VB.NET: Reading a CSV file into a 2D array - vb.net

Question

I'm working on a project that requires me to take values from a CSV file. I've to do further processing with these values and it'd be great if I can have these values in a 2D array. The number of rows and columns of the CSV files changes at regular intervals.
I'm unable to take these values into a 2D array in VB.NET/C#. Can I please have some help on that?
Here the code that I used:
Imports System.IO
Public Class Form1
Private Sub ReadCSVFileToArray()
Dim strfilename As String
Dim num_rows As Long
Dim num_cols As Long
Dim x As Integer
Dim y As Integer
Dim strarray(1, 1) As String
' Load the file.
strfilename = "test.csv"
'Check if file exist
If File.Exists(strfilename) Then
Dim tmpstream As StreamReader = File.OpenText(strfilename)
Dim strlines() As String
Dim strline() As String
strlines = tmpstream.ReadToEnd().Split(Environment.NewLine)
' Redimension the array.
num_rows = UBound(strlines)
strline = strlines(0).Split(",")
num_cols = UBound(strline)
ReDim strarray(num_rows, num_cols)
' Copy the data into the array.
For x = 0 To num_rows
strline = strlines(x).Split(",")
For y = 0 To num_cols
strarray(x, y) = strline(y)
Next
Next
' Display the data in textbox
For x = 0 To num_rows
For y = 0 To num_cols
TextBox1.Text = TextBox1.Text & strarray(x, y) & ","
Next
TextBox1.Text = TextBox1.Text & Environment.NewLine
Next
End If
End Sub
Private Sub Button1_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles Button1.Click
ReadCSVFileToArray()
End Sub
End Class

Answer 1

You use the CsvReader from here to easily and conveniently read a csv (or other similar data) into a DataTable (or an IDataReader). It is always my first choice for this scenario - fast and pretty robust.

Answer 2

I'm not disagreeing with Marc Gravell, because you shouldn't try to re-invent the wheel. His solution would perform the best, and there can be many CSV format nuances that can screw up a simple parser such as the one demonstrated below.
With that said, you asked for a CSV parser that returns a 2D array. The code below does just that (jagged array), and should work for a very simple CSV file. This skips a header row in the file.
(sorry it's not in VB, but you can put it in a helper class in your project)
private string[][] GetData(string fileName)
{
List<string[]> data = new List<string[]>();
using (StreamReader sr = new StreamReader(fileName))
{
bool headerRow = true;
string line;
while ((line = sr.ReadLine()) != null)
{
if (headerRow)
{
headerRow = false;
}
else
{
string[] split = line.Split(new char[] { ',' });
data.Add(split);
}
}
}
return data.ToArray();
}

Answer 3

Parsing a CSV file can be pretty hard due to all the possible variations of CSV. See: http://en.wikipedia.org/wiki/Comma-separated_values
E.g., think about embedded quotes, commas etc.
So it will not be a simple matter of reading lines/strings and splitting them.
Perhaps you are better of by using a third party library

Answer 4

Since the number of columns and rows change frequently, you could use dynamic lists rather than a fixed 2D array.
List<List<string>> data = new List<List<string>>();
Then as you parse the CSV file you can build up the data dynamically in the above structure, adding a new List<string> to data for each new row in the CSV file.
Update: The VB.NET equivalent is something like
data As New List(Of List(Of String))
That being said, #Mark Gravell's suggestion is a far better solution than doing this yourself.