feat(wing): BM25 scorer as the independent cross-check (#205 S1)

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

ingestion/internal/wing/bm25scorer_test.go

@@ -0,0 +1,45 @@
+package wing_test
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"github.com/mathiasbq/hyperguild/ingestion/internal/wing"
+)
+
+func writePage(t *testing.T, dir, rel, body string) {
+	t.Helper()
+	full := filepath.Join(dir, rel)
+	require.NoError(t, os.MkdirAll(filepath.Dir(full), 0o755))
+	require.NoError(t, os.WriteFile(full, []byte(body), 0o644))
+}
+
+func scorerBrain(t *testing.T) string {
+	dir := t.TempDir()
+	writePage(t, dir, "wiki/infra/facts/k3s.md", "---\ntitle: K3s cluster\n---\nkubernetes cluster nodes scheduling pods deployment\n")
+	writePage(t, dir, "wiki/finance/facts/invoice.md", "---\ntitle: Invoices\n---\ninvoice payment bookkeeping ledger reconciliation\n")
+	return dir
+}
+
+func TestBM25Scorer_RanksTheWingWhosePagesMatch(t *testing.T) {
+	s := wing.NewBM25Scorer(scorerBrain(t), []string{"infra", "finance"}, "")
+	got := s.Score("notes about kubernetes cluster scheduling and pods")
+	assert.Greater(t, got["infra"], got["finance"])
+	assert.Greater(t, got["infra"], 0.0)
+}
+
+func TestBM25Scorer_ExcludesThePageBeingClassified(t *testing.T) {
+	// A page must not vouch for its own wing when it is the backtest sample.
+	s := wing.NewBM25Scorer(scorerBrain(t), []string{"infra", "finance"}, "wiki/infra/facts/k3s.md")
+	got := s.Score("kubernetes cluster nodes scheduling pods deployment")
+	assert.Zero(t, got["infra"], "the excluded page is the only infra match")
+}
+
+func TestBM25Scorer_NoUsableTermsScoresNothing(t *testing.T) {
+	s := wing.NewBM25Scorer(scorerBrain(t), []string{"infra", "finance"}, "")
+	assert.Empty(t, s.Score("a an to of"))
+}