v8.10.1

Compare Source

===================

• perf(document): only call undoReset() 1x/document #​15257 #​15255
• perf(schema): clear childSchemas when overwriting existing path to avoid performance degradations #​15256 #​15253
• perf: some more micro optimizations for find() and findOne() #​14906 #​15250
• fix(model): avoid adding timeout on Model.init() buffering to avoid unintentional dangling open handles #​15251 #​15241
• fix: avoid connection buffering on init if autoCreate: false #​15247 #​15241
• fix: infer discriminator key if set in $set with overwriteDiscriminatorKey #​15243 #​15218
• types(middleware): make this in document middleware the hydrated doc type, not raw doc type #​15246 #​15242
• types(schema): support options parameter to Schema.prototype.discriminator() #​15249 #​15244
• types(schema): allow calling Schema.prototype.number() with no message arg #​15237 #​15236
• docs(typescript): recommend using HydratedSingleSubdocument over Types.Subdocument #​15240 #​15211

v8.10.0

Compare Source

===================

• feat(schema+schematype): add toJSONSchema() method to convert schemas and schematypes to JSON schema #​15184 #​11162
• feat(connection): make connection helpers respect bufferTimeoutMS #​15229 #​15201
• feat(document): support schematype-level transform option #​15163 #​15084
• feat(model): add insertOne() function to insert a single doc #​15162 #​14843
• feat(connection): support Connection.prototype.aggregate() for db-level aggregations #​15153
• feat(model): make syncIndexes() not call createIndex() on indexes that already exist #​15175 #​12250
• feat(model): useConnection(connection) function #​14802
• fix(model): disallow updateMany(update) and fix TypeScript types re: updateMany() #​15199 #​15190
• fix(collection): avoid buffering if creating a collection during a connection interruption #​15187 #​14971
• fix(model): throw error if calling create() with multiple docs in a transaction unless ordered: true #​15100
• fix(model): skip createCollection() in syncIndexes() if autoCreate: false #​15155
• fix(model): make hydrate() handle hydrating deeply nested populated docs with hydratedPopulatedDocs #​15130
• types(document): make sure toObject() and toJSON() apply versionKey __v #​15097
• ci(NODE-6505): CI Setup for Encryption Support #​15139 aditi-khare-mongoDB

v8.9.7

Compare Source

==================

• fix: avoid applying defaults on map embedded paths #​15217 #​15196
• types: add missing $median operator to aggregation types #​15233 #​15209
• docs(document): clarify that toObject() returns a POJO that may contain non-POJO values #​15232 #​15208

v8.9.6

Compare Source

==================

• fix(document): allow setting values to undefined with set(obj) syntax with strict: false #​15207 #​15192
• fix(schema): improve reason for UUID cast error, currently a TypeError #​15215 #​15202
• fix(aggregate): improve error when calling near() with invalid coordinates #​15206 #​15188

v8.9.5

Compare Source

==================

• fix: disallow nested $where in populate match CVE-2025-23061
• fix(schema): handle bitwise operators on Int32 #​15176 #​15170

v8.9.4

Compare Source

==================

• fix(document): fix document not applying manual populate when using a function in schema.options.ref #​15138 IchirokuXVI
• fix(model): make Model.validate() static correctly cast document arrays #​15169 #​15164
• fix(model): allow passing validateBeforeSave option to bulkSave() to skip validation #​15161 #​15156
• fix(schema): allow multiple self-referencing discriminator schemas using Schema.prototype.discriminator #​15142 #​15120
• types: avoid BufferToBinary<> wiping lean types when passed to generic functions #​15160 #​15158
• docs: fix <code> in header ids #​15159
• docs: fix header in field-level-encryption.md #​15137 damieng

v8.9.3

Compare Source

==================

• fix(schema): make duplicate index error a warning for now to prevent blocking upgrading #​15135 #​15112 #​15109
• fix(model): handle document array paths set to non-array values in Model.castObject() #​15124 #​15075
• fix(document): avoid using childSchemas.path for compatibility with pre-Mongoose-8.8 schemas #​15131 #​15071
• fix(model): avoid throwing unnecessary error if updateOne() returns null in save() #​15126
• perf(cursor): clear the stack every time if using populate with batchSize to avoid stack overflows with large docs #​15136 #​10449
• types: make BufferToBinary avoid Document instances #​15123 #​15122
• types(model+query): avoid stripping out virtuals when calling populate with paths generic #​15132 #​15111
• types(schema): add missing removeIndex #​15134
• types: add cleanIndexes() to IndexManager interface #​15127
• docs: move search endpoint to netlify #​15119

v8.9.2

Compare Source

==================

• fix(schema): avoid throwing duplicate index error if index spec keys have different order or index has a custom name #​15112 #​15109
• fix(map): clean modified subpaths when overwriting values in map of subdocs #​15114 #​15108
• fix(aggregate): pull session from transaction local storage for aggregation cursors #​15094 IchirokuXVI
• types: correctly handle union types in BufferToBinary and related helpers #​15103 #​15102 #​15057
• types: add UUID to RefType #​15115 #​15101
• docs: remove link to Mongoose 5.x docs from dropdown #​15116
• docs(connection+document+model): remove remaining references to remove(), clarify that deleteOne() does not execute until then() or exec() #​15113 #​15107

v8.9.1

Compare Source

==================

• fix(connection): remove heartbeat check in load balanced mode #​15089 #​15042 #​14812
• fix(discriminator): gather childSchemas when creating discriminator to ensure $getAllSubdocs() can properly get all subdocs #​15099 #​15088 #​15092
• fix(model): handle discriminators in castObject() #​15096 #​15075
• fix(schema): throw error if duplicate index definition using unique in schema path and subsequent .index() call #​15093 #​15056
• fix: mark documents that are populated using hydratedPopulatedDocs option as populated in top-level doc #​15080 #​15048
• fix(document+schema): improve error message for get() on invalid path #​15098 #​15071
• docs: remove more callback doc references & some small other changes #​15095

v8.9.0

Compare Source

==================

• feat: upgrade mongodb -> 6.12
• feat: add int32 schematype #​15054 aditi-khare-mongoDB
• feat: add double schematype #​15061 aditi-khare-mongoDB
• feat: allow specifying error message override for duplicate key errors unique: true #​15059 #​12844
• feat(connection): add support for Connection.prototype.bulkWrite() with MongoDB server 8.0 #​15058 #​15028
• feat: add forceRepopulate option for populate() to allow avoiding repopulating already populated docs #​15044 #​14979
• fix(connection): remove heartbeat check in load balanced mode #​15089 #​15042
• fix(query): clone PopulateOptions when setting _localModel to avoid state leaking between subpopulate instances #​15082 #​15026
• types: add splice() to DocumentArray to allow adding partial objects with splice() #​15085 #​15041
• types(aggregate): add $firstN, $lastN, $bottom, $bottomN, $minN and $maxN operators #​15087 mlomnicki
• docs: Remove merge conflict markers #​15090 sponrad

v8.8.4

Compare Source

==================

• fix: cast using overwritten embedded discriminator key when set #​15076 #​15051
• fix: avoid throwing error if saveOptions undefined when invalidating subdoc cache #​15062

v8.8.3

Compare Source

==================

• fix: disallow using $where in match
• perf: cache results from getAllSubdocs() on saveOptions, only loop through known subdoc properties #​15055 #​15029
• fix(model+query): support overwriteDiscriminatorKey for bulkWrite updateOne and updateMany, allow inferring discriminator key from update #​15046 #​15040

v8.8.2

Compare Source

==================

• fix(model): handle array filters when casting bulkWrite #​15036 #​14978
• fix(model): make diffIndexes() avoid trying to drop default timeseries collection index #​15035 #​14984
• fix: save execution stack in query as string #​15039 durran
• types(cursor): correct asyncIterator and asyncDispose for TypeScript with lib: 'esnext' #​15038
• docs(migrating_to_8): add note about removing findByIdAndRemove #​15024 dragontaek-lee