feat: repair Redis candidates from durable source

server/store/candidate_projection_test.go

@@ -0,0 +1,48 @@
+package store
+
+import (
+	"context"
+	"testing"
+	"time"
+
+	"github.com/alicebob/miniredis/v2"
+	"github.com/cosmic-clash/cosmic-clash/server/domain"
+	"github.com/redis/go-redis/v9"
+)
+
+func TestCandidateProjectionRepairsPartialRedisStateFromDurableSource(t *testing.T) {
+	mini, err := miniredis.Run()
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer mini.Close()
+	client := redis.NewClient(&redis.Options{Addr: mini.Addr()})
+	defer client.Close()
+	now := time.Unix(1000, 0).UTC()
+	candidate := domain.Candidate{TicketID: "repair-ticket", PlayerID: "repair-player", EnqueuedAt: now}
+	index := RedisCandidateIndex{Client: client, Prefix: "repair", TTL: time.Minute}
+	_, orderKey := index.keys()
+	if err := client.ZAdd(context.Background(), orderKey, redis.Z{Score: float64(now.UnixNano()), Member: candidate.TicketID}).Err(); err != nil {
+		t.Fatal(err)
+	}
+	projection := CandidateProjection{Index: index, Source: func(context.Context, time.Time) ([]domain.Candidate, error) {
+		return []domain.Candidate{candidate}, nil
+	}}
+	got, err := projection.Snapshot(context.Background(), now)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(got) != 1 || got[0].TicketID != candidate.TicketID {
+		t.Fatalf("repaired projection = %+v", got)
+	}
+}
+
+func TestCandidateProjectionDoesNotReturnCacheWhenRepairSourceFails(t *testing.T) {
+	index := RedisCandidateIndex{TTL: time.Minute}
+	projection := CandidateProjection{Index: index, Source: func(context.Context, time.Time) ([]domain.Candidate, error) {
+		return nil, context.DeadlineExceeded
+	}}
+	if _, err := projection.Snapshot(context.Background(), time.Unix(1000, 0)); err == nil {
+		t.Fatal("cache projection succeeded without a usable Redis/index source")
+	}
+}

server/store/redis_candidates.go

@@ -19,6 +19,44 @@ type RedisCandidateIndex struct {
 	TTL    time.Duration
 }
 
+// DurableCandidateSource is the authoritative queue projection used to
+// repair Redis. Implementations must apply queue state and expiry rules before
+// returning candidates.
+type DurableCandidateSource func(context.Context, time.Time) ([]domain.Candidate, error)
+
+// CandidateProjection couples the transient index to its durable repair
+// source. A cache miss, partial write, malformed payload, or Redis restart is
+// repaired before candidates are returned to a matcher.
+type CandidateProjection struct {
+	Index  RedisCandidateIndex
+	Source DurableCandidateSource
+}
+
+func (p CandidateProjection) Repair(ctx context.Context, now time.Time) error {
+	if p.Source == nil || now.IsZero() {
+		return fmt.Errorf("invalid candidate repair source")
+	}
+	candidates, err := p.Source(ctx, now)
+	if err != nil {
+		return err
+	}
+	return p.Index.Rebuild(ctx, candidates)
+}
+
+func (p CandidateProjection) Snapshot(ctx context.Context, now time.Time) ([]domain.Candidate, error) {
+	if p.Source == nil {
+		return nil, fmt.Errorf("invalid candidate repair source")
+	}
+	candidates, err := p.Index.Snapshot(ctx, now)
+	if err == nil {
+		return candidates, nil
+	}
+	if err := p.Repair(ctx, now); err != nil {
+		return nil, err
+	}
+	return p.Index.Snapshot(ctx, now)
+}
+
 func (r RedisCandidateIndex) keys() (string, string) {
 	prefix := r.Prefix
 	if prefix == "" {
@@ -105,17 +143,25 @@ func (r RedisCandidateIndex) Snapshot(ctx context.Context, now time.Time) ([]dom
 		return nil, err
 	}
 	result := make([]domain.Candidate, 0, len(payloads))
-	for _, raw := range payloads {
-		text, ok := raw.(string)
-		if !ok {
-			continue
+	for i, raw := range payloads {
+		var encoded []byte
+		switch value := raw.(type) {
+		case string:
+			encoded = []byte(value)
+		case []byte:
+			encoded = value
+		default:
+			return nil, fmt.Errorf("candidate payload missing for %s", tickets[i])
 		}
 		var candidate domain.Candidate
-		if err := json.Unmarshal([]byte(text), &candidate); err != nil {
-			continue
+		if err := json.Unmarshal(encoded, &candidate); err != nil {
+			return nil, fmt.Errorf("invalid candidate payload for %s: %w", tickets[i], err)
 		}
-		if err := validateRedisCandidate(candidate); err != nil || candidate.EnqueuedAt.After(now) {
-			continue
+		if err := validateRedisCandidate(candidate); err != nil {
+			return nil, fmt.Errorf("invalid candidate payload for %s: %w", tickets[i], err)
+		}
+		if candidate.EnqueuedAt.After(now) {
+			return nil, fmt.Errorf("candidate payload is newer than its index for %s", tickets[i])
 		}
 		result = append(result, candidate)
 	}