LinearTsvParser 1.1.7

Linear TSV Parser for .NET Core (read, write)

Install-Package LinearTsvParser -Version 1.1.7
dotnet add package LinearTsvParser --version 1.1.7
<PackageReference Include="LinearTsvParser" Version="1.1.7" />
For projects that support PackageReference, copy this XML node into the project file to reference the package.
paket add LinearTsvParser --version 1.1.7
The NuGet Team does not provide support for this client. Please contact its maintainers for support.

Linear TSV Parser

Reading and writing Linear TSV files in a safe, lossless way. Both async and sync I/O operations are supported.

NuGet package

Available at: https://www.nuget.org/packages/LinearTsvParser

To include it in a .NET Core project:

$ dotnet add package LinearTsvParser

Examples

Reading a .tsv.gz file in async mode:

using System.IO;
using System.IO.Compression;
using System.Threading.Tasks;
using System.Collections.Generic;
using LinearTsvParser;

public class Example {
    public async Task ReadTsv() {
        using var input = File.OpenRead("/tmp/test.tsv.gz");
        using var gzip = new GZipStream(input, CompressionMode.Decompress);
        using var tsvReader = new TsvReader(gzip);

        while(!tsvReader.EndOfStream) {
            List<string> fields = await tsvReader.ReadLineAsync();
        }
    }
}

Writing a .tsv.gz file in sync mode:

using System.IO;
using System.IO.Compression;
using System.Collections.Generic;
using LinearTsvParser;

public class Example {
    public void WriteTsv(List<string[]> data) {
        using var outfile = File.Create("/tmp/test.tsv.gz");
        using var gzip = new GZipStream(outfile, CompressionMode.Compress);
        using var tsvWriter = new TsvWriter(gzip);

        tsvWriter.WriteLine(new List<string>{ "One", "Two\tTwo", "Three" });

        foreach(string[] fields in data) {
            tsvWriter.WriteLine(fields);
        }
    }
}

The writer accepts any enumerable of strings, let it be string[] or List&lt;string&gt;.

You can output the TSV to the standard output by using Console.Out in the constructor:

using System;
using System.IO;
using LinearTsvParser;

public class Example {
    public void WriteTsv() {
        using var tsvWriter = new TsvWriter(Console.Out));

        tsvWriter.WriteLine(new string[] {"One", "Two", "Three"});
    }
}

You can ensure that the data hits the physical drive by calling FlushAsync.

public async Task WriteTsv(Stream output) {
    using var tsvWriter = new TsvWriter(output);
    
    await tsvWriter.WriteLineAsync(new List<string>{ "One", "Two\tTwo", "Three" });
    await tsvWriter.FlushAsync();
}

The Linear TSV format

  • Fields are separated by TAB characters
  • Text encoding is UTF-8
  • The reader can parse lines with any of these three endings: \n, \r\n, \r
  • The writer is restricted to output only the \n character as EOL
  • Special characters inside the fields are replaced (both ways):
    • Newline => &quot;\n&quot;
    • Carriage return => &quot;\r&quot;
    • Tab => &quot;\t&quot;
    • &quot;\&quot; (backslash) => &quot;\\&quot;
  • The column counts are not validated, they can vary per line.

Benchmark

The benchmark test compares the performace of this library with "native" solutions, which use string replace operations. The solution with string replace (native) uses more memory and is slower than this library (lib). The benchmark test can be found here: Benchmark.cs

| Method | Mean | Error | StdDev | Allocated |
|---------------- |---------:|--------:|--------:|----------:|
| LibReadTest | 225.0 ms | 4.42 ms | 5.90 ms | 139.51 MB |
| NativeReadTest | 309.7 ms | 3.84 ms | 3.59 ms | 148.66 MB |
| LibWriteTest | 158.4 ms | 1.92 ms | 1.71 ms | 72.4 MB |
| NativeWriteTest | 270.6 ms | 5.36 ms | 6.38 ms | 86.11 MB |

Configurations

Run the unit tests: (You can also run them one-by-one from VS Code)

$ dotnet build -c Debug
$ dotnet test

Run the benchmark:

$ dotnet run -c Release

Create package for NuGet:

$ dotnet build -c Prod
$ dotnet pack -c Prod

Linear TSV Parser

Reading and writing Linear TSV files in a safe, lossless way. Both async and sync I/O operations are supported.

NuGet package

Available at: https://www.nuget.org/packages/LinearTsvParser

To include it in a .NET Core project:

$ dotnet add package LinearTsvParser

Examples

Reading a .tsv.gz file in async mode:

using System.IO;
using System.IO.Compression;
using System.Threading.Tasks;
using System.Collections.Generic;
using LinearTsvParser;

public class Example {
    public async Task ReadTsv() {
        using var input = File.OpenRead("/tmp/test.tsv.gz");
        using var gzip = new GZipStream(input, CompressionMode.Decompress);
        using var tsvReader = new TsvReader(gzip);

        while(!tsvReader.EndOfStream) {
            List<string> fields = await tsvReader.ReadLineAsync();
        }
    }
}

Writing a .tsv.gz file in sync mode:

using System.IO;
using System.IO.Compression;
using System.Collections.Generic;
using LinearTsvParser;

public class Example {
    public void WriteTsv(List<string[]> data) {
        using var outfile = File.Create("/tmp/test.tsv.gz");
        using var gzip = new GZipStream(outfile, CompressionMode.Compress);
        using var tsvWriter = new TsvWriter(gzip);

        tsvWriter.WriteLine(new List<string>{ "One", "Two\tTwo", "Three" });

        foreach(string[] fields in data) {
            tsvWriter.WriteLine(fields);
        }
    }
}

The writer accepts any enumerable of strings, let it be string[] or List&lt;string&gt;.

You can output the TSV to the standard output by using Console.Out in the constructor:

using System;
using System.IO;
using LinearTsvParser;

public class Example {
    public void WriteTsv() {
        using var tsvWriter = new TsvWriter(Console.Out));

        tsvWriter.WriteLine(new string[] {"One", "Two", "Three"});
    }
}

You can ensure that the data hits the physical drive by calling FlushAsync.

public async Task WriteTsv(Stream output) {
    using var tsvWriter = new TsvWriter(output);
    
    await tsvWriter.WriteLineAsync(new List<string>{ "One", "Two\tTwo", "Three" });
    await tsvWriter.FlushAsync();
}

The Linear TSV format

  • Fields are separated by TAB characters
  • Text encoding is UTF-8
  • The reader can parse lines with any of these three endings: \n, \r\n, \r
  • The writer is restricted to output only the \n character as EOL
  • Special characters inside the fields are replaced (both ways):
    • Newline => &quot;\n&quot;
    • Carriage return => &quot;\r&quot;
    • Tab => &quot;\t&quot;
    • &quot;\&quot; (backslash) => &quot;\\&quot;
  • The column counts are not validated, they can vary per line.

Benchmark

The benchmark test compares the performace of this library with "native" solutions, which use string replace operations. The solution with string replace (native) uses more memory and is slower than this library (lib). The benchmark test can be found here: Benchmark.cs

| Method | Mean | Error | StdDev | Allocated |
|---------------- |---------:|--------:|--------:|----------:|
| LibReadTest | 225.0 ms | 4.42 ms | 5.90 ms | 139.51 MB |
| NativeReadTest | 309.7 ms | 3.84 ms | 3.59 ms | 148.66 MB |
| LibWriteTest | 158.4 ms | 1.92 ms | 1.71 ms | 72.4 MB |
| NativeWriteTest | 270.6 ms | 5.36 ms | 6.38 ms | 86.11 MB |

Configurations

Run the unit tests: (You can also run them one-by-one from VS Code)

$ dotnet build -c Debug
$ dotnet test

Run the benchmark:

$ dotnet run -c Release

Create package for NuGet:

$ dotnet build -c Prod
$ dotnet pack -c Prod

  • .NETCoreApp 3.1

    • No dependencies.

NuGet packages (1)

Showing the top 1 NuGet packages that depend on LinearTsvParser:

Package Downloads
FatCatDB
Zero configuration, high performance database library for ETL workflows

GitHub repositories

This package is not used by any popular GitHub repositories.

Version History

Version Downloads Last updated
1.1.7 614 1/18/2020
1.1.6 230 1/14/2020
1.1.5 129 1/14/2020
1.1.4 177 1/12/2020
1.1.3 191 1/12/2020
1.1.2 196 1/12/2020
1.1.1 204 1/12/2020
1.1.0 196 1/12/2020
1.0.7 146 1/12/2020
1.0.6 142 1/12/2020
1.0.5 160 1/12/2020
1.0.4 159 1/12/2020
1.0.3 162 1/12/2020
1.0.2 220 1/12/2020
1.0.1 232 1/11/2020
1.0.0 242 1/11/2020