codec: make cbor and json codecs use ErrUnexpectedEOF

In particular, if the input doesn't have a single token, we would return
io.EOF, which can be easily mistaken by an error that can be ignored.
In particular, go-ipfs-cmds ignored it as such.

Make them return a clearly bad "unexpected EOF" error,
which mirrors the behavior found in many other unmarshal APIs
such as encoding/json.Unmarshal.

Finally, add a table-driven test that we can extend in the future,
which tests all four codecs.

For ipfs/kubo#8378.

codec/dagcbor/unmarshal.go

@@ -71,6 +71,9 @@ func Unmarshal(na datamodel.NodeAssembler, tokSrc shared.TokenSource, options De
 func unmarshal1(na datamodel.NodeAssembler, tokSrc shared.TokenSource, gas *int, options DecodeOptions) error {
 	var tk tok.Token
 	done, err := tokSrc.Step(&tk)
+	if err == io.EOF {
+		return io.ErrUnexpectedEOF
+	}
 	if err != nil {
 		return err
 	}

codec/dagjson/unmarshal.go

@@ -79,6 +79,9 @@ func Unmarshal(na datamodel.NodeAssembler, tokSrc shared.TokenSource, options De
 	var st unmarshalState
 	st.options = options
 	done, err := tokSrc.Step(&st.tk[0])
+	if err == io.EOF {
+		return io.ErrUnexpectedEOF
+	}
 	if err != nil {
 		return err
 	}

codec/decode_test.go

@@ -0,0 +1,38 @@
+package codec_test
+
+import (
+	"errors"
+	"io"
+	"strings"
+	"testing"
+
+	_ "github.com/ipld/go-ipld-prime/codec/cbor"
+	_ "github.com/ipld/go-ipld-prime/codec/dagcbor"
+	_ "github.com/ipld/go-ipld-prime/codec/dagjson"
+	_ "github.com/ipld/go-ipld-prime/codec/json"
+	mcregistry "github.com/ipld/go-ipld-prime/multicodec"
+	basicnode "github.com/ipld/go-ipld-prime/node/basic"
+	"github.com/multiformats/go-multicodec"
+)
+
+func TestDecodeZero(t *testing.T) {
+	for _, code := range []multicodec.Code{
+		multicodec.Cbor,
+		multicodec.DagCbor,
+		multicodec.Json,
+		multicodec.DagJson,
+	} {
+		t.Run(code.String(), func(t *testing.T) {
+			nb := basicnode.Prototype.Any.NewBuilder()
+			decode, err := mcregistry.LookupDecoder(uint64(code))
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			err = decode(nb, strings.NewReader(""))
+			if !errors.Is(err, io.ErrUnexpectedEOF) {
+				t.Fatalf("unexpected error: %v", err)
+			}
+		})
+	}
+}

go.mod

@@ -5,6 +5,7 @@ go 1.16
 require (
 	github.com/frankban/quicktest v1.11.3
 	github.com/ipfs/go-cid v0.0.4
+	github.com/multiformats/go-multicodec v0.3.0
 	github.com/multiformats/go-multihash v0.0.15
 	github.com/polydawn/refmt v0.0.0-20201211092308-30ac6d18308e
 	github.com/smartystreets/goconvey v1.6.4 // indirect

go.sum

@@ -28,6 +28,8 @@ github.com/multiformats/go-base32 v0.0.3 h1:tw5+NhuwaOjJCC5Pp82QuXbrmLzWg7uxlMFp
 github.com/multiformats/go-base32 v0.0.3/go.mod h1:pLiuGC8y0QR3Ue4Zug5UzK9LjgbkL8NSQj0zQ5Nz/AA=
 github.com/multiformats/go-multibase v0.0.1 h1:PN9/v21eLywrFWdFNsFKaU04kLJzuYzmrJR+ubhT9qA=
 github.com/multiformats/go-multibase v0.0.1/go.mod h1:bja2MqRZ3ggyXtZSEDKpl0uO/gviWFaSteVbWT51qgs=
+github.com/multiformats/go-multicodec v0.3.0 h1:tstDwfIjiHbnIjeM5Lp+pMrSeN+LCMsEwOrkPmWm03A=
+github.com/multiformats/go-multicodec v0.3.0/go.mod h1:qGGaQmioCDh+TeFOnxrbU0DaIPw8yFgAZgFG0V7p1qQ=
 github.com/multiformats/go-multihash v0.0.10/go.mod h1:YSLudS+Pi8NHE7o6tb3D8vrpKa63epEDmG8nTduyAew=
 github.com/multiformats/go-multihash v0.0.15 h1:hWOPdrNqDjwHDx82vsYGSDZNyktOJJ2dzZJzFkOV1jM=
 github.com/multiformats/go-multihash v0.0.15/go.mod h1:D6aZrWNLFTV/ynMpKsNtB40mJzmCl4jb1alC0OvHiHg=

multicodec/registry.go

@@ -70,6 +70,8 @@ func (r *Registry) ListEncoders() []uint64 {
 	return encoders
 }
 
+// TODO(mvdan): turn most of these uint64s into multicodec.Code
+
 // RegisterDecoder updates a simple map of multicodec indicator number to codec.Decoder function.
 // The decoder functions registered can be subsequently looked up using LookupDecoder.
 func (r *Registry) RegisterDecoder(indicator uint64, decodeFunc codec.Decoder) {