如何定义项目（RMO 编程）
通过使用复制管理对象 (RMO)，能够以编程方式来定义项目。用来定义项目的 RMO 类取决于要为其定义项目的发布的类型。
为快照发布或事务发布定义项目
使用 ServerConnection 类创建与发布服务器的连接。
创建 TransArticle 类的实例，将 ConnectionContext 属性（继承自 ReplicationObject）设为步骤 1 中的 ServerConnection 实例，并且设置以下属性：
将 Name 设置为项目的名称。
将 PublicationName 设置为发布的名称。
将 DatabaseName 设置为发布数据库的名称。
（可选）如果项目有一个静态行筛选器，则将 FilterClause 属性指定为筛选子句。
调用 Create 方法。
（可选）如果项目有列筛选器：
调用 RemoveReplicatedColumns 方法以从项目中删除列。传递包含待删除列的名称的 String 数组。传递一个 null 值则表明删除所有列。
调用 AddReplicatedColumns 方法，向项目中添加列。传递包含待添加列的名称的 String 数组。传递一个 null 值表明添加所有列。
定义合并发布的项目
使用 ServerConnection 类创建与发布服务器的连接。
创建 MergeArticle 类的实例，将 ConnectionContext 属性（继承自 ReplicationObject）设置为步骤 1 中的 ServerConnection 实例，并设置以下属性：
将 Name 设置为项目的名称。
将 PublicationName 设置为发布的名称。
将 DatabaseName 设置为发布数据库的名称。
（可选）如果对项目进行水平筛选，则将 FilterClause 属性指定为行筛选器子句。使用此属性可指定静态筛选器或参数化筛选器。有关详细信息，请参阅参数化行筛选器。
调用 Create 方法。
（可选）如果项目有经过筛选的列：
调用 RemoveReplicatedColumns 方法以从项目中删除列。传递包含待删除列的名称的 String 数组。传递一个 null 值则表明删除所有列。
调用 AddReplicatedColumns 方法，向项目中添加列。传递包含待添加列的名称的 String 数组。传递一个 null 值表明添加所有列。
（可选）如果必须基于在另一个项目中定义的行筛选器来过滤项目，可创建 MergeJoinFilter 类的实例来定义两个项目之间的联接筛选器。然后，设置以下属性：
将 JoinArticleName 属性设置为要筛选的子项目（来自步骤 2）的名称。
将 ArticleName 属性设置为将与此项目进行联接的现有已筛选项目的名称。
将 FilterName 属性设置为联接筛选器的名称。
将 JoinFilterClause 属性设置为定义该联接的表达式。
将 FilterTypes 属性的值设置为 JoinFilter。
有关详细信息，请参阅联接筛选器。
（可选）对表示已筛选项目（来自步骤 2）的对象调用 AddMergeJoinFilter 方法。传递步骤 5 中的 MergeJoinFilter 对象以创建联接筛选器。
示例
下例向一个事务发布添加一个带有行和列筛选器的项目。
// Define the Publisher, publication, and article names.
string publisherName = publisherInstance;
string publicationName = "AdvWorksProductTran";
string publicationDbName = "AdventureWorks";
string articleName = "Product";
string schemaOwner = "Production";
TransArticle article;
// Create a connection to the Publisher.
ServerConnection conn = new ServerConnection(publisherName);
// Create a filtered transactional articles in the following steps:
// 1) Create the article with a horizontal filter clause.
// 2) Add columns to or remove columns from the article.
try
{
// Connect to the Publisher.
conn.Connect();
// Define a horizontally filtered, log-based table article.
article = new TransArticle();
article.ConnectionContext = conn;
article.Name = articleName;
article.DatabaseName = publicationDbName;
article.SourceObjectName = articleName;
article.SourceObjectOwner = schemaOwner;
article.PublicationName = publicationName;
article.Type = ArticleOptions.LogBased;
article.FilterClause = "DiscontinuedDate IS NULL";
// Ensure that we create the schema owner at the Subscriber.
article.SchemaOption |= CreationScriptOptions.Schema;
if (!article.IsExistingObject)
{
// Create the article.
article.Create();
}
else
{
throw new ApplicationException(String.Format(
"The article {0} already exists in publication {1}.",
articleName, publicationName));
}
// Create an array of column names to remove from the article.
String[] columns = new String[1];
columns[0] = "DaysToManufacture";
// Remove the column from the article.
article.RemoveReplicatedColumns(columns);
}
catch (Exception ex)
{
// Implement appropriate error handling here.
throw new ApplicationException("The article could not be created.", ex);
}
finally
{
conn.Disconnect();
}
' Define the Publisher, publication, and article names.
Dim publisherName As String = publisherInstance
Dim publicationName As String = "AdvWorksProductTran"
Dim publicationDbName As String = "AdventureWorks"
Dim articleName As String = "Product"
Dim schemaOwner As String = "Production"
Dim article As TransArticle
' Create a connection to the Publisher.
Dim conn As ServerConnection = New ServerConnection(publisherName)
' Create a filtered transactional articles in the following steps:
' 1) Create the article with a horizontal filter clause.
' 2) Add columns to or remove columns from the article.
Try
' Connect to the Publisher.
conn.Connect()
' Define a horizontally filtered, log-based table article.
article = New TransArticle()
article.ConnectionContext = conn
article.Name = articleName
article.DatabaseName = publicationDbName
article.SourceObjectName = articleName
article.SourceObjectOwner = schemaOwner
article.PublicationName = publicationName
article.Type = ArticleOptions.LogBased
article.FilterClause = "DiscontinuedDate IS NULL"
' Ensure that we create the schema owner at the Subscriber.
article.SchemaOption = article.SchemaOption Or _
CreationScriptOptions.Schema
If Not article.IsExistingObject Then
' Create the article.
article.Create()
Else
Throw New ApplicationException(String.Format( _
"The article {0} already exists in publication {1}.", _
articleName, publicationName))
End If
' Create an array of column names to remove from the article.
Dim columns() As String = New String(0) {}
columns(0) = "DaysToManufacture"
' Remove the column from the article.
article.RemoveReplicatedColumns(columns)
Catch ex As Exception
' Implement appropriate error handling here.
Throw New ApplicationException("The article could not be created.", ex)
Finally
conn.Disconnect()
End Try
下例将三个项目添加到一个合并发布。这些项目具有列筛选器，而且使用两个联接筛选器来将一个参数化行筛选器传播给其他项目。
// Define the Publisher and publication names.
string publisherName = publisherInstance;
string publicationName = "AdvWorksSalesOrdersMerge";
string publicationDbName = "AdventureWorks";
// Specify article names.
string articleName1 = "Employee";
string articleName2 = "SalesOrderHeader";
string articleName3 = "SalesOrderDetail";
// Specify join filter information.
string filterName12 = "SalesOrderHeader_Employee";
string filterClause12 = "Employee.EmployeeID = " +
"SalesOrderHeader.SalesPersonID";
string filterName23 = "SalesOrderDetail_SalesOrderHeader";
string filterClause23 = "SalesOrderHeader.SalesOrderID = " +
"SalesOrderDetail.SalesOrderID";
string salesSchema = "Sales";
string hrSchema = "HumanResources";
MergeArticle article1 = new MergeArticle();
MergeArticle article2 = new MergeArticle();
MergeArticle article3 = new MergeArticle();
MergeJoinFilter filter12 = new MergeJoinFilter();
MergeJoinFilter filter23 = new MergeJoinFilter();
// Create a connection to the Publisher.
ServerConnection conn = new ServerConnection(publisherName);
// Create three merge articles that are horizontally partitioned
// using a parameterized row filter on Employee.EmployeeID, which is
// extended to the two other articles using join filters.
try
{
// Connect to the Publisher.
conn.Connect();
// Create each article.
// For clarity, each article is defined separately.
// In practice, iterative structures and arrays should
// be used to efficiently create multiple articles.
// Set the required properties for the Employee article.
article1.ConnectionContext = conn;
article1.Name = articleName1;
article1.DatabaseName = publicationDbName;
article1.SourceObjectName = articleName1;
article1.SourceObjectOwner = hrSchema;
article1.PublicationName = publicationName;
article1.Type = ArticleOptions.TableBased;
// Define the parameterized filter clause based on Hostname.
article1.FilterClause = "Employee.LoginID = HOST_NAME()";
// Set the required properties for the SalesOrderHeader article.
article2.ConnectionContext = conn;
article2.Name = articleName2;
article2.DatabaseName = publicationDbName;
article2.SourceObjectName = articleName2;
article2.SourceObjectOwner = salesSchema;
article2.PublicationName = publicationName;
article2.Type = ArticleOptions.TableBased;
// Set the required properties for the SalesOrderDetail article.
article3.ConnectionContext = conn;
article3.Name = articleName3;
article3.DatabaseName = publicationDbName;
article3.SourceObjectName = articleName3;
article3.SourceObjectOwner = salesSchema;
article3.PublicationName = publicationName;
article3.Type = ArticleOptions.TableBased;
if (!article1.IsExistingObject) article1.Create();
if (!article2.IsExistingObject) article2.Create();
if (!article3.IsExistingObject) article3.Create();
// Select published columns for SalesOrderHeader.
// Create an array of column names to vertically filter out.
// In this example, only one column is removed.
String[] columns = new String[1];
columns[0] = "CreditCardApprovalCode";
// Remove the column.
article2.RemoveReplicatedColumns(columns);
// Define a merge filter clauses that filter
// SalesOrderHeader based on Employee and
// SalesOrderDetail based on SalesOrderHeader.
// Parent article.
filter12.JoinArticleName = articleName1;
// Child article.
filter12.ArticleName = articleName2;
filter12.FilterName = filterName12;
filter12.JoinUniqueKey = true;
filter12.FilterTypes = FilterTypes.JoinFilter;
filter12.JoinFilterClause = filterClause12;
// Add the join filter to the child article.
article2.AddMergeJoinFilter(filter12);
// Parent article.
filter23.JoinArticleName = articleName2;
// Child article.
filter23.ArticleName = articleName3;
filter23.FilterName = filterName23;
filter23.JoinUniqueKey = true;
filter23.FilterTypes = FilterTypes.JoinFilter;
filter23.JoinFilterClause = filterClause23;
// Add the join filter to the child article.
article3.AddMergeJoinFilter(filter23);
}
catch (Exception ex)
{
// Do error handling here and rollback the transaction.
throw new ApplicationException(
"The filtered articles could not be created", ex);
}
finally
{
conn.Disconnect();
}
' Define the Publisher and publication names.
Dim publisherName As String = publisherInstance
Dim publicationName As String = "AdvWorksSalesOrdersMerge"
Dim publicationDbName As String = "AdventureWorks"
' Specify article names.
Dim articleName1 As String = "Employee"
Dim articleName2 As String = "SalesOrderHeader"
Dim articleName3 As String = "SalesOrderDetail"
' Specify join filter information.
Dim filterName12 As String = "SalesOrderHeader_Employee"
Dim filterClause12 As String = "Employee.EmployeeID = " + _
"SalesOrderHeader.SalesPersonID"
Dim filterName23 As String = "SalesOrderDetail_SalesOrderHeader"
Dim filterClause23 As String = "SalesOrderHeader.SalesOrderID = " + _
"SalesOrderDetail.SalesOrderID"
Dim salesSchema As String = "Sales"
Dim hrSchema As String = "HumanResources"
Dim article1 As MergeArticle = New MergeArticle()
Dim article2 As MergeArticle = New MergeArticle()
Dim article3 As MergeArticle = New MergeArticle()
Dim filter12 As MergeJoinFilter = New MergeJoinFilter()
Dim filter23 As MergeJoinFilter = New MergeJoinFilter()
' Create a connection to the Publisher.
Dim conn As ServerConnection = New ServerConnection(publisherName)
' Create three merge articles that are horizontally partitioned
' using a parameterized row filter on Employee.EmployeeID, which is
' extended to the two other articles using join filters.
Try
' Connect to the Publisher.
conn.Connect()
' Create each article.
' For clarity, each article is defined separately.
' In practice, iterative structures and arrays should
' be used to efficiently create multiple articles.
' Set the required properties for the Employee article.
article1.ConnectionContext = conn
article1.Name = articleName1
article1.DatabaseName = publicationDbName
article1.SourceObjectName = articleName1
article1.SourceObjectOwner = hrSchema
article1.PublicationName = publicationName
article1.Type = ArticleOptions.TableBased
' Define the parameterized filter clause based on Hostname.
article1.FilterClause = "Employee.LoginID = HOST_NAME()"
' Set the required properties for the SalesOrderHeader article.
article2.ConnectionContext = conn
article2.Name = articleName2
article2.DatabaseName = publicationDbName
article2.SourceObjectName = articleName2
article2.SourceObjectOwner = salesSchema
article2.PublicationName = publicationName
article2.Type = ArticleOptions.TableBased
' Set the required properties for the SalesOrderDetail article.
article3.ConnectionContext = conn
article3.Name = articleName3
article3.DatabaseName = publicationDbName
article3.SourceObjectName = articleName3
article3.SourceObjectOwner = salesSchema
article3.PublicationName = publicationName
article3.Type = ArticleOptions.TableBased
' Create the articles, if they do not already exist.
If article1.IsExistingObject = False Then
article1.Create()
End If
If article2.IsExistingObject = False Then
article2.Create()
End If
If article3.IsExistingObject = False Then
article3.Create()
End If
' Select published columns for SalesOrderHeader.
' Create an array of column names to vertically filter out.
' In this example, only one column is removed.
Dim columns() As String = New String(0) {}
columns(0) = "CreditCardApprovalCode"
' Remove the column.
article2.RemoveReplicatedColumns(columns)
' Define a merge filter clauses that filter
' SalesOrderHeader based on Employee and
' SalesOrderDetail based on SalesOrderHeader.
' Parent article.
filter12.JoinArticleName = articleName1
' Child article.
filter12.ArticleName = articleName2
filter12.FilterName = filterName12
filter12.JoinUniqueKey = True
filter12.FilterTypes = FilterTypes.JoinFilter
filter12.JoinFilterClause = filterClause12
' Add the join filter to the child article.
article2.AddMergeJoinFilter(filter12)
' Parent article.
filter23.JoinArticleName = articleName2
' Child article.
filter23.ArticleName = articleName3
filter23.FilterName = filterName23
filter23.JoinUniqueKey = True
filter23.FilterTypes = FilterTypes.JoinFilter
filter23.JoinFilterClause = filterClause23
' Add the join filter to the child article.
article3.AddMergeJoinFilter(filter23)
Catch ex As Exception
' Do error handling here and rollback the transaction.
Throw New ApplicationException( _
"The filtered articles could not be created", ex)
Finally
conn.Disconnect()
End Try