Pad/internal/editor/chunked_buffer_test.go

package editor

import (
	"bytes"
	"fmt"
	"strings"
	"testing"

	"pad/internal/io/pool/types"
)

// newTestBuffer builds a fully-loaded (in-range) buffer from content using the
// on-open path (SetContent), so all chunks are resident.
func newTestBuffer(t *testing.T, content []byte) *ChunkedBuffer {
	t.Helper()
	cb := NewChunkedBuffer("/test.txt", DefaultChunkSize, nil, "")
	cb.SetContent(content)
	return cb
}

// TestSetContent_MultipleChunks verifies the whole-file load splits content
// into ordered chunks and Content reconstructs it exactly.
func TestSetContent_MultipleChunks(t *testing.T) {
	content := make([]byte, 200*1024)
	for i := range content {
		content[i] = byte(i % 251)
	}
	cb := newTestBuffer(t, content)

	// 200KB / 64KB = 3 full chunks (64KB each) + 1 partial (8KB) = 4 chunks.
	if got := len(cb.chunks); got != 4 {
		t.Fatalf("expected 4 chunks, got %d", got)
	}
	if cb.FileLen() != int64(len(content)) {
		t.Fatalf("FileLen=%d, want %d", cb.FileLen(), len(content))
	}
	full, err := cb.FullContent()
	if err != nil {
		t.Fatalf("FullContent error: %v", err)
	}
	if full != string(content) {
		t.Fatalf("FullContent mismatch after SetContent")
	}
	// Spot-check a range that spans the chunk-0/chunk-1 boundary.
	seg := cb.Content(64*1024-10, 64*1024+10)
	if seg != string(content[64*1024-10:64*1024+10]) {
		t.Fatalf("Content across chunk boundary mismatch: got %q", seg)
	}
}

// TestInsertShiftsLaterChunks is the core drift-fix test: an insert in chunk 0
// must shift every later byte, and Content must still return the correct byte
// at the shifted offset (the old fixed-slot i*chunkSize model failed here).
func TestInsertShiftsLaterChunks(t *testing.T) {
	content := make([]byte, 200*1024)
	for i := range content {
		content[i] = byte(i % 251)
	}
	cb := newTestBuffer(t, content)

	const insLen = 3
	cb.Insert(0, strings.Repeat("X", insLen))

	// A byte originally at position 100000 (chunk 1) is now at 100003.
	want := content[100000]
	if got := cb.Content(100000+insLen, 100001+insLen); got != string(want) {
		t.Fatalf("byte at shifted offset 100003 = %q, want %q", got, string(want))
	}
	// The inserted prefix is present.
	if got := cb.Content(0, insLen); got != strings.Repeat("X", insLen) {
		t.Fatalf("prefix = %q, want %q", got, strings.Repeat("X", insLen))
	}
	// Full content is the insert + original.
	full, _ := cb.FullContent()
	if full != strings.Repeat("X", insLen)+string(content) {
		t.Fatalf("FullContent after insert mismatch (len=%d, want %d)", len(full), insLen+len(content))
	}
}

// TestDeleteShiftsLaterChunks verifies a deletion in an early chunk shifts
// later bytes left and Content/FullContent stay exact.
func TestDeleteShiftsLaterChunks(t *testing.T) {
	content := make([]byte, 200*1024)
	for i := range content {
		content[i] = byte(i % 251)
	}
	cb := newTestBuffer(t, content)

	const delLen = 100
	cb.Delete(0, delLen)

	// A byte originally at 100000 is now at 99900.
	want := content[100000]
	if got := cb.Content(100000-delLen, 100001-delLen); got != string(want) {
		t.Fatalf("byte at shifted offset = %q, want %q", got, string(want))
	}
	full, _ := cb.FullContent()
	if full != string(content[delLen:]) {
		t.Fatalf("FullContent after delete mismatch (len=%d, want %d)", len(full), len(content)-delLen)
	}
}

// TestDeleteSpanningChunks verifies a deletion that spans multiple chunk
// boundaries is exact.
func TestDeleteSpanningChunks(t *testing.T) {
	content := make([]byte, 200*1024)
	for i := range content {
		content[i] = byte(i % 251)
	}
	cb := newTestBuffer(t, content)

	// Delete 64KB+50 bytes starting at 30KB (spans chunks 0,1,2).
	const start = 30 * 1024
	const n = 64*1024 + 50
	cb.Delete(start, n)

	expected := make([]byte, 0, len(content)-n)
	expected = append(expected, content[:start]...)
	expected = append(expected, content[start+n:]...)
	full, _ := cb.FullContent()
	if !bytes.Equal([]byte(full), expected) {
		t.Fatalf("FullContent after spanning delete mismatch (len=%d, want %d)", len(full), len(expected))
	}
	if cb.FileLen() != int64(len(expected)) {
		t.Fatalf("FileLen=%d, want %d", cb.FileLen(), len(expected))
	}
}

// TestRuneIndexToByteAfterEdit verifies the IME rune->byte bridge stays correct
// after a length-changing edit (ASCII + multibyte).
func TestRuneIndexToByteAfterEdit(t *testing.T) {
	// 2 chunks of ASCII so the edit crosses into chunk arithmetic.
	var b bytes.Buffer
	for i := 0; i < 100*1024; i++ {
		b.WriteByte(byte('a' + i%26))
	}
	cb := newTestBuffer(t, b.Bytes())

	// Insert a multibyte rune ("é" = 2 bytes) at rune 0.
	cb.Insert(0, "é")
	// Rune 0 is now 'é' (bytes 0..1). Rune 1 is the original first 'a' at byte 2.
	if got := cb.RuneIndexToByte(0); got != 0 {
		t.Fatalf("RuneIndexToByte(0)=%d, want 0", got)
	}
	if got := cb.RuneIndexToByte(1); got != 2 {
		t.Fatalf("RuneIndexToByte(1)=%d, want 2", got)
	}
	// New rune R (R>0) is the original rune R-1 (originally at byte R-1), now
	// at byte (R-1)+2 = R+1 (é is 1 rune, 2 bytes). Check one deep in chunk 1
	// (past the 64KB boundary) to confirm the shift survives chunk boundaries.
	deep := 70000
	if got := cb.RuneIndexToByte(deep); got != deep+1 {
		t.Fatalf("RuneIndexToByte(%d)=%d, want %d", deep, got, deep+1)
	}
}

// TestContent_BoundsAndEmpty guards edge cases in the rewritten Content.
func TestContent_BoundsAndEmpty(t *testing.T) {
	cb := newTestBuffer(t, []byte("hello world"))
	if got := cb.Content(0, 5); got != "hello" {
		t.Fatalf("Content(0,5)=%q, want hello", got)
	}
	if got := cb.Content(6, 11); got != "world" {
		t.Fatalf("Content(6,11)=%q, want world", got)
	}
	// Out-of-range end clamps to fileLen.
	if got := cb.Content(0, 1000); got != "hello world" {
		t.Fatalf("Content(0,1000)=%q, want full", got)
	}
	// start >= end is empty.
	if got := cb.Content(5, 5); got != "" {
		t.Fatalf("Content(5,5)=%q, want empty", got)
	}
	// Empty buffer.
	empty := newTestBuffer(t, nil)
	if got := empty.Content(0, 10); got != "" {
		t.Fatalf("empty Content=%q, want empty", got)
	}
	if got, _ := empty.FullContent(); got != "" {
		t.Fatalf("empty FullContent=%q, want empty", got)
	}
}

// TestAppendAtEnd inserts at the very end of the buffer (pos == fileLen), which
// maps to the last chunk's end.
func TestAppendAtEnd(t *testing.T) {
	cb := newTestBuffer(t, []byte("abcdef"))
	cb.Insert(6, "XYZ")
	full, _ := cb.FullContent()
	if full != "abcdefXYZ" {
		t.Fatalf("append at end = %q, want abcdefXYZ", full)
	}
	if cb.FileLen() != 9 {
		t.Fatalf("FileLen=%d, want 9", cb.FileLen())
	}
}

// assertMaxChunkSize verifies the chunking invariant: no chunk may exceed
// twice the target size (Insert splits oversized chunks in half).
func assertMaxChunkSize(t *testing.T, cb *ChunkedBuffer) {
	t.Helper()
	for i, c := range cb.chunks {
		if len(c) > 2*cb.chunkSize {
			t.Fatalf("chunk %d is %d bytes, want <= %d (2 x %d)", i, len(c), 2*cb.chunkSize, cb.chunkSize)
		}
	}
}

// TestInsert_SustainedTyping_SplitsOversizedChunks simulates sustained typing
// at one spot (many small inserts into the same chunk). Without the split, the
// target chunk grows without bound and every edit copies the whole grown
// chunk (quadratic total). With the split, chunks stay O(chunkSize).
func TestInsert_SustainedTyping_SplitsOversizedChunks(t *testing.T) {
	const chunkSize = 64
	cb := NewChunkedBuffer("/test.txt", chunkSize, nil, "")
	seed := strings.Repeat("seed line\n", chunkSize/8) // 64 bytes = 1 chunk
	cb.SetContent([]byte(seed))

	// Type 2000 characters, each inserted immediately after the previous one,
	// starting at position 10 (inside chunk 0).
	var typed strings.Builder
	pos := 10
	for i := 0; i < 2000; i++ {
		r := rune('a' + i%26)
		cb.Insert(pos, string(r))
		pos++
		typed.WriteRune(r)
		if i%200 == 0 {
			assertMaxChunkSize(t, cb)
		}
	}
	assertMaxChunkSize(t, cb)
	want := seed[:10] + typed.String() + seed[10:]
	full, _ := cb.FullContent()
	if full != want {
		t.Fatalf("FullContent mismatch after sustained inserts (len %d vs %d)", len(full), len(want))
	}
	if cb.FileLen() != int64(len(want)) {
		t.Fatalf("FileLen=%d, want %d", cb.FileLen(), len(want))
	}
}

// TestInsert_LargePaste_EmptyBuffer verifies that a large first paste into an
// empty buffer is chunked (not one oversized chunk) and reconstructs exactly.
func TestInsert_LargePaste_EmptyBuffer(t *testing.T) {
	cb := NewChunkedBuffer("/test.txt", 64*1024, nil, "")
	paste := strings.Repeat("0123456789", 100*1024) // 1 MB
	cb.Insert(0, paste)
	assertMaxChunkSize(t, cb)
	if got := len(cb.chunks); got != 16 {
		t.Fatalf("expected 16 chunks for 1MB/64KB, got %d", got)
	}
	full, _ := cb.FullContent()
	if full != paste {
		t.Fatalf("FullContent mismatch after large paste")
	}
}

// TestInsert_SplitContentIntegrity checks that Content stays exact across the
// NEW chunk boundary created by a split, including spans that cross it.
func TestInsert_SplitContentIntegrity(t *testing.T) {
	const chunkSize = 64
	cb := NewChunkedBuffer("/test.txt", chunkSize, nil, "")
	cb.SetContent([]byte(strings.Repeat("x", 64)))
	// 200 inserts of 2 chars at pos 0: the first chunk grows to 464+ bytes and
	// must be split several times.
	for i := 0; i < 200; i++ {
		cb.Insert(0, "AB")
	}
	want := strings.Repeat("AB", 200) + strings.Repeat("x", 64)
	full, _ := cb.FullContent()
	if full != want {
		t.Fatalf("FullContent mismatch: got %d bytes, want %d", len(full), len(want))
	}
	// Span across several chunk boundaries near the front.
	seg := cb.Content(0, 200)
	if seg != want[:200] {
		t.Fatalf("Content(0,200) mismatch across split boundary")
	}
	seg = cb.Content(50, 300)
	if seg != want[50:300] {
		t.Fatalf("Content(50,300) mismatch across split boundary")
	}
}

// int32SlicesEqual reports whether two []int32 are element-wise equal.
func int32SlicesEqual(a, b []int32) bool {
	if len(a) != len(b) {
		return false
	}
	for i := range a {
		if a[i] != b[i] {
			return false
		}
	}
	return true
}

// lineStartOffsets computes the ground-truth line-start byte offsets for
// content: position 0 plus the byte after each '\n'.
func lineStartOffsets(content []byte) []int32 {
	off := make([]int32, 0, 16)
	off = append(off, 0)
	for i, b := range content {
		if b == '\n' {
			off = append(off, int32(i+1))
		}
	}
	return off
}

// assertLineIndexMatchesOracle verifies the buffer's incrementally-maintained
// LineIndex is exactly the one a full recomputation from the content would
// produce. This is the regression guard for newline handling in
// UpdateLineIndexAfterInsert / UpdateLineIndexAfterDelete.
func assertLineIndexMatchesOracle(t *testing.T, cb *ChunkedBuffer, label string) {
	t.Helper()
	full, err := cb.FullContent()
	if err != nil {
		t.Fatalf("%s: FullContent: %v", label, err)
	}
	want := lineStartOffsets([]byte(full))
	if cb.LineIndex == nil {
		t.Fatalf("%s: LineIndex is nil", label)
	}
	if !int32SlicesEqual(cb.LineIndex.Offsets, want) {
		t.Fatalf("%s: line index drift\n got  %v\n want %v", label, cb.LineIndex.Offsets, want)
	}
	if cb.LineIndex.Size != int64(len(full)) {
		t.Fatalf("%s: LineIndex.Size=%d, want %d", label, cb.LineIndex.Size, len(full))
	}
}

// TestLineIndex_InsertNewlines drives inserts that add newlines through the
// incremental update and checks the oracle invariant after each step.
func TestLineIndex_InsertNewlines(t *testing.T) {
	base := "aaa\nbbb\nccc"
	cb := newTestBuffer(t, []byte(base))
	cb.LineIndex = types.NewLineIndex(lineStartOffsets([]byte(base)), 0, int64(len(base)))

	steps := []struct {
		pos  int
		text string
	}{
		{7, "\n"},                 // empty line after "bbb"
		{0, "\n"},                 // leading empty line
		{len(base) + 2, "x\ny\n"}, // append two lines at EOF (after edits)
		{4, "QQ\n"},               // insert at start of "bbb" line, with newline
		{1000000, "zz"},           // insert far past EOF (clamps in Insert)
	}
	for i, s := range steps {
		// Insert clamps pos to EOF; use the same clamped pos for both the
		// buffer edit and the index update, as production code does (the
		// cursor is always within the buffer).
		p := s.pos
		if p > int(cb.FileLen()) {
			p = int(cb.FileLen())
		}
		cb.Insert(p, s.text)
		cb.UpdateLineIndexAfterInsert(p, s.text)
		assertLineIndexMatchesOracle(t, cb, fmt.Sprintf("insert step %d (pos=%d text=%q)", i, p, s.text))
	}
}

// TestLineIndex_DeleteNewlines drives deletions that remove newlines / whole
// lines / from line starts through the incremental update and checks the
// oracle invariant after each step.
func TestLineIndex_DeleteNewlines(t *testing.T) {
	base := "aaa\nbbb\nccc\nddd"
	cb := newTestBuffer(t, []byte(base))
	cb.LineIndex = types.NewLineIndex(lineStartOffsets([]byte(base)), 0, int64(len(base)))

	// (start, end) ranges chosen to exercise: mid-line, across a newline,
	// a whole line, from a line start, and the file start.
	steps := []struct {
		start, end int
	}{
		{3, 4}, // delete the '\n' after aaa -> merge aaa+bbb
		{4, 7}, // delete "bbb" (after merge: content is aaabbb\nddd...)
		{0, 3}, // delete from file start
		{3, 8}, // delete "bb\n" region
	}
	for i, s := range steps {
		full, _ := cb.FullContent()
		if s.start > len(full) || s.end > len(full) {
			t.Fatalf("step %d: range [%d,%d) beyond content len %d", i, s.start, s.end, len(full))
		}
		cb.Delete(s.start, s.end-s.start)
		cb.UpdateLineIndexAfterDelete(s.start, s.end)
		assertLineIndexMatchesOracle(t, cb, fmt.Sprintf("delete step %d [%d,%d)", i, s.start, s.end))
	}
}

// TestLineIndex_MixedEditSequence interleaves inserts and deletes (including
// IME-style replace = delete+insert at the same point) and checks the oracle
// after each operation.
func TestLineIndex_MixedEditSequence(t *testing.T) {
	base := "hello\nworld\nfoo\nbar\nbaz"
	cb := newTestBuffer(t, []byte(base))
	cb.LineIndex = types.NewLineIndex(lineStartOffsets([]byte(base)), 0, int64(len(base)))

	ops := []struct {
		desc  string
		isIns bool
		pos   int
		text  string // for insert
		start int    // for delete
		end   int    // for delete
	}{
		{desc: "insert newline mid", isIns: true, pos: 3, text: "\n"},
		{desc: "delete across nl", start: 6, end: 10},
		{desc: "replace with multi-line", isIns: false, pos: 2, text: "X\nY\nZ", start: 2, end: 4},
		{desc: "delete whole line", start: 0, end: 4},
		{desc: "insert at eof", isIns: true, pos: 999, text: "end\n"},
	}
	for i, op := range ops {
		full, _ := cb.FullContent()
		switch {
		case op.isIns:
			p := op.pos
			if p > len(full) {
				p = len(full)
			}
			cb.Insert(p, op.text)
			cb.UpdateLineIndexAfterInsert(p, op.text)
		case op.text != "":
			// replace: delete [start,end) then insert at start, mirroring
			// HandleReplaceRange ordering.
			s, e := op.start, op.end
			if s > len(full) {
				s = len(full)
			}
			if e > len(full) {
				e = len(full)
			}
			if e > s {
				cb.Delete(s, e-s)
				cb.UpdateLineIndexAfterDelete(s, e)
			}
			cb.Insert(s, op.text)
			cb.UpdateLineIndexAfterInsert(s, op.text)
		default:
			s, e := op.start, op.end
			if s > len(full) {
				s = len(full)
			}
			if e > len(full) {
				e = len(full)
			}
			if e > s {
				cb.Delete(s, e-s)
				cb.UpdateLineIndexAfterDelete(s, e)
			}
		}
		assertLineIndexMatchesOracle(t, cb, fmt.Sprintf("op %d (%s)", i, op.desc))
	}
}