基于 Microsoft.VisualBasic.FileIO.TextFieldParser 的简单 CSV 分析器

A simple CSV parser based on Microsoft.VisualBasic.FileIO.TextFieldParser.  The goal of this library is to provide a simple alternative to the `TextFieldParser` class in the `Microsoft.VisualBasic` assembly that works with .NET Core, or just that doesn't require you to use a VisualBasic assembly to a C# project. Installation ============ There's just one real file in this library: [CsvTextFieldParser](CsvTextFieldParser/CsvTextFieldParser.cs). So the recommended way to use this is to just copy that file into your project. This project is available under either of two licenses: [MIT](LICENSE) or [The Unlicense](UNLICENSE). If you're able to use the unlicense, then you don't have to provide any attribution at all. You want to throw this code into your own project and tell everyone you wrote it? Go for it. Alternatively, there are a couple more traditional ways to use this library: * Install the [NuGet package](https://www.nuget.org/packages/CsvTextFieldParser/) * Download the assembly from the [latest release](https://github.com/22222/CsvTextFieldParser/releases/latest) and reference it manually Getting Started =============== Using this library is almost exactly the same as using the VisualBasic `TextFieldParser` class. Here's an example of using this library to parse CSV input into dictionaries: ```csharp public static IEnumerable<IDictionary<string, string>> ParseCsvWithHeader(string csvInput) { using (var csvReader = new StringReader(csvInput)) using (var parser = new NotVisualBasic.FileIO.CsvTextFieldParser(csvReader)) { if (parser.EndOfData) { yield break; } string[] headerFields = parser.ReadFields(); while (!parser.EndOfData) { string[] fields = parser.ReadFields(); int fieldCount = Math.Min(headerFields.Length, fields.Length); IDictionary<string, string> fieldDictionary = new Dictionary<string, string>(fieldCount); for (var i = 0; i < fieldCount; i++) { string headerField = headerFields[i]; string field = fields[i]; fieldDictionary[headerField] = field; } yield return fieldDictionary; } } } ``` But let's take a step back. You've got some CSV data you need to parse: ```csv Name,Birth Date Apollo Creed,1942-08-17 Ivan Drago,1961-11-03 ``` That's easy, it only takes a few lines of code: ```csharp public static IEnumerable<string[]> ParseBasicCsv(string input) { using (var reader = new StringReader(input)) { string line; while ((line = reader.ReadLine()) != null) { if (!string.IsNullOrEmpty(line)) { yield return line.Split(','); } } } } public static void ProcessCsv(string csvInput) { foreach (var csvLine in ParseBasicCsv(csvInput).Skip(1)) { var name = csvLine[0]; var birthDate = csvLine[1]; Console.WriteLine($"{name} was born on {birthDate}"); } } ``` If that's the kind of CSV data you have, then throw this library in the garbage and just use that. Parsing CSV data is simple when your values will never contain any characters that require special handling. But wait, now they've changed the format of the CSV file. Some of the values contain commas, so now you have to put those values in quotes: ```csv Name,Birth Date "Creed, Apollo",1942-08-17 Ivan Drago,1961-11-03 ``` Everything's a little more complicated now And maybe it turns out they sometimes put newlines in those quoted values, and some of the quoted values also have quote characters in them: ```csv Name,Birth Date "Creed, Apollo",1942-08-17 "Ivan Drago",1961-11-03 "Robert ""Rocky"" Balboa",1945-07-06 ``` Now maybe it's time to consider a CSV library. And what luck, the .NET standard library happens to have one built-in. Here's your code now: ```csharp public static void ProcessCsv(string csvInput) { using (var csvReader = new StringReader(csvInput)) using (var parser = new Microsoft.VisualBasic.FileIO.TextFieldParser(csvReader)) { parser.SetDelimiters(","); parser.TrimWhiteSpace = false; // Skip the header line if (!parser.EndOfData) parser.ReadFields(); while (!parser.EndOfData) { var csvLine = parser.ReadFields(); var name = csvLine[0]; var birthDate = csvLine[1]; Console.WriteLine($"{name} was born on {birthDate}"); } } } ``` That works great. It's kind of a weird library, but it handles the quoting thing so you can have commas and newlines in your fields now. So once again, this library can go straight into the garbage. But there's one thing that's kind of gross about this solution: we have to add a dependency on the `Microsoft.VisualBasic` assembly to our C# project. It's still .NET, so that should all work fine. But it feels a little wrong, doesn't it? And there's another problem. The `Microsoft.VisualBasic` probably isn't in .NET Standard, and probably won't ever be. So if you want CSV parsing in a .NET Core application, you're probably back to having to pick a library. One option is to use this one. It's almost exactly the same as using the VisualBasic TextFieldParser: ```csharp public static void ProcessCsv(string csvInput) { using (var csvReader = new StringReader(csvInput)) using (var parser = new NotVisualBasic.FileIO.CsvTextFieldParser(csvReader)) { // Skip the header line if (!parser.EndOfData) parser.ReadFields(); while (!parser.EndOfData) { var csvLine = parser.ReadFields(); var name = csvLine[0]; var birthDate = csvLine[1]; Console.WriteLine($"{name} was born on {birthDate}"); } } } ``` So you can start with either this parser or the VisualBasic one, and you should be able to freely switch between them with only minimal code changes. Error Handling ============== There are two types of input that can cause an error when parsing a CSV file: * A quoted field has trailing characters after the closing quote character * A quoted field is started but never closed by another unescaped quote character Either of these cases will cause a `CsvMalformedLineException` to be thrown. Getting information about these errors works just like handling `Microsoft.VisualBasic.FileIO.MalformedLineException` exceptions from the `Microsoft.VisualBasic.FileIO.TextFieldParser` class: * The exception contains a `LineNumber` property that tells you the number of the line in the file is malformed * The parser class contains an `ErrorLine` property that provides the actual malformed line * The parser class also contains an `ErrorLineNumber` property with the same value as the exception's `LineNumber` property Here's an example of using this parser with error handling: ```csharp public static IEnumerable<IDictionary<string, string>> ParseCsvWithHeaderIgnoreErrors(string csvInput) { using (var csvReader = new StringReader(csvInput)) using (var parser = new NotVisualBasic.FileIO.CsvTextFieldParser(csvReader)) { if (parser.EndOfData) { yield break; } string[] headerFields; try { headerFields = parser.ReadFields(); } catch (NotVisualBasic.FileIO.CsvMalformedLineException ex) { Console.Error.WriteLine($"Failed to parse header line {ex.LineNumber}: {parser.ErrorLine}"); yield break; } while (!parser.EndOfData) { string[] fields; try { fields = parser.ReadFields(); } catch (NotVisualBasic.FileIO.CsvMalformedLineException ex) { Console.Error.WriteLine($"Failed to parse line {ex.LineNumber}: {parser.ErrorLine}"); continue; } int fieldCount = Math.Min(headerFields.Length, fields.Length); IDictionary<string, string> fieldDictionary = new Dictionary<string, string>(fieldCount); for (var i = 0; i < fieldCount; i++) { string headerField = headerFields[i]; string field = fields[i]; fieldDictionary[headerField] =