MongoDB聚合：嵌套数组架构下$match与$sum操作异常排查

Got it, let's break down why your aggregation is returning an empty array and fix it to get the total quantity of items across all your carts.

First, Let's Recap Your Schema Setup #

I'll assume your schemas look something like this (based on your description):

// CartItem Sub-Schema
const CartItemSchema = new Schema({
  productId: { type: Schema.Types.ObjectId, ref: 'Product' },
  quantity: { type: Number, required: true, min: 1 }
});

// Top-level Cart Schema
const CartSchema = new Schema({
  userId: { type: Schema.Types.ObjectId, ref: 'User', required: true },
  items: [CartItemSchema], // Nested array of cart items
  createdAt: { type: Date, default: Date.now }
});

The Problem With Your Original Aggregation #

If your original code looked something like this (just using $sum directly as a pipeline stage):

// ❌ This returns empty array because $sum can't be a top-level stage
Cart.aggregate([
  { $sum: "$items.quantity" }
])

The issue is that $sum is an accumulator operator, not a standalone pipeline stage. It only works inside stages like $group or $project. Plus, since items is a nested array, we need to handle that nested quantity sum first.

Fixed Aggregation Pipelines #

Here are two working approaches to get the total item quantity across all carts:

Approach 1: Group Globally (No Unwind Needed)

This is more efficient because it avoids unwinding the array:

// ✅ Calculate total quantity without unwinding the items array
Cart.aggregate([
  {
    $group: {
      _id: null, // Group all documents into a single global result
      totalQuantity: {
        // Inner $sum adds up quantities for one cart's items
        // Outer $sum adds that total across all carts
        $sum: { $sum: "$items.quantity" }
      }
    }
  },
  // Optional: Clean up the output to remove _id and rename the field
  {
    $project: {
      _id: 0,
      totalItems: "$totalQuantity"
    }
  }
])

Approach 2: Unwind First (For Granular Processing)

If you need to do extra filtering or processing on individual cart items first, use $unwind:

// ✅ Unwind the items array, then sum all quantities
Cart.aggregate([
  // Split each cart item into its own document
  { $unwind: "$items" },
  {
    $group: {
      _id: null,
      totalQuantity: { $sum: "$items.quantity" }
    }
  },
  { $project: { _id: 0, totalItems: "$totalQuantity" } }
])

Verification With Your mLab Data #

Let's say you have a sample cart document like this (from your mLab setup):

// Sample Cart Document
{
  "_id": ObjectId("60d21b4667d0d8992e610c85"),
  "userId": ObjectId("60d21b2467d0d8992e610c84"),
  "items": [
    { "productId": ObjectId("60d21b1067d0d8992e610c83"), "quantity": 2 },
    { "productId": ObjectId("60d21b0067d0d8992e610c82"), "quantity": 3 }
  ],
  "createdAt": ISODate("2021-06-23T12:00:00Z")
}

Running either of the fixed pipelines will return:

[ { "totalItems": 5 } ]

Key Takeaways #

• Never use $sum as a top-level pipeline stage — it must be inside $group or $project.
• For nested arrays, use a nested $sum (to calculate per-cart totals) before summing across all carts, or unwind the array first if you need to process individual items.

内容的提问来源于stack exchange，提问作者ozer