MongoDB — Documents, CRUD & the Aggregation Framework

db.orders.insertOne({
  customerId: "cust_4521",
  status: "PENDING",
  items: [{ sku: "SKU-100", qty: 2, price: 799 }],
  total: 1598
});

db.orders.findOne({ customerId: "cust_4521" });

db.orders.updateOne(
  { _id: ObjectId("64f1a2b3c4d5e6f7a8b9c0d1") },
  { $set: { status: "SHIPPED" }, $inc: { version: 1 } }
);

db.orders.deleteOne({ _id: ObjectId("64f1a2b3c4d5e6f7a8b9c0d1") });

find() returns a cursor, not an array — the driver fetches results in batches as you iterate, not all at once. For a small result set this distinction rarely matters; for a query that could match millions of documents, iterating the cursor (rather than materializing it into a list up front) is what keeps memory bounded on both the client and the server.

// total > 1000 AND status in [SHIPPED, DELIVERED] — multiple top-level fields in
// one filter object are an IMPLICIT $and; $or must be written explicitly
db.orders.find({
  total: { $gt: 1000 },
  status: { $in: ["SHIPPED", "DELIVERED"] }
});

// at least one item with qty > 1 AND sku "SKU-100" — on the SAME array element,
// not just "any qty > 1 somewhere and any SKU-100 somewhere"
db.orders.find({
  items: { $elemMatch: { qty: { $gt: 1 }, sku: "SKU-100" } }
});

A reporting task: "the 10 customers who spent the most on delivered orders this month, with their name attached." This genuinely needs a filter, a per-customer sum across a nested array, a join to another collection, and a sort — a good showcase of stages actually composing.

db.orders.aggregate([
  // 1. filter down to what we care about FIRST — uses an index on {status, createdAt}
  { $match: {
      status: "DELIVERED",
      createdAt: { $gte: ISODate("2026-01-01T00:00:00Z") }
  }},

  // 2. one order document with 3 items becomes 3 documents, one per item
  { $unwind: "$items" },

  // 3. group back up, but now by customer, summing item price*qty across the WHOLE month
  { $group: {
      _id: "$customerId",
      totalSpent: { $sum: { $multiply: ["$items.price", "$items.qty"] } },
      orderIds: { $addToSet: "$_id" }
  }},

  // 4. reshape — rename _id back to something readable, count distinct orders
  { $project: {
      _id: 0,
      customerId: "$_id",
      totalSpent: 1,
      ordersCount: { $size: "$orderIds" }
  }},

  // 5. join in the customer's name from the customers collection
  { $lookup: {
      from: "customers",
      localField: "customerId",
      foreignField: "_id",
      as: "customer"
  }},
  { $unwind: "$customer" },

  { $sort: { totalSpent: -1 } },
  { $limit: 10 }
]);