Click here to Skip to main content
Click here to Skip to main content
Search within:



 
Comments & Discussions (60)

Very Lightweight Data Access Layer in C# and .NET 2.0

By , 4 Apr 2007
 

Introduction

In my spare time I write a lot of small applications for my friends and family. The common thing for all these small applications is that they grab some data from the database, display the data to the user and save the changes back to database.

When I design programs, I break them into tiers. Most of the time I have three logical tiers, which are User Interface Layer, Business Logic and Data Access Layer. I usually start with the class diagram, and only after I am happy with the class diagram then I build the tables and stored procedures to access the data from/to tables. User Interface is usually dictated by the user.

As I was building more and more of these small apps, I realized that I was spending a lot of time on my Data Access Layer, since I had to create mappings for populating the business objects from the tables in database, and since each application had a different business logic and business objects, I ended up writing Data Access Layer from scratch. So to make my life easier, I decided to build a generic data access helper class that could be reused on all my projects with little or no changes.

To be able to populate the object with the data from the database, the business object needs to have public properties with GET and SET methods. Then using reflection I could query the object for the public properties and if the property name matched the name of the field in the table, then the object would be populated with the data from that field.

There were times when the properties and the fields in the database were different or that the object could have more properties than the fields in database, so I decided to have two ways of populating the object with the data from the database:

  • Use a mapping class that would provide information of what property is mapped to what field, and
  • Decorate the properties with a custom attribute to show to what field the property is mapped.

The first step is to build a mapping class. This is a very simple class that holds a collection of a strings. The first string would be in the format property=field, where property is the name of the property of the object, and the field is the name of the field in database. So the mapping class, after implementation, looks like this:

/// <summary>
/// This class holds information about mapping a database field to a 
/// object property.
/// </summary>
public class Mapper
{
    private List<string> mMappingInfo;

    /// <summary>
    /// Default constructor
    /// </summary>
    public Mapper()
    {
        mMappingInfo = new List<string>();
    }


    /// <summary>
    /// Add mapping information. This method can be used to add more than 
    /// one mapping at a time.
    /// You could use it like: mapper.Add("property1=field1", 
    ///     "property2=field2", "property3=field3", ...)
    /// </summary>
    /// <param name="mappings">mapping information in format 
    //     "[property name]=[field name]"</param>
    public void Add(params string[] mappings)
    {
        foreach (string map in mappings)
            mMappingInfo.Add(map);
    }


    /// <summary>
    /// Return mapping information held in this class as string array
    /// </summary>
    public string[] MappingInformation
    {
        get
        {
            string[] mappings = new string[mMappingInfo.Count];
            mMappingInfo.CopyTo(mappings);

            return mappings;
        }
    }


    /// <summary>
    /// Indexer property. By providing the name it returns the mapping info
    /// for that property.
    /// If the mapping information for the provided property does not exist,
    /// the indexer 
    /// return null. 
    /// You could use it like: string mapInfo = mapper["property1"];
    /// </summary>
    /// <param name="propertyName">the name of the property to 

return 
    /// mapping information</param>
    /// <returns>mapping information for the property 

provided</returns>
    public string this[string propertyName]
    {
        get
        {
            foreach (string map in mMappingInfo)
            {
                string[] spplitedString = map.Split('=');
                if (spplitedString[0] == propertyName)
                    return map;
            }

            return null;
        }
    }


    /// <summary>
    /// Another indexer property. This property returns mapping information,
    /// that is stored in the list, in order.
    /// </summary>
    /// <param name="index">the index</param>
    /// <returns>mapping information</returns>
    public string this[int index]
    {
        get
        {
            if (index < mMappingInfo.Count)
                return mMappingInfo[index];
            else
                return null;
        }
    }

    /// <summary>
    /// Get the property name from the mapping information
    /// </summary>
    /// <param name="map">mapping information</param>
    /// <returns>the name of the property from the provided mapping 
    /// information</returns>
    public static string GetProperty(string map)
    {
        // split the mapping info and return the name of the property
        string[] spplitedString = map.Split('=');
        return spplitedString[0];
    }


    /// <summary>
    /// Get the field name from the mapping information
    /// </summary>
    /// <param name="map">mapping information</param>
    /// <returns>the name of the field from the provided mapping 
    /// information</returns>
    public static string GetField(string map)
    {
        // split the mapping info and return the name of the field
        string[] spplitedString = map.Split('=');
        return spplitedString[1];
    }
}

The next thing to implement is the custom attribute that will be used to map properties with the fields in database. Again this is a very simple attribute that will be used to store the name of the field. The implementation looks like:

/// <summary>
/// Specifies the name of the field in the table that the property maps to 
/// </summary>
[AttributeUsage(AttributeTargets.Property)]
    public class DBFieldAttribute : Attribute
    {
        private string mFieldName;

        /// <summary>
        /// constructor
        /// </summary>
        /// <param name="fieldName">name of the field that the 

property will
        /// be mapped to</param>
        public DBFieldAttribute(string fieldName)
        {
            mFieldName = fieldName;
        }

        public string FieldName
        {
            get { return mFieldName; }
        }
    }

Since this attribute can only be used on properties, the class is decorated with AttributeUsage(AttributeTargets.Property) attribute.

And now finally the only thing left is the helper class that will be used to retrieve data from the database. The implemented class looks like:

public class DBHelper
{
    /// <summary>
    /// Generic method. Gets an object of type T from the data reader. It 
    /// uses mapping information provided to read a field from the reader, 
    /// and gets the property name and sets the value of the property with 
    /// the data which are held in database field
    /// </summary>
    /// <typeparam name="T>The type of object to be 

instantiated</typeparam>
    /// <param name="rdr">Data Reader where the data will be 

read from</param>
    /// <param name="mappings">mapping 

information</param>
    /// <returns>an instance of type T with the properties populated from 
    /// database</returns>
private static T GetItemFromReader<T>(IDataReader rdr, Mapper mappings) 
    where T : class
{
    Type type = typeof(T); 
    T item = Activator.CreateInstance<T>(); // create an instance of the 
                                            // type provided
    foreach(string map in mappings.MappingInformation)
    {
        // for each mapping information 
        string property = Mapper.GetProperty(map);
        string field = Mapper.GetField(map);

        PropertyInfo propInfo = type.GetProperty(property);//get the property
                                                         //by name

        if (Convert.IsDBNull(rdr[field])) // data in database is null, so do
                                          // not set the value of the property
            continue;

        if (propInfo.PropertyType == rdr[field].GetType()) 
        // if the property and database field are the same
            propInfo.SetValue(item, rdr[field], null); // set the value of 
                                                       // property
        else
        {
            // change the type of the data in table to that of property and 
            // set the value of the property
            propInfo.SetValue(item, Convert.ChangeType(rdr[field], 
                propInfo.PropertyType), null); 
        }
    }
    return item;
}

/// <summary>
/// Generic method. Gets an object of type T from the data reader. It uses 
/// attribute information 
/// applied to a property to read a field from the reader, and gets the 
/// property name and sets
/// the value of the property with the data which are held in database field
/// </summary>
/// <typeparam name="T">The type of object to be 

instantiated</typeparam>
/// <param name="rdr">Data Reader where the data will be read 

from</param>
/// <returns>an instance of type T with the properties populated from 
/// database</returns>
private static T GetItemFromReader<T>(IDataReader rdr) where T : class
{
    Type type = typeof(T);
    T item = Activator.CreateInstance<T>();
    PropertyInfo[] properties = type.GetProperties();

    foreach (PropertyInfo property in properties)
    {
        // for each property declared in the type provided check if the 
        // property is decorated with the DBField attribute
        if (Attribute.IsDefined(property, typeof(DBFieldAttribute)))
        {
            DBFieldAttribute attrib = (DBFieldAttribute)Attribute.
                GetCustomAttribute(property, typeof(DBFieldAttribute));

            if (Convert.IsDBNull(rdr[attrib.FieldName])) 
            // data in database is null, so do not set the value of the 
            // property
                continue;

            if (property.PropertyType == rdr[attrib.FieldName].GetType()) 
            // if the property and database field are the same
                property.SetValue(item, rdr[attrib.FieldName], null); 
                // set the value of property
            else
            {
                // change the type of the data in table to that of property 
                // and set the value of the property
                property.SetValue(item, Convert.ChangeType(
                    rdr[attrib.FieldName], property.PropertyType), null);
            }
        }
    }

    return item;
}

/// <summary>
/// Get one object from the database by using the attribute information
/// </summary>
/// <typeparam name="T">the type of object the collection will 

hold</typeparam>
/// <param name="command">DbCommand that is used to read data 

from the 
///  database</param>
/// <returns>populated object from the database</returns>
public static T ReadObject<T>(IDbCommand command) where T : class
{
    IDataReader reader = command.ExecuteReader();
    if (reader.Read())
        return GetItemFromReader<T>(reader);
    else
        return default(T);
}

/// <summary>
/// Get one object from the database by using the mapping information 
/// provided by Mapper class
/// </summary>
/// <typeparam name="T">the type of object the collection will 

hold</typeparam>
/// <param name="command">DbCommand that is used to read data 

from the 
/// database</param>
/// <returns>populated object from the database</returns>
public static T ReadObject<T>(IDbCommand command, Mapper mappingInfo) 
    where T : class
{
    IDataReader reader = command.ExecuteReader();
    if (reader.Read())
        return GetItemFromReader<T>(reader, mappingInfo);
    else
        return default(T);
}

/// <summary>
/// Get a collection of objects from the database by using the attribute 
/// information
/// </summary>
/// <typeparam name="T>the type of object the collection will 

hold</typeparam>
/// <param name="command">DbCommand that is used to read data 

from the 
/// database</param>
/// <returns>a collection of populated objects from the 

database</returns>
public static List<T> ReadCollection<T>(IDbCommand command) where T 

: class
{
    List<T> collection = new List<T>();
    IDataReader reader = command.ExecuteReader();
    while (reader.Read())
    {
        T item = GetItemFromReader<T>(reader);
        collection.Add(item);
    }

    return collection;
}

/// <summary>
/// Get a collection of objects from the database by using the mapping 
/// information provided 
/// by Mapper class
/// </summary>
/// <typeparam name="T">the type of object the collection will 

hold</typeparam>
/// <param name="command">DbCommand that is used to read data 

from the 
/// database</param>
/// <returns>a collection of populated objects from the 

database</returns>
public static List<T> ReadCollection<T>(IDbCommand command, 
    Mapper mappingInfo) where T : class
{
    List<T> collection = new List<T>();
    IDataReader reader = command.ExecuteReader();
    while (reader.Read())
    {
        T item = GetItemFromReader<T>(reader, mappingInfo);
        collection.Add(item);
    }

    return collection;
}
}

DBHelper is a static class which means that it does not need to be instantiated and all the methods are static as well. It is generic in a sense that you provide the type of the object to be loaded from the database, and the class creates the objects and populates its public properties (the ones that either are decorated with the DBField attribute or mapping information is provided) from the data read from the table.

As you can see from the code above we are using constraints on generics, which basically means that the generic type T has to be an object (reference type) and can't be of value type (primitives like int, float, byte, and so on, or struct which is also of value type). The above class also uses reflection to check if the properties have the DBField set, and if yes then the code reads the attribute, gets the field name from the attribute and reads the data from the table.

Using the code

To use the class above is very easy, as you can see from the following example. Suppose we have a table that contains persons' details and its definition is like the below:

and a class Person that looks like the code below:

public class Person
{
    private int mID;
    private string mName;
    private string mSurname;
    private DateTime mDob;
    private string mProfession;

    public Person()
    {
    }


    [DBField("ID")]
    public int ID
    {
        get { return mID; }
        set { mID = value; }
    }

    [DBField("Name")]
    public string Name
    {
        get { return mName; }
        set { mName = value; }
    }

    [DBField("Surname")]
    public string Surname
    {
        get { return mSurname; }
        set { mSurname = value; }
    }

    [DBField("DOB")]
    public DateTime DateOfBirth
    {
        get { return mDob; }
        set { mDob = value; }
    }

    [DBField("Profession")]
    public string Profession
    {
        get { return mProfession; }
        set { mProfession = value; }
    }


    public int CalculateAge()
    {
        int age = DateTime.Now.Year - mDob.Year;
        return age;
    }
}

As you can see from the code of class Person, public properties are decorated with the attribute DBField. Each attribute corresponds with the field name in the table. Now to read the data from the table and return populated objects we would use the DBHelper class like this:

public List<Person> GetAllPersonsFromDB()
{
    using (SqlConnection connection = new SqlConnection(connectionString))
    {
        SqlCommand command = new SqlCommand("Select * from Persons order 

by 
            _Surname", connection);
        connection.Open();
        List<Person> persons = 

DBHelper.ReadCollection<Person>(command);
        return persons;
    }
}

or, if you need to retrieve a particular object, you could use the code below:

public Person GetPersonByID(int id)
{
    using (SqlConnection connection = new SqlConnection(connectionString))
    {
        SqlCommand command = new SqlCommand("Select * from Persons where 

ID 
            _= @ID", connection);
        SqlParameter param = command.Parameters.Add("@ID", 

SqlDbType.Int);
        param.Value = id;

        connection.Open();
        Person person = DBHelper.ReadObject<Person>(command);
        return person;
    }
}

If for some reason you need to get data from another table where the fields are named differently from the attributes, we need to provide mapping information by using Mapper class to the DBHelper.

Example: If we need to read data from another table and the table definition is as follows:

To read the data from the above table, we need to provide mapping information to DBHelper as follows:

public List<Person> GetAllPersonsFromDB()
{
    using (SqlConnection connection = new SqlConnection(connectionString))
    {
        SqlCommand command = new SqlCommand("Select * from Persons order 

by 
            _Surname", connection);
        Mapper mapper = new Mapper();
        
        // Provide the mapping information in format: "[Property Name]=
        // [Field Name]" for the appropriate fields 
        mapper.Add("ID=ID", "Name=FirstName", 

"Surname=Surname", 
            _"Profession=Profession", 

"DateOfBirth=DateOfBirth");

        connection.Open();
        List<Person> persons = 

DBHelper.ReadCollection<Person>(command, 
            _mapper);
        return persons;
    }
}

As you can see from the examples above, the DBHelper class can be used easily on different projects to get the data from the database and convert them to objects in a convenient way. So the first step is to declare the appropriate class with default constructor (parameterless constructor) and provide public properties with get/set methods, then either use DBField attribute on properties to map them to database fields, or use Mapper class to provide mapping information.

This class could be extended to support storing object to the database as well.

History

I have changed the source code, so now the DBHelper class can be used to insert/update business objects to database. I have also created a small demo program (attached to source solution).

License

This article, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

About the Author

 Fitim Skenderi
Software Developer (Senior) KOS-LIGHTS.COM
United Kingdom United Kingdom
Member
No Biography provided

Article Top
 
Sign Up to vote   Poor Excellent
Add a reason or comment to your vote: x
Votes of 3 or less require a comment

Comments and Discussions

 
You must Sign In to use this message board.
Search this forum  
    Spacing  Noise  Layout  Per page   
View All ThreadsFirst Prev Next
GeneralPerformancememberITEK13 Mar '07 - 4:20 
What type of performance gain do you find using this approach? I can see the value from a coding approach but in terms of data access speed. Do you see the data downloading faster.
 
Thanks,
 
Eric
 
Eric
 
Sign In·Permalink
GeneralRe: Performance PinmemberFitim Skenderi13 Mar '07 - 5:28 
I don't think you gain anything in speed when accessing data (It should be equal or slightly faster than using DataSets). The biggest gain as you say is from a coding approach, since it reduces quite a bit the amount of code to be written, hence less errors and more productive.
Sign In·Permalink
GeneralRe: Performance PinmemberCraig G. Wilson5 Apr '07 - 2:50 
You will probably lose some speed because you are using reflection. I have a similiar library that uses XML files to map result sets to objects. The speed boost will come from using a DynamicMethod and actually "compiling" the reflective calls.
Sign In·Permalink
GeneralRe: Performance Pinmembertechnicaltitch7 Oct '08 - 6:52 
Great code thanks! Does anyone know what the performance hit will be compared to using the readers explicitly without reflection? A random browse on the internet reveals mixed opinions on this.. If the performance hit is small, why is this entitled "*Very lightweight* DAL.."?
 
Thanks for any thoughts
Chris
Sign In·Permalink
Last Visit: 31 Dec '99 - 18:00     Last Update: 20 May '13 - 12:38Refresh1

General General    News News    Suggestion Suggestion    Question Question    Bug Bug    Answer Answer    Joke Joke    Rant Rant    Admin Admin   

About Article
Very Lightweight Data Access Layer in C# and .NET 2.0
Type Article
Licence CPOL
First Posted 9 Mar 2007
Views 126,216
Downloads 1,921
Bookmarked 124 times

, +
Top News

Math Primers for Programmers

Get the Insider News free each morning.
Related Videos
C++ Programming - 1004 vector Data Access
JavaScript: Using Objects Part 3
N-Tier Architecture and Tips
A N-Tier Architecture Sample with ASP.NET MVC3, WCF, and Entity Framework
Back-propagation Neural Net
A Lightweight Indexed File for Storing Persistent Data and Objects
ADO.NET for the Object-Oriented Programmer – Part Two
C# ASP.NET MySQL Connection Tutorial with MySQL Connector
From Mocks to Isolation
How design patterns can help you in developing unit testing-enabled applications
Raw Ethernet Packet Sending
Building a Multi-Tier Web Application in the .NET 3.5 Framework Using LINQ to SQL
EHSDataCompliance Framework
In memory data access, a methodology to simplify agile development
Tier Pressure and Isolationism
Access the Data Rows Filtered by the BindingSource.Filter Property
Enterprise Application Architecture with LINQ to SQL
Three-tier .NET Application Utilizing Three ORM Technologies
SoapNULL Transport for WSE 3.0
Facebook Integration via Hammock
Provider Design Patterns in NET 2.0
A Task Management System using Three Layer Architecture
Introduction to Object-Oriented Tiered Application Design
Permalink | Advertise | Privacy | Mobile
Web03 | 2.6.130516.1 | Last Updated 4 Apr 2007
Article Copyright 2007 by Fitim Skenderi
Everything else Copyright © CodeProject, 1999-2013
Terms of Use
Layout: fixed | fluid