Reading CSV file some missing columns

Question

I am trying to read in a CSV file into my VB.net application using the following code:

While Not EOF(1)
    Input(1, dummy)
    Input(1, phone_number)
    Inp         


        

Answer 1

One thing that you should come to grips with is that those Filexxxx methods are all but officially and formally deprecated. When using them, Intellisense pops up with:

...The My feature gives you better productivity and performance in file I/O operations than FileOpen. For more information, see Microsoft.VisualBasic.FileIO.FileSystem.

They are talking about My.Computer.FileSystem but there are some even more useful NET methods.

The post doesnt reveal how the data will be stored, but if it is an array of any sort and/or a structure, those are at least suboptimal if not also outdated. This will store it in a class so that the numeric data can be stored as numbers and a List will be used in place of an array.

I made a quick file similar to yours with some random data: {"CustName", "Phone", "UserName", "Product", "Cost", "Price", "Profit", "SaleDate", "RefCode"}:

• The CustName is present 70% of the time
• The username is never present
• The RefCode is present 30% of the time
• I added a SaleDate to illustrate that data conversion

Ziggy Aurantium,132-5562,,Cat Food,8.26,9.95,1.69,08/04/2016,
Catrina Caison,899-8599,,Knife Sharpener,4.95,6.68,1.73,10/12/2016,X-873-W3
,784-4182,,Vapor Compressor,11.02,12.53,1.51,09/12/2016,

Code to Parse the CSV

Note: this is a bad way to parse a CSV. There are lots of problems that can arise doing it this way; plus it takes more code. It is presented because it is a simple way to not have to deal with the missing fields. See The Right Way

' form/class level var:
Private SalesItems As List(Of SaleItem)

SaleItem is a simple class to store the elements you care about. SalesItems is a collection which can store only SaleItem objects. The properties in that class allow Price and Cost to be stored as Decimal and the date as a DateTime.

' temp var
Dim item As SaleItem
' create the collection
SalesItems = New List(Of SaleItem)
    
' load the data....all of it
Dim data = File.ReadAllLines("C:\Temp\custdata.csv")

' parse data lines 
' Start at 1 to skip a Header
For n As Int32 = 0 To data.Length - 1
    Dim split = data(n).Split(","c)

    ' check if it is a good line
    If split.Length = 9 Then
        ' create a new item
        item = New SaleItem
        ' store SOME data to it
        item.CustName = split(0)
        item.Phone = split(1)
        ' dont care anout user name (2)
        item.Product = split(3)
        ' convert numbers
        item.Price = Convert.ToDecimal(split(4))
        item.Cost = Convert.ToDecimal(split(5))
        ' dont use the PROFIT, calculate it in the class (6)

        ' convert date
        item.SaleDate = Convert.ToDateTime(split(7))

        ' ignore nonexistant RefCode (8)

        ' add new item to collection
        ' a List sizes itself as needed!
        SalesItems.Add(item)
    Else
        ' To Do: make note of a bad line format
    End If
Next

' show in DGV for approval/debugging
dgvMem.DataSource = SalesItems

Result:

Notes
It is generally a bad idea to store something which can be simply calculated. So the Profit property is:

Public ReadOnly Property Profit As Decimal
    Get
        Return (Cost - Price)
    End Get
End Property

It can never be "stale" if the cost or price is updated.

As shown, using the resulting collection can be displayed to the user very easily. Given a DataSource, the DataGridView will create the columns and populate the rows.

The Right Way

String.Split(c) is a very bad idea because if the product is: "Hose, Small Green" it will chop that up and treat it as 2 fields. There are a number of tools which will do nearly all the work for you:

1. Read the file
2. Parse the lines
3. Map the CSV data to a class
4. convert the text into the proper data type
5. create an economical collecton

Aside from the class, all the above could be done in just a few lines using CSVHelper:

Private CustData As List(Of SaleItem)
...
Using sr As New StreamReader("C:\Temp\custdata.csv", False),
     csv = New CsvReader(sr)
    csv.Configuration.HasHeaderRecord = True

    CustData = csv.GetRecords(Of SaleItem)().ToList()
End Using

Two or three lines of code to read, parse, and create a collection of 250 items.

Even if you want to do it manually for some reason, CSVHelper can help. Rather than create a List(Of SaleItem) for you, you can use it to read and parse the data:

... like above
csv.Configuration.HasHeaderRecord = True
Do Until csv.Read() = False
    For n As Int32 = 0 To csv.Parser.FieldCount - 1
        DoSomethingWith(csv.GetField(n))
    Next
Loop

This will return the fields to you one by one. It wont convert any dates or prices, but it wont choke on missing data elements either.

Resources

• Five Minute Intro To Classes and Lists
• CSVHelper
• Class vs Structure
  • Eric Lippert's Thoughts on the matter
• The File Class has gobs of useful methods