Refactors
- Updated all imports and type references to use zod/v4.
- Replaced deprecated or changed Zod APIs and options across the codebase.
- Adjusted validation schemas and error messages to match Zod v4 requirements.
- Updated dependencies and lockfiles to use Zod v4.
* refactor(adapter): `createAdapter` helper
* update: Kysely adapter support
* update: memory adapter
* chore: cleanup
* update: MongoDB adapter supported
* update: mongodb adapter file names
* update: support for prisma adapter
- also fixed memory adapter with `getField`
- disabled all debug logs by default
* chore: lint
* fix: generate an `id` if it isn't already provided
* update(test): init config snapshot
* update: existing adapters tests to include an `id` existence check
* fix: renamed prisma adapter file names
* ^
* update(mongo): Removed `disableIdGeneration` for mongo adapter
* fix: dont add `id` from fields if `disableIdGeneration` is false
* fix: adapter converting dates to string using the wrong value data
* add: create-adapter tests
* update: moved `adapter` to `create-adapter` under the existing `adapters` folder
* WIP: docs
* fix: create-adapter test importing from wrong path
* WIP: docs
* update: key transformation map options are now objs not fns
* chore: lint
* update: cleanup & added debugLogs/usePlural option to some adapters
* update: tests
* chore: cleanup
* update: fields name & made customTransforminput/output to run after supportsDates/supportsBoolean/SupportsJSON to run
* chore: lint
* Update: docs
* update: update-many to not use transformOutput
* chore: lint
* fix: adapter options breaking
* update: set the default limit on findMany
* update: docs
* update: docs & cleaned up code
* update: create schema comes with `tables` now
* This is useless
* fix: api-key shouldn't handle id gen in row creation
* chore: lint
* WIP: useNumberId
* chore: cleanup
* removed all cases of `id` during adapter.create call
* update: useNumberId
* update: drizzle adapter schema generation
- update: pg now sets `id` field to use `uuid`.
- fix: sets the correct reference `onDelete` action.
- refactor: cleaned up the code a lot more
* update: added `supportNumericIds` in create adapter options
* chore: cleanup
* update(tests): added option to disable any adapter tests & created a `useNumberId` test
* fix: skip tests options were using old method
* chore: lint
* WIP: Drizzle tests with number id
* chore: lint
* fix: memory adapter failing on number id adapter tests
* chore: cleanup
* fix: adapter tests fail due to emails needing to be unique
* update: support drizzle adapter use-number-id
* fix: cleanup Drizzle Adapter tests
* fix: drizzle schema generation for reference id fields
* fix: type errs in memory adapter tests
* fix: memory adapter tests
* Update init.test.ts.snap
* fix: prisma tests correctly uses the right primsa client per schema
* Delete run-vitest.ts
* update: create-adapter to support `getDefaultFieldName`
* update: create-adapter adapter options updated
* fix: adapters using older function names
* fix: create-adapter now converts where clauses & added tests
* update: new `getFieldAttributes` option in create-adapter, and fixed convert where clause fn
* chore: cleanup
* update: convertWhereClause in create-adapter improved
* update: BAOptions to start using `database` & updated adapter tests
* fix: type errors
* chore: lint
* update: default config values for the create-adapter config
* fix: getModelName doesnt take into account of plural
* chore: cleanup
* update: improved getDefaultModelname
* fix: create-adapter transform input doesn't take into account reference IDs
* fix: transformInput on reference IDs doesn't take into account of array IDs
* chore: lint
* fix: prisma tests
* fix: Prisma adapter tests running one after another
* fix: init snapshot should state supportJSON as `false`
* fix: drizzle adapter tests
* update: adapter test options to allow passing a test prefix
* chore: add state.txt to gitignore
* remove: state.txt from gitignore, it's useless
* chore: lint
* Update adapter.prisma.numberid.test.ts
* fix: get-migration for mysql & mssql `id` fields to use varchar
* chore: cleanup
* update: adapter comes with adapter-test-debug-log functions
* update: made adapterTestDebugLog functions only be in adapter if config enabled it
* fix: transactionId for debug logs not working correctly
This was due to the adapter being reinitialized each test case
* update: Added colors to debug logs
* update: adapter tests a little more refined
* add: deepmerge dev deps to deep merge better auth options config for adapters
* fix: create-adapter types
* update: revert back to old types
* fix: prisma adapter tests now run one after another
* fix: kysely adapter to work when no `id` is provided
* update: mongoDB to set `supportNumericids` to false
* update(docs): discuss new `database` object in BetterAuthOptions
* add: support for custom ID generation
* fix: docs had incorrect default value for `supportsJSON`
* update(docs): added number id adapter test documentation
* fix: drizzle-adapter mysql tests
* update: drizzle-adapter tests to use deepmerge
* add: drizzle-adapter mysql tests to test useNumberId
* add: Prisma generate to support number id
* update: support the old `generateId` option, but warn the user for deprecation
* update: init test snapshot
* update: adapter tests not included in normal tests
* Update pnpm-lock.yaml
* fix: typo in file names for kysley test files
* update: separated adapter tests
* Update pnpm-lock.yaml
* fix: please tell me I actually fixed this
* fix: pnpm-lock merge conflict
* chore: lint
* fix: sveltekit pkg.json merge conflicts (hopefully)
* Delete pnpm-lock.yaml
* update: createSchema takes into account of rate-limit and secondaryStorage
* improve docs
* add drizzle returning id retriver
* chore: fix test script
---------
Co-authored-by: Bereket Engida <bekacru@gmail.com>
