Introduction

Welcome to Wolfgang.Etl.SqlBulkCopy!

Overview

A loader that uses SqlBulkCopy for fast inserts into a Microsoft SQL database.

It consumes an IAsyncEnumerable<TRecord> and streams it to SQL Server through SqlBulkCopy, deriving the schema, table and column mapping from the record type itself — so there is no manual ColumnMappings wiring to keep in sync. It is built on Wolfgang.Etl.Abstractions, so it composes with the rest of the ETL family.

Key Features

  • Streaming bulk load — consumes IAsyncEnumerable<TRecord> and writes in configurable batches, so memory stays flat regardless of source size.
  • Type-driven mapping — [Table], [Column] and [NotMapped] attributes drive the schema, table and column names.
  • Nested tables — child collections are recursively written to their own tables inside the same bulk-copy session.
  • Pre/post actions — built-in TruncateTable / DeleteAllRecords, plus custom-action delegates for schema-aware work.
  • Data validation — opt in with EnableDataValidation; DataAnnotations failures either throw or skip per ValidationFailureBehavior, with OnValidationFailed / OnNestedValidationFailed callbacks.
  • Transactions — an optional SqlTransaction participates in both the bulk load and the pre/post commands.
  • Dry run — set IsDryRun to enumerate, map, validate and report with no SQL side effects.
  • Progress reporting — IProgress<SqlBulkCopyReport> reports rows written (CurrentItemCount), rows skipped and batch count.
  • Native AOT ready — mark a record [BulkCopyable] to get compile-time source-generated accessors instead of runtime IL emission.
  • Async-only — a banned-symbol analyzer keeps the library free of synchronous fallbacks.

Getting Help

If you need help with Wolfgang.Etl.SqlBulkCopy, please:

Edit this page