tests: Port flate fuzz test (#660)

Author: klauspost

flate/fuzz_test.go

@@ -0,0 +1,113 @@
+//go:build go1.18
+// +build go1.18
+
+package flate
+
+import (
+	"bytes"
+	"io"
+	"strconv"
+	"testing"
+
+	"github.com/klauspost/compress/internal/fuzz"
+)
+
+func FuzzEncoding(f *testing.F) {
+	fuzz.AddFromZip(f, "testdata/regression.zip", true, false)
+	fuzz.AddFromZip(f, "testdata/fuzz/encode-raw-corpus.zip", true, testing.Short())
+	fuzz.AddFromZip(f, "testdata/fuzz/FuzzEncoding.zip", false, testing.Short())
+	// Fuzzing tweaks:
+	const (
+		// Test a subset of encoders.
+		startFuzz = HuffmanOnly
+		endFuzz   = BestCompression
+
+		// Also tests with dictionaries...
+		testDicts = true
+
+		// Max input size:
+		maxSize = 1 << 20
+	)
+	decoder := NewReader(nil)
+	buf := new(bytes.Buffer)
+
+	f.Fuzz(func(t *testing.T, data []byte) {
+		if len(data) > maxSize {
+			return
+		}
+		for level := startFuzz; level <= endFuzz; level++ {
+			msg := "level " + strconv.Itoa(level) + ":"
+			buf.Reset()
+			fw, err := NewWriter(buf, level)
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			n, err := fw.Write(data)
+			if n != len(data) {
+				t.Fatal(msg + "short write")
+			}
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			err = fw.Close()
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			decoder.(Resetter).Reset(buf, nil)
+			data2, err := io.ReadAll(decoder)
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			if !bytes.Equal(data, data2) {
+				t.Fatal(msg + "not equal")
+			}
+			// Do it again...
+			msg = "level " + strconv.Itoa(level) + " (reset):"
+			buf.Reset()
+			fw.Reset(buf)
+			n, err = fw.Write(data)
+			if n != len(data) {
+				t.Fatal(msg + "short write")
+			}
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			err = fw.Close()
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			decoder.(Resetter).Reset(buf, nil)
+			data2, err = io.ReadAll(decoder)
+			if err != nil {
+				t.Fatal(msg + err.Error())
+			}
+			if !bytes.Equal(data, data2) {
+				t.Fatal(msg + "not equal")
+			}
+		}
+
+		// Split into two and use history...
+		buf.Reset()
+		err := StatelessDeflate(buf, data[:len(data)/2], false, nil)
+		if err != nil {
+			t.Error(err)
+		}
+
+		// Use top half as dictionary...
+		dict := data[:len(data)/2]
+		err = StatelessDeflate(buf, data[len(data)/2:], true, dict)
+		if err != nil {
+			t.Error(err)
+		}
+
+		decoder.(Resetter).Reset(buf, nil)
+		data2, err := io.ReadAll(decoder)
+		if err != nil {
+			t.Error(err)
+		}
+		if !bytes.Equal(data, data2) {
+			//fmt.Printf("want:%x\ngot: %x\n", data1, data2)
+			t.Error("not equal")
+		}
+	})
+}

flate/testdata/fuzz/FuzzEncoding.zip

@@ -0,0 +1,124 @@
+//go:build go1.18
+// +build go1.18
+
+package fuzz
+
+import (
+	"archive/zip"
+	"bytes"
+	"fmt"
+	"go/ast"
+	"go/parser"
+	"go/token"
+	"io"
+	"os"
+	"strconv"
+	"testing"
+)
+
+// AddFromZip will read the supplied zip and add all as corpus for f.
+func AddFromZip(f *testing.F, filename string, raw, short bool) {
+	file, err := os.Open(filename)
+	if err != nil {
+		f.Fatal(err)
+	}
+	fi, err := file.Stat()
+	if err != nil {
+		f.Fatal(err)
+	}
+	zr, err := zip.NewReader(file, fi.Size())
+	if err != nil {
+		f.Fatal(err)
+	}
+	for i, file := range zr.File {
+		if short && i%10 != 0 {
+			continue
+		}
+		rc, err := file.Open()
+		if err != nil {
+			f.Fatal(err)
+		}
+
+		b, err := io.ReadAll(rc)
+		if err != nil {
+			f.Fatal(err)
+		}
+		rc.Close()
+		raw := raw
+		if bytes.HasPrefix(b, []byte("go test fuzz")) {
+			raw = false
+		}
+		if raw {
+			f.Add(b)
+			continue
+		}
+		vals, err := unmarshalCorpusFile(b)
+		if err != nil {
+			f.Fatal(err)
+		}
+		for _, v := range vals {
+			f.Add(v)
+		}
+	}
+}
+
+// unmarshalCorpusFile decodes corpus bytes into their respective values.
+func unmarshalCorpusFile(b []byte) ([][]byte, error) {
+	if len(b) == 0 {
+		return nil, fmt.Errorf("cannot unmarshal empty string")
+	}
+	lines := bytes.Split(b, []byte("\n"))
+	if len(lines) < 2 {
+		return nil, fmt.Errorf("must include version and at least one value")
+	}
+	var vals = make([][]byte, 0, len(lines)-1)
+	for _, line := range lines[1:] {
+		line = bytes.TrimSpace(line)
+		if len(line) == 0 {
+			continue
+		}
+		v, err := parseCorpusValue(line)
+		if err != nil {
+			return nil, fmt.Errorf("malformed line %q: %v", line, err)
+		}
+		vals = append(vals, v)
+	}
+	return vals, nil
+}
+
+// parseCorpusValue
+func parseCorpusValue(line []byte) ([]byte, error) {
+	fs := token.NewFileSet()
+	expr, err := parser.ParseExprFrom(fs, "(test)", line, 0)
+	if err != nil {
+		return nil, err
+	}
+	call, ok := expr.(*ast.CallExpr)
+	if !ok {
+		return nil, fmt.Errorf("expected call expression")
+	}
+	if len(call.Args) != 1 {
+		return nil, fmt.Errorf("expected call expression with 1 argument; got %d", len(call.Args))
+	}
+	arg := call.Args[0]
+
+	if arrayType, ok := call.Fun.(*ast.ArrayType); ok {
+		if arrayType.Len != nil {
+			return nil, fmt.Errorf("expected []byte or primitive type")
+		}
+		elt, ok := arrayType.Elt.(*ast.Ident)
+		if !ok || elt.Name != "byte" {
+			return nil, fmt.Errorf("expected []byte")
+		}
+		lit, ok := arg.(*ast.BasicLit)
+		if !ok || lit.Kind != token.STRING {
+			return nil, fmt.Errorf("string literal required for type []byte")
+		}
+		s, err := strconv.Unquote(lit.Value)
+		if err != nil {
+			return nil, err
+		}
+		return []byte(s), nil
+	}
+	return nil, fmt.Errorf("expected []byte")
+}