.Net Entity Framework to CSV

Question

I'm using the latest Entity Framework with DBContext. I have a result set that I want to convert to comma separated values. I've done something similar with DataTables in VB DataTable to CSV extraction. I've got the QuoteName method working. I've also get a derivative of the GetCSV method working using a foreach. The problem is that it is a lot slower than the comparable code for a DataTable. So I'm hoping someone will have some suggestions.

    public static string GetCSV(this IQueryable entity)
    {
        if (entity == null)
        {
            throw new ArgumentNullException("entity");
        }
        Type T = entity.ElementType;
        var props = T.GetProperties(BindingFlags.Public | BindingFlags.Instance);
        string s = string.Empty;
        int iCols = props.Count();

        try
        {
            s += string.Join(",", (from int ii in Enumerable.Range(0, iCols)
                                   select props[ii].Name.QuoteName("[]")).ToArray());
            s += Environment.NewLine;
            foreach (var dr in entity)
            {
                s += string.Join(",", (from int ii in Enumerable.Range(0, iCols)
                                       select
                                           props[ii].GetValue(dr)
                                                    .ToString()
                                                    .QuoteName("\"\"", ",")).ToArray());
                s += Environment.NewLine;
            }
            s = s.TrimEnd(new char[] { (char)0x0A, (char)0x0D });
        }
        catch (Exception)
        {

            throw;
        }
        return s;
    }

Answer 1

Don't use a string to create your file. Try using the StringBuilder class (http://msdn.microsoft.com/en-us/library/system.text.stringbuilder.aspx)

Strings are immutable objects - that is, once a string is created, it cannot be changed. Every time you change a string (such as concatenate it), you're actually creating a brand new string. Using a string is very inefficient here.

Instead, create a stringbuilder object:

StringBuilder builder = new StringBuilder();

builder.Append("my data");

At the end, just call

builder.ToString();

Answer 2

Check out nuget package CsvHelper (http://nuget.org/packages/CsvHelper/).

Here is an example of the usage from https://github.com/JoshClose/CsvHelper/wiki/Basics

using (var csv = new CsvWriter( new StreamWriter( "Actors.csv" ) )) 
{
    csv.WriteRecords( actorsList );
}

Answer 3

I got some help from my little brother on this. He said to use a StringBuilder as well.

This is the code answer:

    /// <summary>
    /// Quotes a string using the following rules:
    /// <list>
    /// <listheader>Rules</listheader>
    /// <item>if the string is not quoted and the string contains the separator string</item>
    /// <item>if the string is not quoted and the string begins or ends with a space</item>
    /// <item>if the string is not quoted and the string contains CrLf</item>
    /// </list>
    /// </summary>
    /// <param name="s">String to be quoted</param>
    /// <param name="quote">
    /// <list>
    /// <listheader>quote characters</listheader>
    /// <item>if len = 0 then double quotes assumed</item>
    /// <item>if len = 1 then quote string is doubled for left and right quote characters</item>
    /// <item>else first character is left quote, second character is right quote</item>
    /// </list>
    /// </param>
    /// <param name="sep">separator string to check against</param>
    /// <returns></returns>
    /// <remarks></remarks>
    public static string QuoteName(this string s, string quote = null, string sep = ",")
    {
        quote = quote == null ? "" : quote;
        switch (quote.Length)
        {
            case 0:
                quote = "\"\"";
                break;
            case 1:
                quote += quote;
                break;
        }
        // Fields with embedded sep are quoted
        if ((!s.StartsWith(quote.Substring(0, 1))) && (!s.EndsWith(quote.Substring(1, 1))))
            if (s.Contains(sep))
                s = quote.Substring(0, 1) + s + quote.Substring(1, 1);
        // Fields with leading or trailing blanks are quoted
        if ((!s.StartsWith(quote.Substring(0, 1))) && (!s.EndsWith(quote.Substring(1, 1))))
            if (s.StartsWith(" ") || s.EndsWith(" "))
                s = quote.Substring(0, 1) + s + quote.Substring(1, 1);
        // Fields with embedded CrLF are quoted
        if ((!s.StartsWith(quote.Substring(0, 1))) && (!s.EndsWith(quote.Substring(1, 1))))
            if (s.Contains(System.Environment.NewLine))
                s = quote.Substring(0, 1) + s + quote.Substring(1, 1);
        return s;
    }

    public static string GetCSV(this IQueryable entity)
    {
        if (entity == null)
        {
            throw new ArgumentNullException("entity");
        }
        Type T = entity.ElementType;
        var props = T.GetProperties(BindingFlags.Public | BindingFlags.Instance);
        var sb = new StringBuilder();
        int iCols = props.Count();

        try
        {
            sb.Append(string.Join(",", Enumerable.Range(0, iCols).Cast<int>().
                Select(ii => props[ii].Name.QuoteName("[]")).ToArray()));

            foreach (var dr in entity)
            {
                sb.AppendLine();
                sb.Append(string.Join(",", Enumerable.Range(0, iCols).Cast<int>().
                    Select(ii => props[ii].GetValue(dr).
                        ToString().QuoteName("\"\"", ",")).ToArray()));
            }
        }
        catch (Exception ex)
        {

            throw;
        }
        return sb.ToString();
    }
}

I hope this is helpful to someone else.