get scraped documents list performance update

internal/tender/repository.go

@@ -21,6 +21,8 @@ type TenderRepository interface {
 	GetByContractNoticeID(ctx context.Context, contractNoticeID string) (*Tender, error)
 	GetByNoticePublicationID(ctx context.Context, noticePublicationID string) (*Tender, error)
 	GetByContractFolderID(ctx context.Context, contractFolderID string) (*Tender, error)
+	// GetLatestByContractFolderIDs returns the most recently updated tender per distinct contract_folder_id.
+	GetLatestByContractFolderIDs(ctx context.Context, contractFolderIDs []string) (map[string]*Tender, error)
 	GetByProcurementProjectID(ctx context.Context, procurementProjectID string) (*Tender, error)
 	FindTendersWithContentXML(ctx context.Context, limit, skip int) ([]Tender, error)
 	GetTenderCountByCountry(ctx context.Context) (map[string]int64, error)
@@ -289,6 +291,76 @@ func (r *tenderRepository) GetByContractFolderID(ctx context.Context, contractFo
 	return &result.Items[0], nil
 }
 
+// GetLatestByContractFolderIDs returns one tender per distinct contract_folder_id (latest updated_at).
+func (r *tenderRepository) GetLatestByContractFolderIDs(ctx context.Context, contractFolderIDs []string) (map[string]*Tender, error) {
+	out := make(map[string]*Tender)
+	if len(contractFolderIDs) == 0 {
+		return out, nil
+	}
+
+	uniq := uniqueNonEmptyTrimmedStrings(contractFolderIDs)
+	if len(uniq) == 0 {
+		return out, nil
+	}
+
+	// Allow a few rows per folder in case duplicate folder rows exist despite partial uniques.
+	findLimit := len(uniq) * 3
+	if findLimit < len(uniq)+1 {
+		findLimit = len(uniq) + 1
+	}
+	if findLimit > 100000 {
+		findLimit = 100000
+	}
+
+	result, err := r.ormRepo.FindAll(ctx,
+		bson.M{"contract_folder_id": bson.M{"$in": uniq}},
+		orm.Pagination{
+			Limit:     findLimit,
+			Skip:      0,
+			SortField: "updated_at",
+			SortOrder: -1,
+		},
+	)
+	if err != nil {
+		r.logger.Error("Failed to load tenders by contract folder ids", map[string]interface{}{
+			"folder_count": len(uniq),
+			"error":        err.Error(),
+		})
+		return nil, err
+	}
+
+	for i := range result.Items {
+		t := &result.Items[i]
+		cid := strings.TrimSpace(t.ContractFolderID)
+		if cid == "" {
+			continue
+		}
+		prev, ok := out[cid]
+		if !ok || t.UpdatedAt > prev.UpdatedAt {
+			out[cid] = t
+		}
+	}
+
+	return out, nil
+}
+
+func uniqueNonEmptyTrimmedStrings(ids []string) []string {
+	seen := make(map[string]struct{}, len(ids))
+	out := make([]string, 0, len(ids))
+	for _, id := range ids {
+		id = strings.TrimSpace(id)
+		if id == "" {
+			continue
+		}
+		if _, dup := seen[id]; dup {
+			continue
+		}
+		seen[id] = struct{}{}
+		out = append(out, id)
+	}
+	return out
+}
+
 // GetByProcurementProjectID finds a tender by UBL ProcurementProject ID (shared across per-lot notices).
 func (r *tenderRepository) GetByProcurementProjectID(ctx context.Context, procurementProjectID string) (*Tender, error) {
 	filter := bson.M{"procurement_project_id": procurementProjectID}