RepoDb.Firebird.BulkOperations 1.17.0

<div align="center"> <a href="https://repodb.net/tutorial/get-started-firebird/"><image src="Firebird.png" style="width:256px;" /></a> <br/> <span style="font-size:28px;font-weight:bold;"><a href="https://repodb.net/tutorial/get-started-firebird/"><strong>RepoDb.Firebird.BulkOperations</strong></a></span> <br/> <span style="font-size:16px;">A high-performance data productivity platform for bulk operations on Firebird in .NET.</span> </div>

<br/>

RepoDb.Firebird.BulkOperations

An extension library of RepoDB that does bulk operations towards a Firebird database. Firebird's ADO.NET provider has no SqlBulkCopy-equivalent bulk-copy class, so this uses FbBatchCommand - a single parameterized statement executed once per row through one round trip (or one round trip per batch, for very large row counts) - kept API-shaped like SqlBulkCopy (FirebirdCommandBatcher.DestinationTableName, ColumnMappings, WriteToServer/WriteToServerAsync).

Important Pages

• GitHub Home — core library and source code.
• Website — full documentation, API reference, and blog.

Core Features

• Async Methods
• BulkInsert
• BulkMerge
• BulkUpdate
• BulkDelete
• BulkDeleteByKey
• Column Mappings Behavior

Community

• GitHub Issues — bug reports and feature requests.
• Microsoft Teams — live Q&A.
• GitHub Discussions — ask questions and share ideas.
• X / Twitter — news and updates.

Known limitations

• Every bulk operation except a plain (no-return-identity) BulkInsert stages rows through a pseudo (staging) table first. Each pseudo table gets a short, per-call unique name, so - unlike some other providers' bulk-operations packages - concurrent callers targeting the same table never race on a shared staging-table name, for either pseudoTableType: Physical or Memory (a genuine Firebird GLOBAL TEMPORARY TABLE ... ON COMMIT PRESERVE ROWS).
• BulkMerge/BulkInsert with identityBehavior: ReturnIdentity read generated identities back via an EXECUTE BLOCK ... SUSPEND loop over the pseudo table - Firebird's RETURNING clause, like Oracle's, only ever returns a single row, so a multi-row result set needs this "loop and yield" shape. Row order is guaranteed by a client-assigned row-order column (not a server-generated one), so correctness never depends on rows being written back in input order.
• BulkMerge where the identity column is also a merge qualifier (the common default case) compiles to a per-row branching EXECUTE BLOCK, for the same reason documented for the core RepoDb.Firebird package's own Merge - see its README's "Known limitations" section.
• There is no bulkCopyOptions parameter (unlike SqlBulkCopy/DB2BulkCopy-based packages) - FbBatchCommand has no equivalent concept.

License

Apache-2.0 — Copyright © 2020 Michael Camara Pendon

Installation

Install-Package RepoDb.Firebird.BulkOperations

Then initialize the bootstrapper once at application startup:

GlobalConfiguration
    .Setup()
    .UseFirebird();

Async Methods

Every synchronous operation has a corresponding Async overload.

BulkInsert

Inserts a list of entities into the database in bulk. Returns the number of inserted rows.

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var insertedRows = connection.BulkInsert<Customer>(customers);
}

Or via table-name:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var insertedRows = connection.BulkInsert("Customer", customers);
}

Or via a DataTable:

using (var connection = new FbConnection(ConnectionString))
{
    var table = GetCustomersAsDataTable();
    var insertedRows = connection.BulkInsert("Customer", table);
}

Returning generated identities:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers(); // Id not set
    connection.BulkInsert<Customer>(customers, identityBehavior: FirebirdBulkImportIdentityBehavior.ReturnIdentity);
    // customers[i].Id now holds the generated identity for each row
}

BulkMerge

Upserts a list of entities in bulk — inserts new rows and updates existing ones based on the defined qualifiers. Returns the number of affected rows.

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var mergedRows = connection.BulkMerge<Customer>(customers);
}

Or with qualifiers:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var mergedRows = connection.BulkMerge<Customer>(customers, qualifiers: e => new { e.LastName, e.DateOfBirth });
}

Or via table-name with qualifiers:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var mergedRows = connection.BulkMerge("Customer", customers, qualifiers: Field.From("LastName", "DateOfBirth"));
}

Or via a DataTable:

using (var connection = new FbConnection(ConnectionString))
{
    var table = GetCustomersAsDataTable();
    var mergedRows = connection.BulkMerge("Customer", table);
}

BulkUpdate

Updates existing rows in the database in bulk, matched by the defined qualifiers. Returns the number of updated rows.

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var rows = connection.BulkUpdate<Customer>(customers);
}

Or with qualifiers:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var rows = connection.BulkUpdate<Customer>(customers, qualifiers: e => new { e.LastName, e.DateOfBirth });
}

Or via a DataTable:

using (var connection = new FbConnection(ConnectionString))
{
    var table = GetCustomersAsDataTable();
    var rows = connection.BulkUpdate("Customer", table);
}

BulkDelete

Deletes existing rows from the database in bulk, matched by the defined qualifiers. Returns the number of deleted rows.

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var deletedRows = connection.BulkDelete<Customer>(customers);
}

Or with qualifiers:

using (var connection = new FbConnection(ConnectionString))
{
    var customers = GetCustomers();
    var deletedRows = connection.BulkDelete<Customer>(customers, qualifiers: e => new { e.LastName, e.DateOfBirth });
}

Or via a DataTable:

using (var connection = new FbConnection(ConnectionString))
{
    var table = GetCustomersAsDataTable();
    var deletedRows = connection.BulkDelete("Customer", table);
}

BulkDeleteByKey

Deletes existing rows from the database in bulk, matched by their primary (or identity) key value alone — no entities or DataTable involved, just the list of key values to remove. Returns the number of deleted rows.

using (var connection = new FbConnection(ConnectionString))
{
    var primaryKeys = new [] { 10045, 10046, 10047 };
    var deletedRows = connection.BulkDeleteByKey("Customer", primaryKeys);
}

Column Mappings Behavior

When no mappings are passed to BulkInsert, BulkMerge, or BulkUpdate (or their async variants), and for BulkDelete (which has no mappings argument), the columns of the source (entities, DataTable, or DbDataReader) are aligned with the columns of the destination table based on the BulkColumnMappingsBehavior setting:

Automatic is the default and is the behavior of the earlier versions. FirebirdBulkOperationsDbSetting inherits FirebirdDbSetting, and UseFirebird(...) registers it as the setting of the connection in place of the default one. Strict and StrictBypass throw a FirebirdBulkColumnMappingsException (naming the offending columns) before any data is written.

GlobalConfiguration
    .Setup()
    .UseFirebird(new FirebirdBulkOperationsDbSetting
    {
        BulkColumnMappingsBehavior = FirebirdBulkColumnMappingsBehavior.StrictBypass
    });

For example, with a Customer table that has the Id, Name, Email, and CreatedDateUtc columns, and a list of entities (the source) with the Email, Name, and Nickname properties:

• Strict throws, because the order differs, Id/CreatedDateUtc are missing, and Nickname does not exist in the destination.
• StrictBypass throws, because Nickname does not exist in the destination.
• Automatic inserts Name and Email, and ignores Nickname.

try
{
    await connection.BulkInsertAsync("Customer", customers);
}
catch (FirebirdBulkColumnMappingsException ex)
{
    // e.g. "... Source column(s) that do not exist on the destination table: 'Nickname'."
    logger.LogError(ex, "Bulk insert rejected due to column misalignment.");
    throw;
}

Explicitly passed mappings always take precedence: the setting is not applied to them. BulkDeleteByKey is not affected, as its source is a list of key values rather than a set of columns.