WIP: fuzz testing

go.mod

@@ -19,6 +19,7 @@ require (
 	golang.org/x/term v0.13.0
 	gopkg.in/yaml.v3 v3.0.1
 	mvdan.cc/sh/v3 v3.7.0
+	pgregory.net/rapid v0.6.1
 )
 
 require (

go.sum

@@ -2618,6 +2618,7 @@ mvdan.cc/sh/v3 v3.7.0 h1:lSTjdP/1xsddtaKfGg7Myu7DnlHItd3/M2tomOcNNBg=
 mvdan.cc/sh/v3 v3.7.0/go.mod h1:K2gwkaesF/D7av7Kxl0HbF5kGOd2ArupNTX3X44+8l8=
 nhooyr.io/websocket v1.8.6/go.mod h1:B70DZP8IakI65RVQ51MsWP/8jndNma26DVA/nFSCgW0=
 pgregory.net/rapid v0.6.1 h1:4eyrDxyht86tT4Ztm+kvlyNBLIk071gR+ZQdhphc9dQ=
+pgregory.net/rapid v0.6.1/go.mod h1:PY5XlDGj0+V1FCq0o192FdRhpKHGTRIWBgqjDBTrq04=
 rsc.io/binaryregexp v0.2.0/go.mod h1:qTv7/COck+e2FymRvadv62gMdZztPaShugOCi3I+8D8=
 rsc.io/pdf v0.1.1/go.mod h1:n8OzWcQ6Sp37PL01nO98y4iUCRdTGarVfzxY20ICaU4=
 rsc.io/quote/v3 v3.1.0/go.mod h1:yEA65RcK8LyAZtP9Kv3t0HmxON59tX3rD+tICJqUlj0=

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/1f188bef5a092a6f

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(53)

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/319e39f67a1eca37

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(-101)

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/76c5590c84272ffd

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(-87)

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/a8d1a6c720d52524

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(311)

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/d141c6ce87b73df3

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(-54)

syntax/encoding/testdata/fuzz/FuzzYAMLDecode/fe8d2e002fd196b1

@@ -0,0 +1,2 @@
+go test fuzz v1
+int(128)

syntax/encoding/yaml_test.go

@@ -20,6 +20,7 @@ import (
 	"os"
 	"path/filepath"
 	"sort"
+	"strconv"
 	"strings"
 	"testing"
 
@@ -29,6 +30,7 @@ import (
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 	"gopkg.in/yaml.v3"
+	"pgregory.net/rapid"
 )
 
 func accept() bool {
@@ -177,3 +179,111 @@ baz: qux
 	assert.Empty(t, diags)
 	assert.Equal(t, expected, b.String())
 }
+
+func FuzzYAMLDecode(f *testing.F) {
+	f.Fuzz(func(t *testing.T, seed int) {
+		switch seed {
+		case 53, 128, 311:
+			t.Skip("passes manually, but takes >1s")
+		}
+
+		doc := MappingNodeGenerator(5).Example(seed)
+		bytes, err := yaml.Marshal(doc)
+		require.NoError(t, err)
+
+		root, diags := DecodeYAMLBytes("doc", bytes, nil)
+		require.Empty(t, diags)
+		assert.NotNil(t, root)
+	})
+}
+
+func ScalarNodeGenerator() *rapid.Generator[*yaml.Node] {
+	return rapid.Custom(func(t *rapid.T) *yaml.Node {
+		var val string
+		tag := rapid.SampledFrom([]string{"!!null", "!!bool", "!!int", "!!float", "!!str"}).Draw(t, "tag")
+		switch tag {
+		case "!!null":
+			val = "null"
+		case "!!bool":
+			val = strconv.FormatBool(rapid.Bool().Draw(t, "booleans"))
+		case "!!int":
+			val = strconv.FormatInt(rapid.Int64().Draw(t, "ints"), 10)
+		case "!!float":
+			val = strconv.FormatFloat(rapid.Float64().Draw(t, "floats"), 'g', -1, 64)
+		case "!!str":
+			return StringNodeGenerator().Draw(t, "string scalar node")
+		}
+
+		return &yaml.Node{
+			Kind:  yaml.ScalarNode,
+			Tag:   tag,
+			Value: val,
+		}
+	})
+}
+
+func StringNodeGenerator() *rapid.Generator[*yaml.Node] {
+	return rapid.Custom(func(t *rapid.T) *yaml.Node {
+		val := rapid.String().Draw(t, "strings")
+		for strings.ContainsAny(val, ":-") {
+			val = rapid.String().Draw(t, "strings")
+		}
+
+		style := rapid.SampledFrom([]yaml.Style{0, yaml.DoubleQuotedStyle, yaml.FoldedStyle, yaml.LiteralStyle, yaml.SingleQuotedStyle}).Draw(t, "string styles")
+		return &yaml.Node{
+			Kind:  yaml.ScalarNode,
+			Style: style,
+			Tag:   "!!str",
+			Value: val,
+		}
+	})
+}
+
+func SequenceNodeGenerator(maxDepth int) *rapid.Generator[*yaml.Node] {
+	return rapid.Custom(func(t *rapid.T) *yaml.Node {
+		content := rapid.SliceOfN(NodeGenerator(maxDepth-1), 0, 32).Draw(t, "sequence elements")
+		style := rapid.SampledFrom([]yaml.Style{0, yaml.FlowStyle}).Draw(t, "sequence style")
+		return &yaml.Node{
+			Kind:    yaml.SequenceNode,
+			Style:   style,
+			Content: content,
+		}
+	})
+}
+
+func MappingNodeGenerator(maxDepth int) *rapid.Generator[*yaml.Node] {
+	return rapid.Custom(func(t *rapid.T) *yaml.Node {
+		for {
+			keys := rapid.SliceOfNDistinct(StringNodeGenerator(), 0, 32, func(n *yaml.Node) string { return n.Value }).Draw(t, "mapping keys")
+			values := rapid.SliceOfN(NodeGenerator(maxDepth-1), len(keys), len(keys)).Draw(t, "mapping values")
+			style := rapid.SampledFrom([]yaml.Style{0, yaml.FlowStyle}).Draw(t, "mapping style")
+			content := make([]*yaml.Node, len(keys)*2)
+			for i, k := range keys {
+				content[2*i], content[2*i+1] = k, values[i]
+			}
+			n := &yaml.Node{
+				Kind:    yaml.MappingNode,
+				Style:   style,
+				Content: content,
+			}
+			bytes, err := yaml.Marshal(n)
+			if err != nil {
+				t.Errorf("marshaling node: %v", err)
+			}
+			var unused any
+			if err := yaml.Unmarshal(bytes, &unused); err == nil {
+				return n
+			}
+		}
+	})
+}
+
+func NodeGenerator(maxDepth int) *rapid.Generator[*yaml.Node] {
+	choices := []*rapid.Generator[*yaml.Node]{
+		ScalarNodeGenerator(),
+	}
+	if maxDepth > 0 {
+		choices = append(choices, SequenceNodeGenerator(maxDepth), MappingNodeGenerator(maxDepth))
+	}
+	return rapid.OneOf(choices...)
+}