C#에서 텍스트 파일을 CSV로 변환

On .NET you can convert text to CSV with built-in file APIs for simple cases, or with CsvHelper — the de-facto CSV library — when quoting and encoding need to be exactly right. Both are shown below with runnable code.

For a single file, the browser tool is faster; the C# code fits services, desktop apps and scheduled jobs.

Install CsvHelper (NuGet: CsvHelper). Read the source with its delimiter and write properly quoted CSV:

using System.Globalization;
using System.IO;
using CsvHelper;
using CsvHelper.Configuration;

var inConfig = new CsvConfiguration(CultureInfo.InvariantCulture) { Delimiter = "\t" };
using var reader = new StreamReader("input.txt");
using var csvIn = new CsvReader(reader, inConfig);
using var writer = new StreamWriter("output.csv");
using var csvOut = new CsvWriter(writer, CultureInfo.InvariantCulture);

while (csvIn.Read())
{
    for (int i = 0; i < csvIn.Parser.Count; i++)
        csvOut.WriteField(csvIn.GetField(i));
    csvOut.NextRecord();
}

CsvHelper escapes fields containing commas, quotes or newlines, producing valid CSV automatically.

For anything beyond clean tab- or space-separated input, prefer CsvHelper over String.Split: manual splitting cannot handle a value that contains the delimiter, an embedded quote or a line break, and those cases are exactly where data corruption hides. CsvHelper reads and writes to the standard, so the output re-imports everywhere.

Encoding deserves attention. Writing with new UTF8Encoding(true) emits a BOM that makes Excel open international characters correctly; omit it (new UTF8Encoding(false)) for files consumed by tools that dislike a BOM. Set the Delimiter on a CsvConfiguration to match your input, and use CultureInfo.InvariantCulture so number and date formatting do not vary by machine locale.

As with every language here, leading zeros survive in the CSV file itself — if they vanish, it is Excel reinterpreting the file on open, so apply the fix at the Excel import step. Verify by reading the output back with CsvHelper and checking the field count of the first few records.

For large files, avoid File.ReadAllLines, which loads everything into memory; read with a streaming StreamReader line by line, or let CsvHelper’s GetRecords yield records lazily. If you map rows to a strongly typed class, CsvHelper can read and write those objects directly, which is cleaner and less error-prone than indexing fields by position when the shape is fixed and known.

In short, reach for CsvHelper whenever data might contain commas, quotes or newlines, fall back to a simple split only for clean input, and always be explicit about encoding and culture. With those settled, the same conversion runs reliably inside a service, a desktop app or a scheduled task. The result is correct, properly quoted CSV that opens in Excel, Google Sheets and databases without any rework.