feat(tasks): 任务列表支持批量修改分类

勾选多个任务后可一次性移动到目标分类,或选「未分类」把它们移出当前分类。

- db: 新增 SetTasksCategory,在单个事务里校验分类存在性并 UPDATE ... WHERE
  id=ANY(...),返回真正被更新的 id 和刷新后的分类(含 task_count)。分类检查
  放在事务内,避免并发删除把 UPDATE 变成外键冲突。单批上限 100。
- server: 新增 POST /api/tasks/category/batch。写入原子,响应里的逐任务失败
  只可能是勾选后又被删掉的任务;顺带把 category_id 的解析(必填但可为 null)
  抽成 parseCategoryIDField 与单任务接口共用。
- server: 把批量 id 的 trim/归一/去重逻辑从 controlTasksBatch 抽成
  normalizeBatchTaskIDs,两个批量端点共用同一套重复判定。
- manager: SetTasksCategory 在 taskStateMu 内完成 DB 写入与运行中任务 DTO
  刷新,防止与单任务更新交错后留下过期的 category_name。
- web: 选中工具栏新增「修改分类」对话框(需确认目标,避免误点批量改动);
  「未分类」哨兵值统一提取为 UNCATEGORIZED_VALUE。

验证: go build / vet / gofmt、db 与 server 新增测试连本地 PG 实跑通过、
tsc --noEmit、biome check、next build 均通过。

server/manager.go

@@ -812,6 +812,49 @@ func (m *Manager) SetTaskCategory(taskID string, categoryID *int64) (*pgdb.TaskC
 	return category, nil
 }
 
+// SetTasksCategory applies one category change to several tasks at once. The
+// database write and the in-memory refresh share taskStateMu, so a concurrent
+// single-task update cannot interleave and leave a live DTO stale. The returned
+// set holds the ids that were actually moved; callers report the rest as gone.
+func (m *Manager) SetTasksCategory(taskIDs []string, categoryID *int64) (map[string]bool, *pgdb.TaskCategory, error) {
+	m.taskStateMu.Lock()
+	defer m.taskStateMu.Unlock()
+	numeric := make([]int64, 0, len(taskIDs))
+	for _, taskID := range taskIDs {
+		id, err := strconv.ParseInt(taskID, 10, 64)
+		if err != nil || id <= 0 {
+			return nil, nil, pgdb.ErrTaskCategoryTaskNotFound
+		}
+		numeric = append(numeric, id)
+	}
+	updatedIDs, category, err := m.pg.SetTasksCategory(numeric, categoryID)
+	if err != nil {
+		return nil, nil, err
+	}
+	categoryName := ""
+	if category != nil {
+		categoryName = category.Name
+	}
+	updated := make(map[string]bool, len(updatedIDs))
+	m.mu.RLock()
+	tasks := make([]*Task, 0, len(updatedIDs))
+	for _, id := range updatedIDs {
+		taskID := strconv.FormatInt(id, 10)
+		updated[taskID] = true
+		if task := m.tasks[taskID]; task != nil {
+			tasks = append(tasks, task)
+		}
+	}
+	m.mu.RUnlock()
+	for _, task := range tasks {
+		task.updateLifecycle(func(state *taskLifecycleState) {
+			state.CategoryID = cloneInt64Ptr(categoryID)
+			state.CategoryName = categoryName
+		})
+	}
+	return updated, category, nil
+}
+
 // DeleteCompanyWithAssets keeps the database cascade and live task handles in
 // one manager-level critical section. This closes the gap where a task could
 // commit its company scope immediately before registration and miss the