Skip to content

migration.md

Latest commit

 

History

History
47 lines (35 loc) · 2.14 KB

migration.md

File metadata and controls

47 lines (35 loc) · 2.14 KB
Raw

12. Migration to SharpCoreDB

Migrating existing data and code from other engines. Deep dives: docs/migration/README.md · docs/migration/MIGRATION_GUIDE.md · docs/migration/SQLITE_VECTORS_TO_SHARPCORE.md

12.1 From SQLite

  • SQL dialect: SharpCoreDB implements a SQLite-compatible dialect subset — INTEGER PRIMARY KEY rowid alias, _rowid, COLLATE, INSERT OR REPLACE-style semantics — so most DDL/DML ports with minimal changes.
  • Data: use the built-in migration tooling in docs/migration/MIGRATION_GUIDE.md to copy tables and rows programmatically.
  • Vectors: replace SQLite vector extensions with SharpCoreDB.VectorSearch (SQLITE_VECTORS_TO_SHARPCORE.md).
  • Performance: switch hot read loops to FindByPrimaryKey / ExecuteQueryStruct and hot ingestion to InsertBatch — see the Performance Guide.

12.2 From LiteDB

  • Replace BSON document access with typed SQL tables and rows.
  • LiteCollection<T>.FindById → db.FindByPrimaryKey(table, key).
  • InsertBulk → db.InsertBatch.
  • You gain SQL, encryption, vector search, and analytics you did not have.

12.3 From RavenDB / MongoDB (network engines)

  • Embedded-first: use SharpCoreDB directly, or SharpCoreDB.Server for the gRPC network layer.
  • RLS replaces per-document authorization filters.
  • Document-centric schemas map cleanly to TEXT/BLOB columns with a ULID PK.

12.4 FluentMigrator

Add versioned migrations to any SharpCoreDB database with the SharpCoreDB.FluentMigrator package — see Providers.

12.5 Check list

  1. Port schema → CREATE TABLE (keep INTEGER PRIMARY KEY / ULID patterns for keys).
  2. Port data via a scripted copy (per-table ExecuteQuery + InsertBatch).
  3. Re-create indexes (CREATE INDEX, hash for point lookups, B-tree for ranges).
  4. Switch hot loops to the v2.0 fast-path APIs.
  5. Run the comparative benchmark to confirm target numbers on your machine.