feat: support for reading gzip,bzip2 and xz files

dundee · dundee · commit c66dcde41e78 · 2024-05-31T22:50:42.000+02:00
diff --git a/go.mod b/go.mod
@@ -6,6 +6,7 @@ require (
 	github.com/dgraph-io/badger/v3 v3.2103.2
 	github.com/fatih/color v1.16.0
 	github.com/gdamore/tcell/v2 v2.7.1
+	github.com/h2non/filetype v1.1.3
 	github.com/maruel/natural v1.1.0
 	github.com/mattn/go-isatty v0.0.20
 	github.com/pbnjay/memory v0.0.0-20210728143218-7b4eea64cf58
@@ -14,6 +15,7 @@ require (
 	github.com/sirupsen/logrus v1.9.3
 	github.com/spf13/cobra v1.8.0
 	github.com/stretchr/testify v1.9.0
+	github.com/ulikunitz/xz v0.5.12
 	golang.org/x/exp v0.0.0-20240205201215-2c58cdc269a3
 	golang.org/x/sys v0.20.0
 	gopkg.in/yaml.v3 v3.0.1
diff --git a/go.sum b/go.sum
@@ -48,6 +48,8 @@ github.com/google/flatbuffers v1.12.1 h1:MVlul7pQNoDzWRLTw5imwYsl+usrS1TXG2H4jg6
 github.com/google/flatbuffers v1.12.1/go.mod h1:1AeVuKshWv4vARoZatz6mlQ0JxURH0Kv5+zNeJKJCa8=
 github.com/google/go-cmp v0.3.0/go.mod h1:8QqcDgzrUqlUb/G2PQTWiueGozuR1884gddMywk6iLU=
 github.com/google/go-cmp v0.5.4/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
+github.com/h2non/filetype v1.1.3 h1:FKkx9QbD7HR/zjK1Ia5XiBsq9zdLi5Kf3zGyFTAFkGg=
+github.com/h2non/filetype v1.1.3/go.mod h1:319b3zT68BvV+WRj7cwy856M2ehB3HqNOt6sy1HndBY=
 github.com/hashicorp/hcl v1.0.0/go.mod h1:E5yfLk+7swimpb2L/Alb/PJmXilQ/rhwaUYs4T20WEQ=
 github.com/inconshreveable/mousetrap v1.0.0/go.mod h1:PxqpIevigyE2G7u3NXJIT2ANytuPF1OarO4DADm73n8=
 github.com/inconshreveable/mousetrap v1.1.0 h1:wN+x4NVGpMsO7ErUn/mUI3vEoE6Jt13X2s0bqwp9tc8=
@@ -116,6 +118,8 @@ github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/
 github.com/stretchr/testify v1.9.0 h1:HtqpIVDClZ4nwg75+f6Lvsy/wHu+3BoSGCbBAcpTsTg=
 github.com/stretchr/testify v1.9.0/go.mod h1:r2ic/lqez/lEtzL7wO/rwa5dbSLXVDPFyf8C91i36aY=
 github.com/ugorji/go/codec v0.0.0-20181204163529-d75b2dcb6bc8/go.mod h1:VFNgLljTbGfSG7qAOspJ7OScBnGdDN/yBr0sguwnwf0=
+github.com/ulikunitz/xz v0.5.12 h1:37Nm15o69RwBkXM0J6A5OlE67RZTfzUxTj8fB3dfcsc=
+github.com/ulikunitz/xz v0.5.12/go.mod h1:nbz6k7qbPmH4IRqmfOplQw/tblSgqTqBwxkY0oWt/14=
 github.com/xordataexchange/crypt v0.0.3-0.20170626215501-b2862e3d0a77/go.mod h1:aYKd//L2LvnjZzWKhF00oedf4jCCReLcmhLdhm1A27Q=
 github.com/yuin/goldmark v1.1.27/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
diff --git a/tui/show_file.go b/tui/show_file.go
@@ -2,11 +2,18 @@ package tui
 
 import (
 	"bufio"
+	"compress/bzip2"
+	"compress/gzip"
+	"io"
 	"os"
 	"strings"
 
 	"github.com/gdamore/tcell/v2"
+	"github.com/h2non/filetype"
+	"github.com/h2non/filetype/matchers"
+	"github.com/pkg/errors"
 	"github.com/rivo/tview"
+	"github.com/ulikunitz/xz"
 
 	"github.com/dundee/gdu/v5/build"
 	"github.com/dundee/gdu/v5/pkg/fs"
@@ -23,18 +30,23 @@ func (ui *UI) showFile() *tview.TextView {
 		return nil
 	}
 
-	f, err := os.Open(selectedFile.GetPath())
+	path := selectedFile.GetPath()
+	f, err := os.Open(path)
 	if err != nil {
 		ui.showErr("Error opening file", err)
 		return nil
 	}
+	scanner, err := getScanner(f)
+	if err != nil {
+		ui.showErr("Error reading file", err)
+		return nil
+	}
 
 	totalLines := 0
-	scanner := bufio.NewScanner(f)
 
 	file := tview.NewTextView()
 	ui.currentDirLabel.SetText("[::b] --- " +
-		strings.TrimPrefix(selectedFile.GetPath(), build.RootPathPrefix) +
+		strings.TrimPrefix(path, build.RootPathPrefix) +
 		" ---").SetDynamicColors(true)
 
 	readNextPart := func(linesCount int) int {
@@ -94,3 +106,41 @@ func (ui *UI) showFile() *tview.TextView {
 
 	return file
 }
+
+func getScanner(f io.ReadSeeker) (scanner *bufio.Scanner, err error) {
+	// We only have to pass the file header = first 261 bytes
+	head := make([]byte, 261)
+	if _, err = f.Read(head); err != nil {
+		return nil, errors.Wrap(err, "error reading file header")
+	}
+
+	if pos, err := f.Seek(0, 0); pos != 0 || err != nil {
+		return nil, errors.Wrap(err, "error seeking file")
+	}
+	scanner = bufio.NewScanner(f)
+
+	typ, err := filetype.Match(head)
+	if err != nil {
+		return nil, errors.Wrap(err, "error matching file type")
+	}
+
+	switch typ.MIME.Value {
+	case matchers.TypeGz.MIME.Value:
+		r, err := gzip.NewReader(f)
+		if err != nil {
+			return nil, errors.Wrap(err, "error creating gzip reader")
+		}
+		scanner = bufio.NewScanner(r)
+	case matchers.TypeBz2.MIME.Value:
+		r := bzip2.NewReader(f)
+		scanner = bufio.NewScanner(r)
+	case matchers.TypeXz.MIME.Value:
+		r, err := xz.NewReader(f)
+		if err != nil {
+			return nil, errors.Wrap(err, "error creating xz reader")
+		}
+		scanner = bufio.NewScanner(r)
+	}
+
+	return scanner, nil
+}
diff --git a/tui/show_file_test.go b/tui/show_file_test.go
@@ -0,0 +1,83 @@
+package tui
+
+import (
+	"bytes"
+	"compress/gzip"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/ulikunitz/xz"
+)
+
+func TestGetScannerForPlainString(t *testing.T) {
+	r := bytes.NewReader([]byte("hello"))
+	s, err := getScanner(r)
+	assert.Nil(t, err)
+
+	assert.Equal(t, true, s.Scan())
+	assert.Equal(t, "hello", s.Text())
+	assert.Equal(t, nil, s.Err())
+}
+
+func TestGetScannerForGzipped(t *testing.T) {
+	b := bytes.NewBuffer([]byte{})
+	w := gzip.NewWriter(b)
+
+	_, err := w.Write([]byte("hello world"))
+	assert.Nil(t, err)
+
+	err = w.Close()
+	assert.Nil(t, err)
+
+	r := bytes.NewReader(b.Bytes())
+	s, err := getScanner(r)
+	assert.Nil(t, err)
+
+	assert.Equal(t, true, s.Scan())
+	assert.Equal(t, "hello world", s.Text())
+	assert.Equal(t, nil, s.Err())
+}
+
+func TestGetScannerForBzipped(t *testing.T) {
+	r := bytes.NewReader([]byte{
+		// bzip2 header
+		0x42, 0x5A, 0x68, 0x39,
+		// bzip2 compressed data: "hello"
+		0x31, 0x41, 0x59, 0x26,
+		0x53, 0x59, 0xC1, 0xC0,
+		0x80, 0xE2, 0x00, 0x00,
+		0x01, 0x41, 0x00, 0x00,
+		0x10, 0x02, 0x44, 0xA0,
+		0x00, 0x30, 0xCD, 0x00,
+		0xC3, 0x46, 0x29, 0x97,
+		0x17, 0x72, 0x45, 0x38,
+		0x50, 0x90, 0xC1, 0xC0,
+		0x80, 0xE2,
+	})
+	s, err := getScanner(r)
+	assert.Nil(t, err)
+
+	assert.Equal(t, true, s.Scan())
+	assert.Equal(t, "hello", s.Text())
+	assert.Equal(t, nil, s.Err())
+}
+
+func TestGetScannerForXzipped(t *testing.T) {
+	b := bytes.NewBuffer([]byte{})
+	w, err := xz.NewWriter(b)
+	assert.Nil(t, err)
+
+	_, err = w.Write([]byte("hello world"))
+	assert.Nil(t, err)
+
+	err = w.Close()
+	assert.Nil(t, err)
+
+	r := bytes.NewReader(b.Bytes())
+	s, err := getScanner(r)
+	assert.Nil(t, err)
+
+	assert.Equal(t, true, s.Scan())
+	assert.Equal(t, "hello world", s.Text())
+	assert.Equal(t, nil, s.Err())
+}
