Add CSV export of scan results

The Review tab gets an Export CSV menu: "This view" exports the current
status filter and search, "Entire scan" exports every scanned slave file.

Each row carries status, match type, a recommendation, slave/master paths,
the link target, the match reason, sizes, codec details for both sides,
the quality verdict, notes, errors and timestamps. Rows are streamed in
pages of 1000 so a large export never holds the database connection a
running scan needs. The file starts with a UTF-8 BOM for Excel, and cells
a spreadsheet would evaluate as formulas are prefixed with an apostrophe.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

internal/api/export_test.go

@@ -0,0 +1,162 @@
+package api
+
+import (
+	"bytes"
+	"context"
+	"encoding/csv"
+	"fmt"
+	"io"
+	"log/slog"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+	"testing/fstest"
+	"time"
+
+	"github.com/cdnutter/playlink/internal/engine"
+	"github.com/cdnutter/playlink/internal/probe"
+	"github.com/cdnutter/playlink/internal/store"
+)
+
+func TestExportCSV(t *testing.T) {
+	ctx := context.Background()
+	root, _ := filepath.EvalSymlinks(t.TempDir())
+	master, slave := filepath.Join(root, "Albums"), filepath.Join(root, "Singles")
+	write := func(base, rel, content string) {
+		p := filepath.Join(base, rel)
+		os.MkdirAll(filepath.Dir(p), 0o755)
+		if err := os.WriteFile(p, []byte(content), 0o644); err != nil {
+			t.Fatal(err)
+		}
+	}
+	// Enough rows to need more than one export page.
+	for i := 0; i < 1005; i++ {
+		rel := fmt.Sprintf("Bulk/Album/%04d.flac", i)
+		write(master, rel, "same")
+		write(slave, rel, "same")
+	}
+	write(master, "Beyoncé/Lemonade/01 Pray You Catch Me.flac", "a")
+	write(slave, "Beyoncé/Lemonade/01 Pray You Catch Me.flac", "b")
+	write(slave, "=HYPERLINK(evil)/song.flac", "x") // a folder name a spreadsheet would run
+
+	st, err := store.Open(filepath.Join(root))
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer st.DB.Close()
+	eng := engine.New(st, probe.Prober{}, []string{root}, slog.New(slog.NewTextHandler(io.Discard, nil)))
+	p := store.Pair{Name: "Chris' Music", MasterRoot: master, SlaveRoot: slave}
+	if _, err := eng.NormalizePair(&p); err != nil {
+		t.Fatal(err)
+	}
+	if err := st.SavePair(ctx, &p); err != nil {
+		t.Fatal(err)
+	}
+	if _, err := eng.StartScan(p.ID, "test"); err != nil {
+		t.Fatal(err)
+	}
+	for deadline := time.Now().Add(20 * time.Second); ; {
+		if j := eng.Job(p.ID); j != nil && j.State != "running" {
+			break
+		}
+		if time.Now().After(deadline) {
+			t.Fatal("scan did not finish")
+		}
+		time.Sleep(20 * time.Millisecond)
+	}
+
+	h := New(Config{}, st, eng, slog.New(slog.NewTextHandler(io.Discard, nil)), fstest.MapFS{}).Handler()
+	get := func(url string, auth bool) *httptest.ResponseRecorder {
+		req := httptest.NewRequest(http.MethodGet, url, nil)
+		if auth {
+			req.AddCookie(&http.Cookie{Name: cookieName, Value: "tok"})
+		}
+		rec := httptest.NewRecorder()
+		h.ServeHTTP(rec, req)
+		return rec
+	}
+	url := fmt.Sprintf("/api/pairs/%d/export.csv", p.ID)
+	if rec := get(url, false); rec.Code != http.StatusUnauthorized {
+		t.Fatalf("export without a session must be refused, got %d", rec.Code)
+	}
+	if err := st.CreateFirstUser(ctx, "chris", "x"); err != nil {
+		t.Fatal(err)
+	}
+	u, _ := st.UserByName(ctx, "chris")
+	if err := st.CreateSession(ctx, hashToken("tok"), u.ID, time.Now().Add(time.Hour).Unix()); err != nil {
+		t.Fatal(err)
+	}
+
+	rec := get(url, true)
+	if rec.Code != 200 {
+		t.Fatalf("status %d: %s", rec.Code, rec.Body.String())
+	}
+	if ct := rec.Header().Get("Content-Type"); !strings.HasPrefix(ct, "text/csv") {
+		t.Fatalf("content type %q", ct)
+	}
+	if cd := rec.Header().Get("Content-Disposition"); !strings.Contains(cd, `filename="playlink-Chris-Music-all-`) {
+		t.Fatalf("content disposition %q", cd)
+	}
+	body := rec.Body.Bytes()
+	if !bytes.HasPrefix(body, []byte{0xEF, 0xBB, 0xBF}) {
+		t.Fatal("missing UTF-8 byte order mark")
+	}
+	rows, err := csv.NewReader(bytes.NewReader(body[3:])).ReadAll()
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(rows) != 1+1005+2 {
+		t.Fatalf("want header + 1007 rows (paging), got %d", len(rows))
+	}
+	if strings.Join(rows[0], ",") != strings.Join(exportHeader, ",") {
+		t.Fatal("unexpected header")
+	}
+	col := map[string]int{}
+	for i, h := range rows[0] {
+		col[h] = i
+	}
+	seen := map[string][]string{}
+	for _, r := range rows[1:] {
+		seen[strings.TrimPrefix(r[col["slave_path"]], "'")] = r
+	}
+	if r := seen["Beyoncé/Lemonade/01 Pray You Catch Me.flac"]; r == nil || r[col["status"]] != "different" || r[col["recommendation"]] != "review" {
+		t.Fatalf("non-ASCII row missing or wrong: %v", r)
+	}
+	if r := seen["=HYPERLINK(evil)/song.flac"]; r == nil || r[col["slave_path"]] != "'=HYPERLINK(evil)/song.flac" {
+		t.Fatalf("a path starting with = must be neutralised, got %v", r)
+	}
+	if r := seen["Bulk/Album/0000.flac"]; r[col["recommendation"]] != "link (identical)" || r[col["match_type"]] != "same path" {
+		t.Fatalf("identical row: %v", r)
+	}
+
+	// Filters narrow the export exactly like the Review list.
+	rec = get(url+"?status=different", true)
+	rows, _ = csv.NewReader(bytes.NewReader(rec.Body.Bytes()[3:])).ReadAll()
+	if len(rows) != 2 {
+		t.Fatalf("status filter: want header + 1 row, got %d", len(rows))
+	}
+	rec = get(url+"?q=0100", true)
+	rows, _ = csv.NewReader(bytes.NewReader(rec.Body.Bytes()[3:])).ReadAll()
+	if len(rows) != 2 {
+		t.Fatalf("search filter: want header + 1 row, got %d", len(rows))
+	}
+}
+
+func TestCSVSafe(t *testing.T) {
+	for in, want := range map[string]string{
+		"=SUM(A1)":     "'=SUM(A1)",
+		"+1":           "'+1",
+		"-x":           "'-x",
+		"@cmd":         "'@cmd",
+		"Artist/Song":  "Artist/Song",
+		"":             "",
+		"Beyoncé/Halo": "Beyoncé/Halo",
+	} {
+		if got := csvSafe(in); got != want {
+			t.Errorf("csvSafe(%q) = %q, want %q", in, got, want)
+		}
+	}
+}