clean up merge conflicts

Author: vkarpov15

.github/workflows/encryption-tests.yml

@@ -4,7 +4,7 @@ on:
   push: 
     branches: ['master']
   pull_request:
-    branches: [ 'master' ]
+    branches: [ 'master', 'csfle' ]
   workflow_dispatch: {}
 
 permissions:

.gitignore

@@ -72,4 +72,4 @@ notes.md
 list.out
 
 data
-*.pid
+fle-cluster-config.json

CHANGELOG.md

@@ -1,3 +1,12 @@
+8.15.0 / 2025-05-16
+===================
+ * feat: CSFLE support #15390 [baileympearson](https://github.com/baileympearson)
+ * feat: add strictFilter option to findOneAndUpdate (#14913) #15402 #14913 [muazahmed-dev](https://github.com/muazahmed-dev)
+ * feat(error): set cause to MongoDB error reason on ServerSelection errors #15420 #15416
+ * fix(model): make bulkSave() rely on document.validateSync() to validate docs and skip bulkWrite casting #15415 #15410
+ * types: stricter projection typing with 1-level deep nesting #15418 #15327 #13840 [pshaddel](https://github.com/pshaddel)
+ * docs: emphasize automatic type inference in TypeScript intro and statics/methods, remove duplicated statics.md #15421
+
 8.14.3 / 2025-05-13
 ===================
  * types(schema): allow post('init') #15413 #15412 #15333

docs/field-level-encryption.md

@@ -16,11 +16,141 @@ The resulting document will look similar to the following to a client that doesn
 
 You can read more about CSFLE on the [MongoDB CSFLE documentation](https://www.mongodb.com/docs/manual/core/csfle/) and [this blog post about CSFLE in Node.js](https://www.mongodb.com/developer/languages/javascript/client-side-field-level-encryption-csfle-mongodb-node/).
 
-Note that Mongoose does **not** currently have any Mongoose-specific APIs for CSFLE.
-Mongoose defers all CSFLE-related work to the MongoDB Node.js driver, so the [`autoEncryption` option](https://mongodb.github.io/node-mongodb-native/5.6/interfaces/AutoEncryptionOptions.html) for `mongoose.connect()` and `mongoose.createConnection()` is where you put all CSFLE-related configuration.
-Mongoose schemas currently don't support CSFLE configuration.
+## Automatic FLE in Mongoose
 
-## Setting Up Field Level Encryption with Mongoose
+Mongoose supports the declaration of encrypted schemas - schemas that, when connected to a model, utilize MongoDB's Client Side
+Field Level Encryption or Queryable Encryption under the hood.  Mongoose automatically generates either an `encryptedFieldsMap` or a
+`schemaMap` when instantiating a MongoClient and encrypts fields on write and decrypts fields on reads.
+
+### Encryption types
+
+MongoDB has two different automatic encryption implementations: client side field level encryption (CSFLE) and queryable encryption (QE).  
+See [choosing an in-use encryption approach](https://www.mongodb.com/docs/v7.3/core/queryable-encryption/about-qe-csfle/#choosing-an-in-use-encryption-approach).
+
+###  Declaring Encrypted Schemas
+
+The following schema declares two properties, `name` and `ssn`.  `ssn` is encrypted using queryable encryption, and
+is configured for equality queries:
+
+```javascript
+const encryptedUserSchema = new Schema({ 
+  name: String,
+  ssn: { 
+    type: String, 
+    // 1
+    encrypt: { 
+      keyId: '<uuid string of key id>',
+      queries: 'equality'
+    }
+  }
+  // 2
+}, { encryptionType: 'queryableEncryption' });
+```
+
+To declare a field as encrypted, you must:
+
+1. Annotate the field with encryption metadata in the schema definition
+2. Choose an encryption type for the schema and configure the schema for the encryption type
+
+Not all schematypes are supported for CSFLE and QE.  For an overview of supported BSON types, refer to MongoDB's documentation.
+
+### Registering Models
+
+Encrypted schemas can be registered on the global mongoose object or on a specific connection, so long as models are registered before the connection
+is established:
+
+```javascript
+// specific connection
+const GlobalUserModel = mongoose.model('User', encryptedUserSchema);
+
+// specific connection
+const connection = mongoose.createConnection();
+const UserModel = connection.model('User', encryptedUserSchema);
+```
+
+### Connecting and configuring encryption options
+
+Field level encryption in Mongoose works by generating the encryption schema that the MongoDB driver expects for each encrypted model on the connection.  This happens automatically when the model's connection is established.
+
+Queryable encryption and CSFLE require all the same configuration as outlined in the [MongoDB encryption in-use documentation](https://www.mongodb.com/docs/manual/core/security-in-use-encryption/), except for the schemaMap or encryptedFieldsMap options.
+
+```javascript
+const keyVaultNamespace = 'client.encryption';
+const kmsProviders = { local: { key } };
+await connection.openUri(`mongodb://localhost:27017`, {
+  // Configure auto encryption
+  autoEncryption: {
+    keyVaultNamespace: 'datakeys.datakeys',
+    kmsProviders
+  }
+});
+```
+
+Once the connection is established, Mongoose's operations will work as usual.  Writes are encrypted automatically by the MongoDB driver prior to sending them to the server and reads are decrypted by the driver after fetching documents from the server.
+
+### Discriminators
+
+Discriminators are supported for encrypted models as well:
+
+```javascript
+const connection = createConnection();
+
+const schema = new Schema({
+  name: {
+    type: String, encrypt: { keyId }
+  }
+}, {
+  encryptionType: 'queryableEncryption'
+});
+
+const Model = connection.model('BaseUserModel', schema);
+const ModelWithAge = model.discriminator('ModelWithAge', new Schema({
+  age: {
+    type: Int32, encrypt: { keyId: keyId2 }
+  }
+}, {
+  encryptionType: 'queryableEncryption'
+}));
+
+const ModelWithBirthday = model.discriminator('ModelWithBirthday', new Schema({
+  dob: {
+    type: Int32, encrypt: { keyId: keyId3 }
+  }
+}, {
+  encryptionType: 'queryableEncryption'
+}));
+```
+
+When generating encryption schemas, Mongoose merges all discriminators together for all of the discriminators declared on the same namespace.  As a result, discriminators that declare the same key with different types are not supported.  Furthermore, all discriminators for the same namespace must share the same encryption type - it is not possible to configure discriminators on the same model for both CSFLE and Queryable Encryption.
+
+## Managing Data Keys
+
+Mongoose provides a convenient API to obtain a [ClientEncryption](https://mongodb.github.io/node-mongodb-native/Next/classes/ClientEncryption.html)
+object configured to manage data keys in the key vault.  A client encryption can be obtained with the `Model.clientEncryption()` helper:
+
+```javascript
+const connection = createConnection();
+
+const schema = new Schema({
+  name: {
+    type: String, encrypt: { keyId }
+  }
+}, {
+  encryptionType: 'queryableEncryption'
+});
+
+const Model = connection.model('BaseUserModel', schema);
+await connection.openUri(`mongodb://localhost:27017`, {
+  autoEncryption: {
+    keyVaultNamespace: 'datakeys.datakeys',
+    kmsProviders: { local: '....' }
+  }
+});
+
+const clientEncryption = Model.clientEncryption();
+```
+
+## Manual FLE in Mongoose
 
 First, you need to install the [mongodb-client-encryption npm package](https://www.npmjs.com/package/mongodb-client-encryption).
 This is MongoDB's official package for setting up encryption keys.

docs/typescript.md

@@ -8,10 +8,46 @@ This guide describes Mongoose's recommended approach to working with Mongoose in
 
 To get started with Mongoose in TypeScript, you need to:
 
-1. Create an interface representing a document in MongoDB.
-2. Create a [Schema](guide.html) corresponding to the document interface.
-3. Create a Model.
-4. [Connect to MongoDB](connections.html).
+1. Create a [Schema](guide.html).
+2. Create a Model.
+3. [Connect to MongoDB](connections.html).
+
+```typescript
+import { Schema, model, connect } from 'mongoose';
+
+// 1. Create a Schema corresponding to the document interface.
+const userSchema = new Schema({
+  name: { type: String, required: true },
+  email: { type: String, required: true },
+  avatar: String
+});
+
+// 2. Create a Model.
+const User = model('User', userSchema);
+
+run().catch(err => console.log(err));
+
+async function run() {
+  // 3. Connect to MongoDB
+  await connect('mongodb://127.0.0.1:27017/test');
+
+  const user = new User({
+    name: 'Bill',
+    email: 'bill@initech.com',
+    avatar: 'https://i.imgur.com/dM7Thhn.png'
+  });
+  await user.save();
+
+  const email: string = user.email;
+  console.log(email); // 'bill@initech.com'
+}
+```
+
+## Using Generics
+
+By default, Mongoose automatically infers the shape of your documents based on your schema definition.
+However, if you modify your schema after your `new Schema()` call (like with plugins) then Mongoose's inferred type may be incorrect.
+For cases where Mongoose's automatic schema type inference is incorrect, you can define a raw document interface that tells Mongoose the type of documents in your database as follows.
 
 ```typescript
 import { Schema, model, connect } from 'mongoose';
@@ -67,8 +103,6 @@ const user: HydratedDocument<IUser> = new User({
 });
 ```
 
-## ObjectIds and Other Mongoose Types
-
 To define a property of type `ObjectId`, you should use `Types.ObjectId` in the TypeScript document interface. You should use `'ObjectId'` or `Schema.Types.ObjectId` in your schema definition.
 
 ```ts
@@ -106,4 +140,4 @@ However, before you do, please [open an issue on Mongoose's GitHub page](https:/
 
 ## Next Up
 
-Now that you've seen the basics of how to use Mongoose in TypeScript, let's take a look at [statics in TypeScript](typescript/statics-and-methods.html).
+Now that you've seen the basics of how to use Mongoose in TypeScript, let's take a look at [methods in TypeScript](typescript/statics-and-methods.html).