encoding/json: fix Decoder.InputOffset regression in goexperiment.jsonv2

The Decoder.InputOffset method was always ambiguous about the exact
offset returned since anything between the end of the previous token
to the start of the next token could be a valid result.

Empirically, it seems that the behavior was to report
the end of the previous token unless Decoder.More is called,
in which case it reports the start of the next token.

This is an odd semantic since a relatively side-effect free method
like More is not quite so side-effect free.
However, our goal is to preserve historical v1 semantic when possible
regardless of whether it made sense.

Note that jsontext.Decoder.InputOffset consistently always reports
the end of the previous token. Users can explicitly choose the
exact position they want by inspecting the UnreadBuffer.

Fixes #75468

Change-Id: I1e946e83c9d29dfc09f2913ff8d6b2b80632f292
Reviewed-on: https://go-review.googlesource.com/c/go/+/703856
Reviewed-by: Damien Neil <dneil@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>

3 files changed, 149 insertions, 1 deletions

diff --git a/src/encoding/json/stream_test.go b/src/encoding/json/stream_test.go
index 9e5d48d39d..0e937cfaa1 100644
--- a/src/encoding/json/stream_test.go
+++ b/src/encoding/json/stream_test.go
@@ -557,3 +557,67 @@ func TestTokenTruncation(t *testing.T) {
 		}
 	}
 }
+
+func TestDecoderInputOffset(t *testing.T) {
+	const input = ` [
+		[ ] , [ "one" ] , [ "one" , "two" ] ,
+		{ } , { "alpha" : "bravo" } , { "alpha" : "bravo" , "fizz" : "buzz" }
+	] `
+	wantOffsets := []int64{
+		0, 1, 2, 5, 6, 7, 8, 9, 12, 13, 18, 19, 20, 21, 24, 25, 30, 31,
+		38, 39, 40, 41, 46, 47, 48, 49, 52, 53, 60, 61, 70, 71, 72, 73,
+		76, 77, 84, 85, 94, 95, 103, 104, 112, 113, 114, 116, 117, 117,
+		117, 117,
+	}
+	wantMores := []bool{
+		true, true, false, true, true, false, true, true, true, false,
+		true, false, true, true, true, false, true, true, true, true,
+		true, false, false, false, false,
+	}
+
+	d := NewDecoder(strings.NewReader(input))
+	checkOffset := func() {
+		t.Helper()
+		got := d.InputOffset()
+		if len(wantOffsets) == 0 {
+			t.Fatalf("InputOffset = %d, want nil", got)
+		}
+		want := wantOffsets[0]
+		if got != want {
+			t.Fatalf("InputOffset = %d, want %d", got, want)
+		}
+		wantOffsets = wantOffsets[1:]
+	}
+	checkMore := func() {
+		t.Helper()
+		got := d.More()
+		if len(wantMores) == 0 {
+			t.Fatalf("More = %v, want nil", got)
+		}
+		want := wantMores[0]
+		if got != want {
+			t.Fatalf("More = %v, want %v", got, want)
+		}
+		wantMores = wantMores[1:]
+	}
+	checkOffset()
+	checkMore()
+	checkOffset()
+	for {
+		if _, err := d.Token(); err == io.EOF {
+			break
+		} else if err != nil {
+			t.Fatalf("Token error: %v", err)
+		}
+		checkOffset()
+		checkMore()
+		checkOffset()
+	}
+	checkOffset()
+	checkMore()
+	checkOffset()
+
+	if len(wantOffsets)+len(wantMores) > 0 {
+		t.Fatal("unconsumed testdata")
+	}
+}
diff --git a/src/encoding/json/v2_stream.go b/src/encoding/json/v2_stream.go
index 28e72c0a52..dcee553ee1 100644
--- a/src/encoding/json/v2_stream.go
+++ b/src/encoding/json/v2_stream.go
@@ -20,6 +20,10 @@ type Decoder struct {
 	dec  *jsontext.Decoder
 	opts jsonv2.Options
 	err  error
+
+	// hadPeeked reports whether [Decoder.More] was called.
+	// It is reset by [Decoder.Decode] and [Decoder.Token].
+	hadPeeked bool
 }
 
 // NewDecoder returns a new decoder that reads from r.
@@ -76,6 +80,7 @@ func (dec *Decoder) Decode(v any) error {
 		}
 		return dec.err
 	}
+	dec.hadPeeked = false
 	return jsonv2.Unmarshal(b, v, dec.opts)
 }
 
@@ -206,6 +211,7 @@ func (dec *Decoder) Token() (Token, error) {
 		}
 		return nil, transformSyntacticError(err)
 	}
+	dec.hadPeeked = false
 	switch k := tok.Kind(); k {
 	case 'n':
 		return nil, nil
@@ -230,6 +236,7 @@ func (dec *Decoder) Token() (Token, error) {
 // More reports whether there is another element in the
 // current array or object being parsed.
 func (dec *Decoder) More() bool {
+	dec.hadPeeked = true
 	k := dec.dec.PeekKind()
 	return k > 0 && k != ']' && k != '}'
 }
@@ -238,5 +245,18 @@ func (dec *Decoder) More() bool {
 // The offset gives the location of the end of the most recently returned token
 // and the beginning of the next token.
 func (dec *Decoder) InputOffset() int64 {
-	return dec.dec.InputOffset()
+	offset := dec.dec.InputOffset()
+	if dec.hadPeeked {
+		// Historically, InputOffset reported the location of
+		// the end of the most recently returned token
+		// unless [Decoder.More] is called, in which case, it reported
+		// the beginning of the next token.
+		unreadBuffer := dec.dec.UnreadBuffer()
+		trailingTokens := bytes.TrimLeft(unreadBuffer, " \n\r\t")
+		if len(trailingTokens) > 0 {
+			leadingWhitespace := len(unreadBuffer) - len(trailingTokens)
+			offset += int64(leadingWhitespace)
+		}
+	}
+	return offset
 }
diff --git a/src/encoding/json/v2_stream_test.go b/src/encoding/json/v2_stream_test.go
index b8951f8205..0885631fb5 100644
--- a/src/encoding/json/v2_stream_test.go
+++ b/src/encoding/json/v2_stream_test.go
@@ -537,3 +537,67 @@ func TestTokenTruncation(t *testing.T) {
 		}
 	}
 }
+
+func TestDecoderInputOffset(t *testing.T) {
+	const input = ` [
+		[ ] , [ "one" ] , [ "one" , "two" ] ,
+		{ } , { "alpha" : "bravo" } , { "alpha" : "bravo" , "fizz" : "buzz" }
+	] `
+	wantOffsets := []int64{
+		0, 1, 2, 5, 6, 7, 8, 9, 12, 13, 18, 19, 20, 21, 24, 25, 30, 31,
+		38, 39, 40, 41, 46, 47, 48, 49, 52, 53, 60, 61, 70, 71, 72, 73,
+		76, 77, 84, 85, 94, 95, 103, 104, 112, 113, 114, 116, 117, 117,
+		117, 117,
+	}
+	wantMores := []bool{
+		true, true, false, true, true, false, true, true, true, false,
+		true, false, true, true, true, false, true, true, true, true,
+		true, false, false, false, false,
+	}
+
+	d := NewDecoder(strings.NewReader(input))
+	checkOffset := func() {
+		t.Helper()
+		got := d.InputOffset()
+		if len(wantOffsets) == 0 {
+			t.Fatalf("InputOffset = %d, want nil", got)
+		}
+		want := wantOffsets[0]
+		if got != want {
+			t.Fatalf("InputOffset = %d, want %d", got, want)
+		}
+		wantOffsets = wantOffsets[1:]
+	}
+	checkMore := func() {
+		t.Helper()
+		got := d.More()
+		if len(wantMores) == 0 {
+			t.Fatalf("More = %v, want nil", got)
+		}
+		want := wantMores[0]
+		if got != want {
+			t.Fatalf("More = %v, want %v", got, want)
+		}
+		wantMores = wantMores[1:]
+	}
+	checkOffset()
+	checkMore()
+	checkOffset()
+	for {
+		if _, err := d.Token(); err == io.EOF {
+			break
+		} else if err != nil {
+			t.Fatalf("Token error: %v", err)
+		}
+		checkOffset()
+		checkMore()
+		checkOffset()
+	}
+	checkOffset()
+	checkMore()
+	checkOffset()
+
+	if len(wantOffsets)+len(wantMores) > 0 {
+		t.Fatal("unconsumed testdata")
+	}
+}