Develop Reference

Basic Read CSV File Questions

Thanks in advance, C# newb here having a few issues.
I this CSV file provided daily, large, and has no header. I only need certain items out of this file.
Here is the code I have so far.
var config = new CsvConfiguration(CultureInfo.InvariantCulture)
{
HasHeaderRecord = false,
};
using (var reader = new StreamReader(iFile.FileName))
using (var csv = new CsvReader(reader, config))
{
var records = new List<BQFile>();
csv.Read();
csv.ReadHeader();
while (csv.Read())
{
var record = new BQFile()
{
SNumber = csv.GetField<string>("SNumber"),
FOBPoint = csv.GetField<string>("FOBPoint")
};
}
What I am not understanding since this CSV files 150+ fields, is how do grab the correct data. For example, if SNumber is column 46, FOBPoint is column 123. I am finding the CSVHelper documentation a little limited to me.
Any help is appreciated.

What I am not understanding since this CSV files 150+ fields, is how do grab the correct data
By index, because there is no header
In your BQFile, decorate the properties with an attribute of [Index(NNN)] where N is the column number (0-based). The IndexAttribute is found in CsvHelper.Configuration.Attributes namespace - I mention this because Entity Framework also has an Index attribute; be sure you use the correct one
pubic class BQFile{
[Index(46)]
public string SNumber { get; set;}
...
}
Then do:
var config = new CsvConfiguration(CultureInfo.InvariantCulture)
{
HasHeaderRecord = false,
};
using (var reader = new StreamReader(iFile.FileName))
using (var csv = new CsvReader(reader, config))
{
var records = csv.GetRecords<BQFile>();
...
records is an enumeration on top of the file stream (via CSVHelper, which reads records as it goes and creates instances of BQFile). You can only enumerate it once, and then after you're done enumerating it the filestream will be at the end - if you wanted to re-read the file you'd have to Seek the stream or renew the reader. Also, the file is only read (in chunks, progressively) as you enumerate. If you return records somewhere, so you drop out of the using and you thus dispose the reader, you'll get an error when you try to start reading from records (because it's disposed)
To work with records, you either foreach it, processing the objects you get as you go:
foreach(BQFile bqf in records){
//do stuff with each BQFile here
}
Or if you want to load it all into memory, you can do something like ToList() it so you end up with a bunch of BQFile in a List, and then you can e.g. access them randomly, read them over and over etc..
var bqfs = records.ToList();
ps; I don't know, when you said "it's column 46" if that's counting from 1 or 0.. You might have to adjust your 46.

C# - Splitting by columns on a CSV file

I want to import some data from a csv file, but I've encountered a small problem I can't really figure out.
The person who gave me this file, added comma seperated values in cells, so when I split them they will be added to the list. Instead, I would like to get all values per column as a string, I just can't really figure out how.
For example, the column I'm talking about, is about the days a restaurant is open. This can be Mo, Tu, We, Su, but it can also be Mo, Tu.
Is there a way I can just loop over de values per column, instead of by the comma seperated values?
I'm currently using it like this, but this just adds each day to the total list of values:
using (var fs = File.OpenRead(csvUrl))
using (var reader = new StreamReader(fs, Encoding.UTF8))
{
while (!reader.EndOfStream)
{
var line = reader.ReadLine();
if (i > 0)
{
var values = line.Split(',');
}
}
}

Use TextFieldParser to parse CSV files:
TextFieldParser parser = new TextFieldParser(new StringReader(lineContent));
parser.SetDelimiters(",");
string[] rawFields = parser.ReadFields();
lineContent is a string with the content of the current line in your file.
TextFieldParser is available in the namespace:
Microsoft.VisualBasic.FileIO
Don't mind abaout the Visual Basic part it works fine in C#
EDIT
In your code you could implement it like this:
using (var fs = File.OpenRead(csvUrl))
using (var reader = new StreamReader(fs, Encoding.UTF8))
{
while (!reader.EndOfStream)
{
var line = reader.ReadLine();
if (i > 0)
{
TextFieldParser parser = new TextFieldParser(new StringReader(lineContent));
parser.SetDelimiters(",");
string[] rawFields = parser.ReadFields();
}
}
}

Best solution so far to deal with CSV values is using the .NET built in libraries:
Its explained here in my StackOverflow answer here:
Reading CSV file and storing values into an array
For easy reference, I am including the code here as well.
using Microsoft.VisualBasic.FileIO;
var path = #"C:\Person.csv"; // Habeeb, "Dubai Media City, Dubai"
using (TextFieldParser csvParser = new TextFieldParser(path))
{
csvParser.CommentTokens = new string[] { "#" };
csvParser.SetDelimiters(new string[] { "," });
csvParser.HasFieldsEnclosedInQuotes = true;
// Skip the row with the column names
csvParser.ReadLine();
while (!csvParser.EndOfData)
{
// Read current line fields, pointer moves to the next line.
string[] fields = csvParser.ReadFields();
string Name = fields[0];
string Address = fields[1];
}
}
More details about the parser is given here: http://codeskaters.blogspot.ae/2015/11/c-easiest-csv-parser-built-in-net.html

Finding all similar lines in a text file

I have a text file that contains some comma separated values. and it looks like this:
3,23500,R,5998,20.38,06/12/2013 01:44:17
2,23500,P,5983,20.234,06/12/2013 01:44:17
3,23501,R,5998,20.38,06/12/2013 01:44:18
2,23501,P,5983,20.235,06/12/2013 01:44:18
3,23502,R,6000,20.4,06/12/2013 01:44:19
2,23502,P,5983,20.236,06/12/2013 01:44:19
3,23503,R,5999,20.39,06/12/2013 01:44:20
2,23503,P,5983,20.236,06/12/2013 01:44:20
My task is to extract lines that start with same number in unique files. Eg in the above case you see some lines are starting with 2 and some with 3...there can be more cases like 4 and etc...
What would be the best and fastes approach to do this? The files that I am working with are quite big and sometimes are in magnitude of gigabytes...
I did split each line and store the first value that will be the number I am looking for in an array and then remove duplicate values from the array...it works but it is very slow!
This is my own code:
private void buttonBeginProcess_Click(object sender, EventArgs e)
{
var file = File.ReadAllLines(_fileName);
var nodeId = new List<int>();
foreach (var line in file)
{
nodeId.Add(int.Parse(line.Split(',')[0]));
}
//Unique numbers
nodeId = nodeId.Distinct().ToList();
}

var lines = File.ReadLines(myFilePath);
var lineGroups = lines
.Where(line => line.Contains(","))
.Select(line => new{key = line.Split(',')[0], line})
.GroupBy(x => x.key);
foreach(var lineGroup in lineGroups)
{
var key = lineGroup.Key;
var keySpecificLines = lineGroup.Select(x => x.line);
//save keySpecificLines to file
}

You could try using StreamReader / StreamWriter to process each file one line at a time:
var writers = new Dictionary<string, StreamWriter>();
using (StreamReader sr = new StreamReader(pathToFile))
{
while (sr.Peek() >= 0)
{
var line = sr.ReadLine();
var key = line.Split(new[]{ ',' },2)[0];
if (!lineGroups.ContainsKey(key))
{
writers[key] = new StreamWriter(GetPathToOutput(key));
}
writers[key].WriteLine(line);
}
}
foreach(StreamWriter sw in writers.Values)
{
sw.Dispose();
}
With this method, you ensure that your code never has to consume the entire input file, so it shouldn't matter how large your input files are. Of course the downside is it would have to keep an arbitrary number of files open throughout the process.

Reading a CSV file in .NET?

How do I read a CSV file using C#?

A choice, without using third-party components, is to use the class Microsoft.VisualBasic.FileIO.TextFieldParser (http://msdn.microsoft.com/en-us/library/microsoft.visualbasic.fileio.textfieldparser.aspx) . It provides all the functions for parsing CSV. It is sufficient to import the Microsoft.VisualBasic assembly.
var parser = new Microsoft.VisualBasic.FileIO.TextFieldParser(file);
parser.TextFieldType = Microsoft.VisualBasic.FileIO.FieldType.Delimited;
parser.SetDelimiters(new string[] { ";" });
while (!parser.EndOfData)
{
string[] row = parser.ReadFields();
/* do something */
}

You can use the Microsoft.VisualBasic.FileIO.TextFieldParser class in C#:
using System;
using System.Data;
using Microsoft.VisualBasic.FileIO;
static void Main()
{
string csv_file_path = #"C:\Users\Administrator\Desktop\test.csv";
DataTable csvData = GetDataTableFromCSVFile(csv_file_path);
Console.WriteLine("Rows count:" + csvData.Rows.Count);
Console.ReadLine();
}
private static DataTable GetDataTableFromCSVFile(string csv_file_path)
{
DataTable csvData = new DataTable();
try
{
using(TextFieldParser csvReader = new TextFieldParser(csv_file_path))
{
csvReader.SetDelimiters(new string[] { "," });
csvReader.HasFieldsEnclosedInQuotes = true;
string[] colFields = csvReader.ReadFields();
foreach (string column in colFields)
{
DataColumn datacolumn = new DataColumn(column);
datacolumn.AllowDBNull = true;
csvData.Columns.Add(datacolumn);
}
while (!csvReader.EndOfData)
{
string[] fieldData = csvReader.ReadFields();
//Making empty value as null
for (int i = 0; i < fieldData.Length; i++)
{
if (fieldData[i] == "")
{
fieldData[i] = null;
}
}
csvData.Rows.Add(fieldData);
}
}
}
catch (Exception ex)
{
}
return csvData;
}

You could try CsvHelper, which is a project I work on. Its goal is to make reading and writing CSV files as easy as possible, while being very fast.
Here are a few ways you can read from a CSV file.
// By type
var records = csv.GetRecords<MyClass>();
var records = csv.GetRecords( typeof( MyClass ) );
// Dynamic
var records = csv.GetRecords<dynamic>();
// Using anonymous type for the class definition
var anonymousTypeDefinition =
{
Id = default( int ),
Name = string.Empty,
MyClass = new MyClass()
};
var records = csv.GetRecords( anonymousTypeDefinition );

I usually use a simplistic approach like this one:
var path = Server.MapPath("~/App_Data/Data.csv");
var csvRows = System.IO.File.ReadAllLines(path, Encoding.Default).ToList();
foreach (var row in csvRows.Skip(1))
{
var columns = row.Split(';');
var field1 = columns[0];
var field2 = columns[1];
var field3 = columns[2];
}

I just used this library in my application. http://www.codeproject.com/KB/database/CsvReader.aspx. Everything went smoothly using this library, so I'm recommending it. It is free under the MIT License, so just include the notice with your source files.
I didn't display the CSV in a browser, but the author has some samples for Repeaters or DataGrids. I did run one of his test projects to test a Sort operation I have added and it looked pretty good.

You can try Cinchoo ETL - an open source lib for reading and writing CSV files.
Couple of ways you can read CSV files
Id, Name
1, Tom
2, Mark
This is how you can use this library to read it
using (var reader = new ChoCSVReader("emp.csv").WithFirstLineHeader())
{
foreach (dynamic item in reader)
{
Console.WriteLine(item.Id);
Console.WriteLine(item.Name);
}
}
If you have POCO object defined to match up with CSV file like below
public class Employee
{
public int Id { get; set; }
public string Name { get; set; }
}
You can parse the same file using this POCO class as below
using (var reader = new ChoCSVReader<Employee>("emp.csv").WithFirstLineHeader())
{
foreach (var item in reader)
{
Console.WriteLine(item.Id);
Console.WriteLine(item.Name);
}
}
Please check out articles at CodeProject on how to use it.
Disclaimer: I'm the author of this library

I recommend Angara.Table, about save/load: http://predictionmachines.github.io/Angara.Table/saveload.html.
It makes column types inference, can save CSV files and is much faster than TextFieldParser. It follows RFC4180 for CSV format and supports multiline strings, NaNs, and escaped strings containing the delimiter character.
The library is under MIT license. Source code is https://github.com/Microsoft/Angara.Table.
Though its API is focused on F#, it can be used in any .NET language but not so succinct as in F#.
Example:
using Angara.Data;
using System.Collections.Immutable;
...
var table = Table.Load("data.csv");
// Print schema:
foreach(Column c in table)
{
string colType;
if (c.Rows.IsRealColumn) colType = "double";
else if (c.Rows.IsStringColumn) colType = "string";
else if (c.Rows.IsDateColumn) colType = "date";
else if (c.Rows.IsIntColumn) colType = "int";
else colType = "bool";
Console.WriteLine("{0} of type {1}", c.Name, colType);
}
// Get column data:
ImmutableArray<double> a = table["a"].Rows.AsReal;
ImmutableArray<string> b = table["b"].Rows.AsString;
Table.Save(table, "data2.csv");

You might be interested in Linq2Csv library at CodeProject. One thing you would need to check is that if it's reading the data when it needs only, so you won't need a lot of memory when working with bigger files.
As for displaying the data on the browser, you could do many things to accomplish it, if you would be more specific on what are your requirements, answer could be more specific, but things you could do:
1. Use HttpListener class to write simple web server (you can find many samples on net to host mini-http server).
2. Use Asp.Net or Asp.Net Mvc, create a page, host it using IIS.

Seems like there are quite a few projects on CodeProject or CodePlex for CSV Parsing.
Here is another CSV Parser on CodePlex
http://commonlibrarynet.codeplex.com/
This library has components for CSV parsing, INI file parsing, Command-Line parsing as well. It's working well for me so far. Only thing is it doesn't have a CSV Writer.

This is just for parsing the CSV. For displaying it in a web page, it is simply a matter of taking the list and rendering it however you want.
Note: This code example does not handle the situation where the input string line contains newlines.
public List<string> SplitCSV(string line)
{
if (string.IsNullOrEmpty(line))
throw new ArgumentException();
List<string> result = new List<string>();
int index = 0;
int start = 0;
bool inQuote = false;
StringBuilder val = new StringBuilder();
// parse line
foreach (char c in line)
{
switch (c)
{
case '"':
inQuote = !inQuote;
break;
case ',':
if (!inQuote)
{
result.Add(line.Substring(start, index - start)
.Replace("\"",""));
start = index + 1;
}
break;
}
index++;
}
if (start < index)
{
result.Add(line.Substring(start, index - start).Replace("\"",""));
}
return result;
}
}

I have been maintaining an open source project called FlatFiles for several years now. It's available for .NET Core and .NET 4.5.1.
Unlike most of the alternatives, it allows you to define a schema (similar to the way EF code-first works) with an extreme level of precision, so you aren't fight conversion issues all the time. You can map directly to your data classes, and there is also support for interfacing with older ADO.NET classes.
Performance-wise, it's been tuned to be one of the fastest parsers for .NET, with a plethora of options for quirky format differences. There's also support for fixed-length files, if you need it.

you can use this library: Sky.Data.Csv
https://www.nuget.org/packages/Sky.Data.Csv/
this is a really fast CSV reader library and it's really easy to use:
using Sky.Data.Csv;
var readerSettings = new CsvReaderSettings{Encoding = Encoding.UTF8};
using(var reader = CsvReader.Create("path-to-file", readerSettings)){
foreach(var row in reader){
//do something with the data
}
}
it also supports reading typed objects with CsvReader<T> class which has a same interface.

C# Exception Handling continue on error

I have a basic C# console application that reads a text file (CSV format) line by line and puts the data into a HashTable. The first CSV item in the line is the key (id num) and the rest of the line is the value. However I've discovered that my import file has a few duplicate keys that it shouldn't have. When I try to import the file the application errors out because you can't have duplicate keys in a HashTable. I want my program to be able to handle this error though. When I run into a duplicate key I would like to put that key into a arraylist and continue importing the rest of the data into the hashtable. How can I do this in C#
Here is my code:
private static Hashtable importFile(Hashtable myHashtable, String myFileName)
{
StreamReader sr = new StreamReader(myFileName);
CSVReader csvReader = new CSVReader();
ArrayList tempArray = new ArrayList();
int count = 0;
while (!sr.EndOfStream)
{
String temp = sr.ReadLine();
if (temp.StartsWith(" "))
{
ServMissing.Add(temp);
}
else
{
tempArray = csvReader.CSVParser(temp);
Boolean first = true;
String key = "";
String value = "";
foreach (String x in tempArray)
{
if (first)
{
key = x;
first = false;
}
else
{
value += x + ",";
}
}
myHashtable.Add(key, value);
}
count++;
}
Console.WriteLine("Import Count: " + count);
return myHashtable;
}

if (myHashtable.ContainsKey(key))
duplicates.Add(key);
else
myHashtable.Add(key, value);

A better solution is to call ContainsKey to check if the key exist before adding it to the hash table instead. Throwing exception on this kind of error is a performance hit and doesn't improve the program flow.

ContainsKey has a constant O(1) overhead for every item, while catching an Exception incurs a performance hit on JUST the duplicate items.
In most situations, I'd say check for the key, but in this case, its better to catch the exception.

Here is a solution which avoids multiple hits in the secondary list with a small overhead to all insertions:
Dictionary<T, List<K>> dict = new Dictionary<T, List<K>>();
//Insert item
if (!dict.ContainsKey(key))
dict[key] = new List<string>();
dict[key].Add(value);
You can wrap the dictionary in a type that hides this or put it in a method or even extension method on dictionary.

If you have more than 4 (for example) CSV values, it might be worth setting the value variable to use a StringBuilder as well since the string concatenation is a slow function.

Hmm, 1.7 Million lines? I hesitate to offer this for that kind of load.
Here's one way to do this using LINQ.
CSVReader csvReader = new CSVReader();
List<string> source = new List<string>();
using(StreamReader sr = new StreamReader(myFileName))
{
while (!sr.EndOfStream)
{
source.Add(sr.ReadLine());
}
}
List<string> ServMissing =
source
.Where(s => s.StartsWith(" ")
.ToList();
//--------------------------------------------------
List<IGrouping<string, string>> groupedSource =
(
from s in source
where !s.StartsWith(" ")
let parsed = csvReader.CSVParser(s)
where parsed.Any()
let first = parsed.First()
let rest = String.Join( "," , parsed.Skip(1).ToArray())
select new {first, rest}
)
.GroupBy(x => x.first, x => x.rest) //GroupBy(keySelector, elementSelector)
.ToList()
//--------------------------------------------------
List<string> myExtras = new List<string>();
foreach(IGrouping<string, string> g in groupedSource)
{
myHashTable.Add(g.Key, g.First());
if (g.Skip(1).Any())
{
myExtras.Add(g.Key);
}
}

Thank you all.
I ended up using the ContainsKey() method. It takes maybe 30 secs longer, which is fine for my purposes. I'm loading about 1.7 million lines and the program takes about 7 mins total to load up two files, compare them, and write out a few files. It only takes about 2 secs to do the compare and write out the files.