I realize this is a newbie question, but I'm looking for a simple solution - it seems like there should be one.
What's the best way to import a CSV file into a strongly-typed data structure? Again simple = better.
I realize this is a newbie question, but I'm looking for a simple solution - it seems like there should be one.
What's the best way to import a CSV file into a strongly-typed data structure? Again simple = better.
There are two articles on CodeProject that provide code for a solution, one that uses StreamReader and one that imports CSV data using the Microsoft Text Driver.
A good simple way to do it is to open the file, and read each line into an array, linked list, data-structure-of-your-choice. Be careful about handling the first line though.
This may be over your head, but there seems to be a direct way to access them as well using a connection string.
Why not try using Python instead of C# or VB? It has a nice CSV module to import that does all the heavy lifting for you.
If you're expecting fairly complex scenarios for CSV parsing, don't even think up of rolling our own parser. There are a lot of excellent tools out there, like FileHelpers, or even ones from CodeProject.
The point is this is a fairly common problem and you could bet that a lot of software developers have already thought about and solved this problem.
Microsoft's TextFieldParser is stable and follows RFC 4180 for CSV files. Don't be put off by the
Microsoft.VisualBasic
namespace; it's a standard component in the .NET Framework, just add a reference to the globalMicrosoft.VisualBasic
assembly.If you're compiling for Windows (as opposed to Mono) and don't anticipate having to parse "broken" (non-RFC-compliant) CSV files, then this would be the obvious choice, as it's free, unrestricted, stable, and actively supported, most of which cannot be said for FileHelpers.
See also: How to: Read From Comma-Delimited Text Files in Visual Basic for a VB code example.
Use an OleDB connection.
I agree with @NotMyself. FileHelpers is well tested and handles all kinds of edge cases that you'll eventually have to deal with if you do it yourself. Take a look at what FileHelpers does and only write your own if you're absolutely sure that either (1) you will never need to handle the edge cases FileHelpers does, or (2) you love writing this kind of stuff and are going to be overjoyed when you have to parse stuff like this:
1,"Bill","Smith","Supervisor", "No Comment"
2 , 'Drake,' , 'O'Malley',"Janitor,
Oops, I'm not quoted and I'm on a new line!