interpres/fuzz_test.go

// Copyright (c) 2026 Petr Balvín <opensource@petrbalvin.org> (https://petrbalvin.org)
// SPDX-License-Identifier: MIT

package interpres

import (
	"math"
	"reflect"
	"testing"
	"time"
)

// FuzzParse drives the parser with arbitrary input and holds it to the
// round-trip invariant: every document Parse accepts must survive its own
// re-emission. Marshal of the parsed tree must succeed, the emitted document
// must parse again, and the re-parsed tree must equal the original one.
func FuzzParse(f *testing.F) {
	seeds := []string{
		"",
		"title = \"interpres\"\n",
		"[server]\nhost = \"localhost\"\nport = 8080\n\n[server.tls]\nenabled = true\n",
		"[[items]]\nname = \"a\"\n\n[[items]]\nname = \"b\"\n",
		"inline = { a = 1, b = [2, 3], c = { d = 4 } }\n",
		"arr = [1, 2.5, \"three\", true, 1979-05-27T07:32:00Z]\n",
		"mix = [1, {a = 2}, \"x\"]\n",
		"when = 1979-05-27T07:32:00Z\nlocal = 1979-05-27T07:32:00.999\nd = 1979-05-27\nt = 07:32:00\n",
		"multi = \"\"\"\nlines\n\"\"\"\nlit = 'literal'\n",
		"esc = \"\\u0000\\t\\n\\\"\\\\\"\n",
		"neg = -0.0\nnan = nan\ninf = -inf\nexp = 1e6\n",
		"\"quoted key\" = 'value'\n'1979-05-27' = 1\na.b.c = { d = \"dotted\" }\n",
		"hex = 0xFF\noct = 0o755\nbin = 0b1010\nsep = 1_000_000\n",
		"x = \"unterminated\n",
		"[a]\n[a]\n",
		"n = 0x1_0000_0000_0000_0000\n",
		// TOML 1.1 forms.
		"t = 13:37\ndt = 1979-05-27T07:32\nodt = 1979-05-27 07:32Z\n",
		"esc = \"\\e\\x41\\x7f\\x00\"\n",
		"m = {\n\ta = 1,\n\tb = [1, 2,],\n\tc = { d = 2 },\n} # close\n",
	}
	for _, s := range seeds {
		f.Add([]byte(s))
	}
	f.Fuzz(func(t *testing.T, data []byte) {
		tree, err := ParseMap(data)
		if err != nil {
			return
		}
		out, err := Marshal(tree)
		if err != nil {
			t.Fatalf("marshal of a parsed tree failed: %v\ntree: %#v", err, tree)
		}
		re, err := ParseMap(out)
		if err != nil {
			t.Fatalf("re-parse of the emitted document failed: %v\ndoc:\n%s", err, out)
		}
		if !tomlEqual(tree, re) {
			t.Fatalf("round-trip changed the tree\ninput: %q\ndoc:\n%s\nwas: %#v\nnow: %#v", data, out, tree, re)
		}
	})
}

// tomlEqual reports whether two parsed trees are equal as TOML values. It
// differs from reflect.DeepEqual where DeepEqual is wrong for this domain:
// NaN compares equal to itself, date-times compare by their canonical TOML
// rendering so two parses of one document stay equal, and the local variants
// compare through their String form, which fully determines the value.
func tomlEqual(a, b any) bool {
	switch av := a.(type) {
	case nil:
		return b == nil
	case float64:
		bv, ok := b.(float64)
		return ok && (av == bv || (math.IsNaN(av) && math.IsNaN(bv)))
	case time.Time:
		bv, ok := b.(time.Time)
		return ok && av.Format(time.RFC3339Nano) == bv.Format(time.RFC3339Nano)
	case LocalDateTime:
		bv, ok := b.(LocalDateTime)
		return ok && av.String() == bv.String()
	case LocalDate:
		bv, ok := b.(LocalDate)
		return ok && av.String() == bv.String()
	case LocalTime:
		bv, ok := b.(LocalTime)
		return ok && av.String() == bv.String()
	case []any:
		bv, ok := b.([]any)
		if !ok || len(av) != len(bv) {
			return false
		}
		for i := range av {
			if !tomlEqual(av[i], bv[i]) {
				return false
			}
		}
		return true
	case []map[string]any:
		bv, ok := b.([]map[string]any)
		if !ok || len(av) != len(bv) {
			return false
		}
		for i := range av {
			if !tomlEqual(av[i], bv[i]) {
				return false
			}
		}
		return true
	case map[string]any:
		bv, ok := b.(map[string]any)
		if !ok || len(av) != len(bv) {
			return false
		}
		for k, v := range av {
			other, ok := bv[k]
			if !ok || !tomlEqual(v, other) {
				return false
			}
		}
		return true
	default:
		return reflect.DeepEqual(a, b)
	}
}

// FuzzMarshal drives the encoder with generated Go values and holds it to
// the same round-trip invariant FuzzParse holds the parser to: a value built
// only of encodable kinds must marshal, the document must re-parse, and the
// tree must equal the value it came from.
func FuzzMarshal(f *testing.F) {
	seeds := [][]byte{
		{},
		{0, 0, 1, 2},
		{1, 1, 2, 3, 2, 2, 3, 4},
		{0, 3, 1, 9, 3, 3, 2, 8, 1, 0, 1, 7},
	}
	for _, s := range seeds {
		f.Add(s)
	}
	f.Fuzz(func(t *testing.T, data []byte) {
		v := fuzzValue(data)
		out, err := Marshal(v)
		if err != nil {
			t.Fatalf("marshal of an encodable value failed: %v\nvalue: %#v", err, v)
		}
		tree, err := ParseMap(out)
		if err != nil {
			t.Fatalf("re-parse of the emitted document failed: %v\ndoc:\n%s", err, out)
		}
		if !tomlEqual(v, tree) {
			t.Fatalf("round-trip changed the value\nvalue: %#v\ndoc:\n%s\ntree: %#v", v, out, tree)
		}
	})
}

// fuzzKeys is the fixed key pool the generated values draw from, so keys are
// always valid bare keys and repeat often.
var fuzzKeys = []string{"alpha", "beta", "gamma", "delta"}

// fuzzValue builds a map[string]any of encodable kinds from data: integers,
// positive floats, short strings, nested tables and scalar arrays. The bytes
// decide the shape deterministically.
func fuzzValue(data []byte) map[string]any {
	root := map[string]any{}
	cur := root
	depth := 0
	for i := 0; i+3 < len(data); i += 4 {
		key := fuzzKeys[int(data[i])%len(fuzzKeys)]
		switch data[i+1] % 5 {
		case 0:
			cur[key] = int64(data[i+2])<<8 | int64(data[i+3])
		case 1:
			cur[key] = float64(int(data[i+2])%1000)/8.0 + 0.125
		case 2:
			cur[key] = string(rune('a' + int(data[i+2])%26))
		case 3:
			cur[key] = []any{
				int64(data[i+2]),
				float64(int(data[i+3])%100)/4.0 + 0.25,
				string(rune('a' + int(data[i+3])%26)),
			}
		case 4:
			if depth < 6 {
				next := map[string]any{}
				cur[key] = next
				cur = next
				depth++
			}
		}
	}
	return root
}