Mongoose 查询、链式调用和精简文档
您将串联 Mongoose 查询辅助方法,使用 .lean() 获得原始 POJO 性能,并将查询 API 与原生驱动程序进行比较。
Mongoose 查询、链式调用和精简文档 是 CoddyKit 上的免费 MongoDB Academy 课时。 这是第 3 节课,共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 MongoDB Academy 学习路径的一部分,你的进度在网页和 CoddyKit 应用中同步。 MongoDB Academy 课程共包含 4 节课。
本课时的部分内容尚未翻译,以英文显示。
Mongoose Query Objects
When you call a Mongoose query method like User.find(), it returns a Query object rather than a Promise. This Query object is lazy—it does not execute until you explicitly call it with .then(), await, or .exec(). Before execution, you can chain additional query modifiers to build up the complete query. This chainable API is one of Mongoose's most ergonomic features.
const User = require('./models/user');
// This does NOT execute immediately — returns a Query object
const query = User.find({ active: true });
// Now execute it with await
const users = await query;
// Or chain modifiers before executing:
const result = await User.find({ active: true })
.sort({ createdAt: -1 })
.limit(10)
.select('name email -_id');
// select() projects fields: '+field' includes, '-field' excludesChaining Query Modifiers
Mongoose query modifiers like .sort(), .limit(), .skip(), .select(), and .populate() can be chained in any order before execution. The underlying Query object accumulates all modifiers and sends a single optimized query to MongoDB. This is functionally equivalent to passing options to the native driver's find(filter, options) but reads more naturally as a fluent builder.
const orders = await Order
.find({ status: 'completed', userId: currentUserId })
.sort({ createdAt: -1 }) // newest first
.skip(page * pageSize) // pagination offset
.limit(pageSize) // page size
.select('_id total status createdAt') // projection
.lean(); // return plain objects (discussed next)
console.log('Orders on this page:', orders.length);The .lean() Method: Raw POJO Performance
By default, Mongoose wraps every document returned by a query in a Mongoose Document instance—a heavy object with change tracking, methods, virtuals, and middleware hooks. The .lean() method tells Mongoose to return plain JavaScript objects (POJOs) instead. Lean queries are typically 2-5x faster and use less memory because Mongoose skips the Document wrapping. Use .lean() for read-only operations where you don't need document methods or save/update hooks.
// Without .lean() — heavy Mongoose Document objects
const docsWithMethods = await User.find({ active: true });
// docsWithMethods[0].save() works, but incurs overhead
// With .lean() — plain JavaScript objects, much faster
const pureObjects = await User.find({ active: true }).lean();
// pureObjects[0].save() does NOT work — it's a plain object
// But JSON.stringify, spread operators, and array methods are all faster
console.log(typeof docsWithMethods[0].save); // 'function'
console.log(typeof pureObjects[0].save); // 'undefined'When to Use .lean() vs Full Documents
Use .lean() when: you are only reading data (GET endpoints), you need to serialize to JSON quickly, or you are processing many documents in bulk. Do not use .lean() when: you need to call .save() on the result, use virtual properties, run document middleware, or access instance methods. A good rule of thumb: API reads → lean, mutation flows → full Mongoose documents.
// API read endpoint — use .lean() for speed
router.get('/products', async (req, res) => {
const products = await Product.find({}).lean(); // fastest, no doc wrapper
res.json(products);
});
// Update endpoint — use full Mongoose document to access instance methods
router.post('/users/:id/deactivate', async (req, res) => {
const user = await User.findById(req.params.id); // full document, NO .lean()
await user.sendDeactivationEmail(); // instance method won't work with .lean()
user.active = false;
await user.save(); // document method won't work with .lean()
res.json({ success: true });
});findById and findOne Convenience Methods
Mongoose adds convenience query methods not available in the native driver. Model.findById(id) is equivalent to Model.findOne({ _id: id }) and automatically converts string IDs to ObjectId. Model.findByIdAndUpdate(id, update, options) and Model.findByIdAndDelete(id) combine lookup and modification in a single atomic operation. These methods greatly reduce boilerplate in CRUD route handlers.
// findById — automatic ObjectId conversion from string
const user = await User.findById('64a1b2c3d4e5f6789012345a').lean();
// findByIdAndUpdate — find, update, and return result atomically
const updatedProduct = await Product.findByIdAndUpdate(
productId,
{ $set: { price: 199.99 }, $inc: { updateCount: 1 } },
{ new: true, runValidators: true } // return new doc, run validators
);
// findByIdAndDelete — find and delete atomically
const deletedUser = await User.findByIdAndDelete(userId);
console.log('Deleted:', deletedUser ? deletedUser.email : 'not found');Counting Documents
Mongoose provides efficient document counting methods. Model.countDocuments(filter) applies a filter and counts matching documents—it scans matching documents and uses indexes. Model.estimatedDocumentCount() uses collection metadata for an approximate but instantaneous count without a filter—useful for dashboard totals on large collections where exact counts are not critical.
// Exact count with filter — uses an index if available
const activeUsers = await User.countDocuments({ active: true, role: 'user' });
console.log('Active users:', activeUsers);
// Fast approximate count — no filter, uses collection stats
const totalProducts = await Product.estimatedDocumentCount();
console.log('Approximate total products:', totalProducts);
// In Express pagination:
const [data, total] = await Promise.all([
User.find({}).skip(offset).limit(pageSize).lean(),
User.countDocuments({})
]);
res.json({ data, total, pages: Math.ceil(total / pageSize) });Populate: Resolving References
.populate() is one of Mongoose's most powerful features—it replaces an ObjectId reference field with the actual referenced document, fetched from another collection. Under the hood, Mongoose issues a second query to the referenced collection and substitutes the IDs. It is equivalent to $lookup in the aggregation pipeline but with a simpler API.
const Order = mongoose.model('Order', new mongoose.Schema({
userId: { type: mongoose.Schema.Types.ObjectId, ref: 'User' },
productIds: [{ type: mongoose.Schema.Types.ObjectId, ref: 'Product' }]
}));
// Populate the userId reference with the full User document
const order = await Order
.findById(orderId)
.populate('userId', 'name email') // only select name and email from User
.populate('productIds', 'name price') // populate array of references
.lean();
console.log(order.userId.email); // 'alice@example.com'
console.log(order.productIds[0].name); // 'Laptop'Mongoose vs Native Driver: When to Choose
Mongoose adds validation, populate, middleware, and a convenient query API—at the cost of some overhead. Choose Mongoose when: your application has well-defined, stable schemas; you want schema validation without JSON Schema validators; you need populate for reference resolution; or you are building a conventional REST API. Choose the native driver when: you need maximum performance, are working with dynamic schemas, are building aggregation-heavy analytics, or are writing a microservice with minimal dependencies.
// Mongoose: ergonomic, validates, populate works
const user = await User.findOne({ email }).select('-password').populate('profile');
// Native driver: faster, raw, no middleware
const user = await db.collection('users')
.findOne({ email }, { projection: { password: 0 } });exec() and Error Handling
Calling .exec() explicitly converts a Mongoose Query to a Promise and is the traditional way to execute queries when using .catch() promise chaining. With async/await, you can omit .exec()—a bare await User.find({}) works fine. However, some developers prefer .exec() for clarity or when building query objects programmatically. Both patterns produce identical results.
// With .exec() — explicit Promise conversion
const user = await User.findOne({ email }).exec();
// Without .exec() — implicit execution via await
const user = await User.findOne({ email });
// Error handling with try/catch (both forms work the same)
try {
const user = await User.findById(id);
if (!user) throw new Error('User not found');
} catch (err) {
if (err.name === 'CastError') {
res.status(400).json({ error: 'Invalid ID format' });
} else {
res.status(500).json({ error: err.message });
}
}Query Builder Pattern
Because Mongoose queries are lazy, you can conditionally build queries in separate statements before execution. This is useful when query parameters are optional—add sort or filters only when the parameter is present. This pattern is much cleaner than constructing dynamic query strings and keeps the code readable.
async function searchProducts(filters) {
let query = Product.find();
if (filters.category) {
query = query.where('category').equals(filters.category);
}
if (filters.maxPrice) {
query = query.where('price').lte(filters.maxPrice);
}
if (filters.inStock) {
query = query.where('stock').gt(0);
}
const sortField = filters.sortBy || 'createdAt';
query = query.sort({ [sortField]: -1 }).limit(50).lean();
return query; // executes here via await in the caller
}Aggregate Pipeline in Mongoose
Mongoose models also support the aggregation pipeline via Model.aggregate(pipeline). Unlike regular Mongoose queries, aggregation bypasses schema casting, Mongoose middleware, and populate—it behaves similarly to calling the native driver's aggregate directly. Aggregate returns a plain array of objects, never Mongoose Documents. Use Model.aggregate() for complex analytics and reporting that don't benefit from Mongoose's abstractions.
// Aggregation in Mongoose — bypasses Mongoose middleware and casting
const salesByRegion = await Order.aggregate([
{ $match: { status: 'completed' } },
{
$group: {
_id: '$region',
totalRevenue: { $sum: '$total' },
orderCount: { $sum: 1 },
avgOrder: { $avg: '$total' }
}
},
{ $sort: { totalRevenue: -1 } }
]);
// salesByRegion is a plain array — no Mongoose Document wrapperQuick Check
Test your understanding of MongoDB & NoSQL Databases concepts from this lesson.
Lesson Recap
In this lesson you learned: Mongoose query methods return lazy Query objects that can be chained with .sort(), .limit(), .skip(), .select(), and .populate() before execution, .lean() returns plain JavaScript objects for better performance on read-only operations, and Model.aggregate() bypasses Mongoose abstractions and behaves like the native driver for analytics pipelines. Next up we explore Mongoose middleware — pre and post hooks for custom logic around save, find, and other operations.
常见问题解答
「Mongoose 查询、链式调用和精简文档」课时是免费的吗?
是的 — 「Mongoose 查询、链式调用和精简文档」的完整文本可在网页上免费阅读。要进行交互式练习(内置代码编辑器和全天候 AI 导师)并解锁 MongoDB Academy 课程的其余内容,请升级到 CoddyKit PRO。 MongoDB Academy 课程共包含 4 节课。
「Mongoose 查询、链式调用和精简文档」这节课中我会学到什么?
您将串联 Mongoose 查询辅助方法,使用 .lean() 获得原始 POJO 性能,并将查询 API 与原生驱动程序进行比较。 你通过在浏览器中直接运行的动手代码来练习 MongoDB Academy,全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 MongoDB Academy 需要有经验吗?
无需任何先前经验。CoddyKit 上的 MongoDB Academy 课程适合初学者到高级学习者,你可以从这里开始或从头开始,按照自己的节奏学习。 这是第 3 节课,共 4 节。
「Mongoose 查询、链式调用和精简文档」课时需要多长时间?
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动,所以你能稳步进步,并在网页和应用中从离开的地方继续。
我能在这节 MongoDB Academy 课中编写并运行代码吗?
能。每节 MongoDB Academy 课都包含内置代码编辑器,你可以在浏览器中直接编写并运行真实代码,并获得即时 AI 反馈 — 无需本地设置。
此课程中的所有课时
- 使用官方 Node.js 驱动程序连接
- Mongoose 模式、模型和虚拟属性
- Mongoose 查询、链式调用和精简文档
- Mongoose 中间件:前置和后置钩子