65.9K
CodeProject is changing. Read more.
Home

Data Access Layer Auto Generation Library

starIconstarIconstarIconstarIcon
emptyStarIcon
starIcon

4.97/5 (35 votes)

Dec 16, 2008

CPOL

2 min read

viewsIcon

88989

downloadIcon

3983

The proposed solution to avoid coding of data access layer in an ASP.NET project.

Introduction

Data Access Layer (DAL) is an important part in almost all architectures. I encountered problems creating DAL for large databases — maintaining DAL for these databases was another nightmare. So I thought of a generic solution to this problem, which was to auto generate the DAL.

I made use of System.CodeDom which is a code generator to produce source code in a particular programming language.

Using the Code

In order to generate DAL in your ASP.NET project, follow the below steps:

  1. Add a reference of DalProvider.dll which is a DAL Auto-Generator Library.
  2. Add the following under compilation tag in the web.config:
            <system.web>
    		<compilation debug="true">
    			<buildProviders>
    				 <add extension=".dal"
                                             type="AnoopsDALGenerator.DalProvider"/>
                    	<buildProviders>
     		</compilation>
    	</system.web>

    This allows the compilation of any file with the ".dal" extension added to the project with our DAL auto generation library.

  3. Add an XML file to the App_Code folder in the below mentioned format with the extension ".dal"
    	<dal>
       		<connectionString>Data Source =XXXXX; Initial Catalog = XXXXX;
                          User Id =XXXXX; Password=XXXXX 
    		</connectionString>
       		<nameSpace>DAL1</nameSpace>
    	</dal>

    In this XML file under the "dal" root tag, we need to mention the connection string of the database as well as the namespace to access the auto generated DAL.

  4. Compile your ASP.NET project. This generates the DAL with the following features:
    • Namespace "DAL1" as specified in the XML file
    • All tables are represented as classes with its columns as member fields that can be accessed using property with name in the following pattern "Col" + Column Name. This enables the users to create objects of tables.
      DAL1.TableName table1 = new DAL1.TableName();
      tabel1.ColColumnName1 = "2";
      table2.ColColumnName2 = "Trial";
      Response.Write ("Data in column2:" +  table2.ColColumnName2);
    • All stored procedure access methods are defined under the DataProcedure class. Thus, to access a stored procedure, the user has to use the respective stored procedure access function ("Access"+StoredProcedureName).
              DAL2.DataProcedure dbObject = new DAL2.DataProcedure();
      	    DataTable dataResult=new DataTable();
                  int returnStatus = dbObject.AccessStoredProcedureName(ref dataResult,
                      "21", "Thomas", "Atlas", "ITDept");
                  if (returnStatus>

    A reference to a datatable has to be passed to this function to retrieve the records fetched using the stored procedure.

Code

The code is big but simple. The workflow of this library is as below:

  • Extract the connection string and the proposed namespace from the XML file during compilation.
  • Query the systemobjects table and systemcolumns table to list out the tables and columns respectively:
    --To extract the table information from database	            
      Select * from sysobjects where type = 'u'
    --To extract the column information with respect to the table 
      Select * from syscolumns
  • Generate classes from the table names and create coulmns as member variables under it.
  • Create a class named DataProcedure to hold the functions to access stored procedures.
  • Query the systemobjects table to list all the stored procedures:
    Select * from sysobjects where type = 'p' and category = 0
  • Query systemcolumns to get the parameters of the respective stored procedure.
    SELECT
    param.name AS [Name],
    ISNULL(baset.name, N'') AS [SystemType],
    CAST(CASE WHEN baset.name IN (N'char', N'varchar', N'binary', N'varbinary',
        N'nchar', N'nvarchar') 
    THEN param.prec ELSE param.length END AS int) AS [Length],
    CAST(param.xprec AS int) AS [NumericPrecision],
    CAST(param.xscale AS int) AS [NumericScale],
    null AS [DefaultValue],
    CAST(
        CASE param.isoutparam WHEN 1 THEN param.isoutparam WHEN 0 THEN 
           CASE param.name WHEN '' 
    THEN 1 ELSE 0 END END AS bit) AS [IsOutputParameter]
    FROM
    dbo.sysobjects AS sp
    INNER JOIN sysusers AS ssp ON ssp.uid = sp.uid
    INNER JOIN syscolumns AS param ON (param.number = 1) AND (param.id=sp.id)
    LEFT OUTER JOIN systypes AS baset ON baset.xusertype = 
        param.xtype and baset.xusertype = baset.xtype
    WHERE
     (sp.xtype = N'P' OR sp.xtype = N'RF')and(sp.id= ?  and ssp.name=N'dbo')
    ORDER BY
    CAST(param.colid AS int) ASC
  • Combine all the information and generate functions to access stored procedures.

Points of Interest

This solution is a good example of using System.CodeDom.

History

  • 17th December, 2008: Initial version