prisma-driver-adapter-implementation

Implement SqlDriverAdapterFactory with a connect method returning a SqlDriverAdapter that provides queryRaw, executeRaw, executeScript, startTransaction, and dispose. Map arguments using both value and ArgType, return column metadata in array row mode, and wrap driver failures in DriverAdapterError.

startTransaction acquires one dedicated connection, issues BEGIN, applies the isolation level, and returns a Transaction bound to that connection. commit and rollback are lifecycle cleanup hooks only, while optional createSavepoint, rollbackToSavepoint, and releaseSavepoint methods handle nested transactions.

Yes, implement SqlMigrationAwareDriverAdapterFactory with connectToShadowDb when the provider can create an isolated shadow database. The shadow database must use unique quoted identifiers, never point at the primary database, and be dropped during disposal or failure cleanup.

P2039 surfaces when an unmapped driver error reaches Prisma without structured mapping. Preserve originalCode and originalMessage in the DriverAdapterError so the fallback remains useful, and rethrow genuinely unexpected non-driver errors instead of fabricating GenericJs ids.

Common mistakes include sharing one connection across concurrent transactions, issuing duplicate SQL COMMIT or ROLLBACK inside lifecycle hooks, tracking savepoint depth globally on the adapter, splitting migration scripts on semicolons, and truncating 64-bit integers during result mapping.