+
+	assert.Equal(t, int64(9+4096*4), dir.Size)
+	assert.Equal(t, 7, dir.ItemCount)
+
+	// test file3
+	assert.Equal(t, "nested", dir.Files[0].GetName())
+	assert.Equal(t, "file3", dir.Files[0].(*Dir).Files[1].GetName())
+	assert.Equal(t, int64(2), dir.Files[0].(*Dir).Files[1].GetSize())
+	assert.Equal(t, ' ', dir.Files[0].(*Dir).Files[1].GetFlag())
+
+	assert.Equal(t, 'e', dir.Files[1].GetFlag())
+}
+
+func TestBrokenSymlinkSkipped(t *testing.T) {
+	fin := testdir.CreateTestDir()
+	defer fin()
+
+	err := os.Mkdir("test_dir/empty", 0644)
+	assert.Nil(t, err)
+
+	err = os.Symlink("xxx", "test_dir/nested/file3")
+	assert.Nil(t, err)
+
+	analyzer := CreateAnalyzer()
+	analyzer.SetFollowSymlinks(true)
+	dir := analyzer.AnalyzeDir(
+		"test_dir", func(_, _ string) bool { return false }, false,
+	).(*Dir)
+	analyzer.GetDone().Wait()
+	dir.UpdateStats(make(fs.HardLinkedItems))
+
+	sort.Sort(sort.Reverse(dir.Files))
+
+	assert.Equal(t, int64(7+4096*4), dir.Size)
+	assert.Equal(t, 6, dir.ItemCount)
+
+	assert.Equal(t, '!', dir.Files[0].GetFlag())
+}
+
 func BenchmarkAnalyzeDir(b *testing.B) {
 	fin := testdir.CreateTestDir()
 	defer fin()