我正在读取*.csv文件。
*.csv-file由分号(";")分隔的两列组成。
我能够阅读*.csv文件使用StreamReader，并能够通过使用Split()函数分离每一行。我想将每一列存储到一个单独的数组中，然后显示它。
有可能做到吗?
当前回答
这是我的2个简单的静态方法，将文本从csv文件转换为列表<列表<字符串>>，反之亦然。每种方法都使用行转换器。
这段代码应该考虑csv文件的所有可能性。您可以定义自己的csv分隔符，该方法尝试纠正转义双引号字符，并处理当所有文本在引号中是一个单元格，csv分隔符在引号字符串中，包括一个单元格中的多行，可以忽略空行。
最后一种方法仅用于测试。所以你可以忽略它，或者用这个测试方法测试你自己的或别人的解决方案:)。为了测试，我使用了这个硬csv, 4行2行:
0,a,""bc,d
"e, f",g,"this,is, o
ne ""lo
ng, cell""",h
这是最终代码。为了简单起见，我删除了所有的try catch块。
using System;
using System.Collections.Generic;
using System.Linq;
public static class Csv {
public static string FromListToString(List<List<string>> csv, string separator = ",", char quotation = '"', bool returnFirstRow = true)
{
string content = "";
for (int row = 0; row < csv.Count; row++) {
content += (row > 0 ? Environment.NewLine : "") + RowFromListToString(csv[row], separator, quotation);
}
return content;
}
public static List<List<string>> FromStringToList(string content, string separator = ",", char quotation = '"', bool returnFirstRow = true, bool ignoreEmptyRows = true)
{
List<List<string>> csv = new List<List<string>>();
string[] rows = content.Split(new string[] { Environment.NewLine }, StringSplitOptions.None);
if (rows.Length <= (returnFirstRow ? 0 : 1)) { return csv; }
List<string> csvRow = null;
for (int rowIndex = 0; rowIndex < rows.Length; rowIndex++) {
(List<string> row, bool rowClosed) = RowFromStringToList(rows[rowIndex], csvRow, separator, quotation);
if (rowClosed) { if (!ignoreEmptyRows || row.Any(rowItem => rowItem.Length > 0)) { csv.Add(row); csvRow = null; } } // row ok, add to list
else { csvRow = row; } // not fully created, continue
}
if (!returnFirstRow) { csv.RemoveAt(0); } // remove header
return csv;
}
public static string RowFromListToString(List<string> csvData, string separator = ",", char quotation = '"')
{
csvData = csvData.Select(element =>
{
if (element.Contains(quotation)) {
element = element.Replace(quotation.ToString(), quotation.ToString() + quotation.ToString());
}
if (element.Contains(separator) || element.Contains(Environment.NewLine)) {
element = "\"" + element + "\"";
}
return element;
}).ToList();
return string.Join(separator, csvData);
}
public static (List<string>, bool) RowFromStringToList(string csvRow, List<string> continueWithRow = null, string separator = ",", char quotation = '"')
{
bool rowClosed = true;
if (continueWithRow != null && continueWithRow.Count > 0) {
// in previous result quotation are fixed so i need convert back to double quotation
string previousCell = quotation.ToString() + continueWithRow.Last().Replace(quotation.ToString(), quotation.ToString() + quotation.ToString()) + Environment.NewLine;
continueWithRow.RemoveAt(continueWithRow.Count - 1);
csvRow = previousCell + csvRow;
}
char tempQuote = (char)162;
while (csvRow.Contains(tempQuote)) { tempQuote = (char)(tempQuote + 1); }
char tempSeparator = (char)(tempQuote + 1);
while (csvRow.Contains(tempSeparator)) { tempSeparator = (char)(tempSeparator + 1); }
csvRow = csvRow.Replace(quotation.ToString() + quotation.ToString(), tempQuote.ToString());
if(csvRow.Split(new char[] { quotation }, StringSplitOptions.None).Length % 2 == 0) { rowClosed = !rowClosed; }
string[] csvSplit = csvRow.Split(new string[] { separator }, StringSplitOptions.None);
List<string> csvList = csvSplit
.ToList()
.Aggregate("",
(string row, string item) => {
if (row.Count((ch) => ch == quotation) % 2 == 0) { return row + (row.Length > 0 ? tempSeparator.ToString() : "") + item; }
else { return row + separator + item; }
},
(string row) => row.Split(tempSeparator).Select((string item) => item.Trim(quotation).Replace(tempQuote, quotation))
).ToList();
if (continueWithRow != null && continueWithRow.Count > 0) {
return (continueWithRow.Concat(csvList).ToList(), rowClosed);
}
return (csvList, rowClosed);
}
public static bool Test()
{
string csvText = "0,a,\"\"bc,d" + Environment.NewLine + "\"e, f\",g,\"this,is, o" + Environment.NewLine + "ne \"\"lo" + Environment.NewLine + "ng, cell\"\"\",h";
List<List<string>> csvList = new List<List<string>>() { new List<string>() { "0", "a", "\"bc", "d" }, new List<string>() { "e, f", "g", "this,is, o" + Environment.NewLine + "ne \"lo" + Environment.NewLine + "ng, cell\"", "h" } };
List<List<string>> csvTextAsList = Csv.FromStringToList(csvText);
bool ok = Enumerable.SequenceEqual(csvList[0], csvTextAsList[0]) && Enumerable.SequenceEqual(csvList[1], csvTextAsList[1]);
string csvListAsText = Csv.FromListToString(csvList);
return ok && csvListAsText == csvText;
}
}
使用例子:
// get List<List<string>> representation of csv
var csvFromText = Csv.FromStringToList(csvAsText);
// read csv file with custom separator and quote
// return no header and ignore empty rows
var csvFile = File.ReadAllText(csvFileFullPath);
var csvFromFile = Csv.FromStringToList(csvFile, ";", '"', false, false);
// get text representation of csvData from List<List<string>>
var csvAsText = Csv.FromListToString(csvData);
注: 这:char tempQuote = (char)162;是ASCI表中的第一个稀有字符。脚本会搜索这个字符，或者不是文本中的第一个接下来的几个ascii字符，并将其用作临时转义和引用字符。
其他回答
你可以这样做:
using System.IO;
static void Main(string[] args)
{
using(var reader = new StreamReader(@"C:\test.csv"))
{
List<string> listA = new List<string>();
List<string> listB = new List<string>();
while (!reader.EndOfStream)
{
var line = reader.ReadLine();
var values = line.Split(';');
listA.Add(values[0]);
listB.Add(values[1]);
}
}
}
如果你需要跳过(头部)行和/或列，你可以使用它来创建一个2维数组:
var lines = File.ReadAllLines(path).Select(a => a.Split(';'));
var csv = (from line in lines
select (from col in line
select col).Skip(1).ToArray() // skip the first column
).Skip(2).ToArray(); // skip 2 headlines
如果您需要在进一步处理数据之前对其进行塑形，这是非常有用的(假设前两行由标题组成，而第一列是行标题—您不需要在数组中拥有行标题，因为您只想考虑数据)。
注意:您可以通过使用以下代码轻松获得标题和第一列:
var coltitle = (from line in lines
select line.Skip(1).ToArray() // skip 1st column
).Skip(1).Take(1).FirstOrDefault().ToArray(); // take the 2nd row
var rowtitle = (from line in lines select line[0] // take 1st column
).Skip(2).ToArray(); // skip 2 headlines
这个代码示例假设你的*.csv文件的结构如下:
注意:如果你需要跳过空行——有时这很方便，你可以通过插入来实现
where line.Any(a=>!string.IsNullOrWhiteSpace(a))
在上面的LINQ代码示例中的from和select语句之间。
我只是一个写硕士论文的学生，但这就是我解决问题的方法，对我来说效果很好。首先从目录中选择文件(仅以csv格式)，然后将数据放入列表中。
List<float> t = new List<float>();
List<float> SensorI = new List<float>();
List<float> SensorII = new List<float>();
List<float> SensorIII = new List<float>();
using (OpenFileDialog dialog = new OpenFileDialog())
{
try
{
dialog.Filter = "csv files (*.csv)|*.csv";
dialog.Multiselect = false;
dialog.InitialDirectory = ".";
dialog.Title = "Select file (only in csv format)";
if (dialog.ShowDialog() == DialogResult.OK)
{
var fs = File.ReadAllLines(dialog.FileName).Select(a => a.Split(';'));
int counter = 0;
foreach (var line in fs)
{
counter++;
if (counter > 2) // Skip first two headder lines
{
this.t.Add(float.Parse(line[0]));
this.SensorI.Add(float.Parse(line[1]));
this.SensorII.Add(float.Parse(line[2]));
this.SensorIII.Add(float.Parse(line[3]));
}
}
}
}
catch (Exception exc)
{
MessageBox.Show(
"Error while opening the file.\n" + exc.Message,
this.Text,
MessageBoxButtons.OK,
MessageBoxIcon.Error
);
}
}
这里有一个特殊的情况，其中一个数据字段有分号(“;”)作为它的数据的一部分，在这种情况下，上面的大多数答案将失败。
这种情况下的解决方案是
string[] csvRows = System.IO.File.ReadAllLines(FullyQaulifiedFileName);
string[] fields = null;
List<string> lstFields;
string field;
bool quoteStarted = false;
foreach (string csvRow in csvRows)
{
lstFields = new List<string>();
field = "";
for (int i = 0; i < csvRow.Length; i++)
{
string tmp = csvRow.ElementAt(i).ToString();
if(String.Compare(tmp,"\"")==0)
{
quoteStarted = !quoteStarted;
}
if (String.Compare(tmp, ";") == 0 && !quoteStarted)
{
lstFields.Add(field);
field = "";
}
else if (String.Compare(tmp, "\"") != 0)
{
field += tmp;
}
}
if(!string.IsNullOrEmpty(field))
{
lstFields.Add(field);
field = "";
}
// This will hold values for each column for current row under processing
fields = lstFields.ToArray();
}
开源的Angara。表库允许将CSV加载到类型化列中，因此您可以从列中获取数组。每个列都可以按名称或索引进行索引。见http://predictionmachines.github.io/Angara.Table/saveload.html。
该库遵循RFC4180 for CSV;它支持类型推断和多行字符串。
例子:
using System.Collections.Immutable;
using Angara.Data;
using Angara.Data.DelimitedFile;
...
ReadSettings settings = new ReadSettings(Delimiter.Semicolon, false, true, null, null);
Table table = Table.Load("data.csv", settings);
ImmutableArray<double> a = table["double-column-name"].Rows.AsReal;
for(int i = 0; i < a.Length; i++)
{
Console.WriteLine("{0}: {1}", i, a[i]);
}
您可以使用类型column来查看列类型，例如。
Column c = table["double-column-name"];
Console.WriteLine("Column {0} is double: {1}", c.Name, c.Rows.IsRealColumn);
因为这个库的重点是f#，你可能需要添加一个对FSharp的引用。Core 4.4总成;在项目中单击“添加引用”，并选择FSharp。“程序集”下的核心4.4 ->“扩展”。
