C# extension method - OrderBy sort expression as string

Recently I came across a question in a forum where a part of the solution requires the sort expression as string type for the OrderBy extension method. The requirement is, upon passing a string of property name separated by comma, the respective sort need to be applied and returned.

Without explaining further, directly I dive into the implementation so that you can easily infer from that. I altered the snippet got from http://www.extensionmethod.net/Details.aspx?ID=124 based on our needs.

Snippet 1 (Core extension method for OrderBy and ThenBy)

public static class IEnumerableExtensions
{
    public static IOrderedEnumerable<T> OrderBy<T>(this IEnumerable<T> list, string sortExpression)
    {
        sortExpression += "";
        string[] parts = sortExpression.Split(' ');
        bool descending = false;
        string property = "";

        if (!(parts.Length > 0 && parts[0] != ""))
        {
            throw new Exception("Invalid sort expression.");
        }

        property = parts[0];

        if (parts.Length > 1)
        {
            descending = parts[1].ToLower().Contains("esc");
        }

        PropertyInfo prop = typeof(T).GetProperty(property);

        if (prop == null)
        {
            throw new Exception("No property '" + property + "' in + " + typeof(T).Name + "'");
        }

        if (descending)
            return list.OrderByDescending(x => prop.GetValue(x, null));
        else
            return list.OrderBy(x => prop.GetValue(x, null));
    }

    public static IOrderedEnumerable<T> ThenBy<T>(this IOrderedEnumerable<T> list, string sortExpression)
    {
        sortExpression += "";
        string[] parts = sortExpression.Split(' ');
        bool descending = false;
        string property = "";

        if (!(parts.Length > 0 && parts[0] != ""))
        {
            throw new Exception("Invalid sort expression.");
        }

        property = parts[0];

        if (parts.Length > 1)
        {
            descending = parts[1].ToLower().Contains("esc");
        }

        PropertyInfo prop = typeof(T).GetProperty(property);

        if (prop == null)
        {
            throw new Exception("No property '" + property + "' in + " + typeof(T).Name + "'");
        }

        if (descending)
            return list.ThenByDescending(x => prop.GetValue(x, null));
        else
            return list.ThenBy(x => prop.GetValue(x, null));
    }

    public static IEnumerable<T> OrderByStringExpression<T>(this IEnumerable<T> queryObj, string orderByProperties)
    {
        if (string.IsNullOrWhiteSpace(orderByProperties))
        {
            //throw new Exception("Invalid sort expression");
            return queryObj;
        }

        string[] orderByPropertyArray = orderByProperties.Split(new char[] { ',' }, StringSplitOptions.RemoveEmptyEntries);
        IOrderedEnumerable<T> orderedItem = null;

        if (orderByPropertyArray.Length > 0)
        {
            orderedItem = queryObj.OrderBy(orderByPropertyArray[0]);

            if (orderByPropertyArray.Length > 1)
            {
                for (int i = 1; i < orderByPropertyArray.Length; i++)
                {
                    orderedItem = orderedItem.ThenBy(orderByPropertyArray[i]);
                }

            }
        }

        return orderedItem;
    }
}

Snippet 2 (Example code)

var sortResult = listInst.OrderByStringExpression("Id,Name desc");

As you see, each property name should be separated by comma and if required you can specify the sort mode (asc/desc) followed by a blank space with the respective parameter.

Now with the above specified method, we can pass simple comma separated string as sort expression and get the related instance sorted. This will helpful in scenarios like implementing the Repository pattern as I did here.  Hope this helps.

References
http://www.extensionmethod.net/Details.aspx?ID=124

Parsing and Manipulating HTML strings - C#

Recently I came across a situation, where I need to retrieve an HTML string created and stored by an ASP application. Upon retrieving the html string, I need to parse the string and manipulate html string by removing some deprecated attributes and us some other equivalent alternate instead. For instance the size attribute in the font element need to be replaced with css font-size attribute.

Snippet 1:

<font size="1">......</font>

to 

<font style="font-size: 9px;">......</font>

One additional thing we need to note here is, since we don't have any direct equivalent of the unit and value for the size attribute, I simply changed it to the expected size for the invoking applications.

Fine, now comes core of the issue about how to manipulate the HTML string. Initially I thought of using regex. Suddenly, I remembered of the nice tool Html Agility Pack which I explored few months back. Its an HTML parser that allows us to select an HTML element and manipulate it very easily. It uses the XPath to navigate through the elements. If you don't know XPath, no problem. You can refer to examples over the net (like in msdn and w3schools ) and find the XPath that suits better your requirement in less than a minute.

Following is the code for the situation I mentioned above.
Snippet 2:

private string RestructureDeprecatedAttributes(string htmlText)
{
    HtmlDocument document = new HtmlDocument();
    HtmlNodeCollection fontNodeCollection;
    
    document.LoadHtml(htmlText);
    fontNodeCollection = document.DocumentNode.SelectNodes("//font[@size]");

    if (fontNodeCollection != null)
    {
        foreach (HtmlNode fontNode in fontNodeCollection)
        {
            HtmlAttribute sizeAttribute = fontNode.Attributes["size"];

            if (!string.IsNullOrWhiteSpace(sizeAttribute.Value) && int.Parse(sizeAttribute.Value) > 0)
            {
                if (sizeAttribute.Value.Equals("1"))
                {
                    fontNode.SetAttributeValue("style", "font-size: 9px;");
                }
                else
                {
                    fontNode.SetAttributeValue("style", "font-size: 11px;");
                }
                
                fontNode.Attributes["size"].Remove();
            }
        }
    }

    return document.DocumentNode.WriteTo();
}

The selectors are really powerful which allows us to select the nodes that perfectly matches our criteria. The Html Agility Pack provides several methods that supports the manipulation of HTML strings in a very easy manner. You can infer this from the above code.