csharpvitamins/CSharpVitamins.Tabulation

Handles conversion of arbitrary data into plain text

CSharpVitamins.Tabulation

This library provides helpers for creating tabular data with minimal fuss.

Easily create tab or comma separated values (via CsvDefinition), or padded and aligned columns of plain text (with PlainTextTable).

Available on NuGet. To install, run the following command in the Package Manager Console:

PM> Install-Package CSharpVitamins.Tabulation

Usage: CsvDefinition

CsvDefinition is designed for quick production of CSV style text - it aims to keep the column header and how to produce the value close together.

// vars to be used inside column value funcs
const decimal VIP_SPEND = 1000000M;

// the definition
var fields = new CsvDefinition<MyEntity>
{
  { "Username",      row => string.Concat(row.Prefix, ":", row.Username) },
  { "Email",         row => row.Email },
  { "Status",        row => row.TotalSpend >= VIP_SPEND ? "VIP!$!" : "Oh No!" },
  { "Spend ($)",     row => row.TotalSpend == null ? "None" : row.TotalSpend.ToString("n2") },
  { "Year of Birth", row => row.Yob > 1900 && row.Yob <= DateTime.UtcNow.Year ? row.Yob.Value.ToString() : null },
  { "Enabled",       row => row.IsEnabled ? "Y" : "N" },
  { "Flags",         row => string.Join(", ", new string[]
                     {
                       row.SomeFlag ? "some" : null,
                       row.AnotherFlag ? "another" : null,
                       row.YetAnotherFlag ? "even-more" : null
                     }.Where(x => x != null))
  },
  { "Timestamp",     row => DateTime.UtcNow.ToString("yyyy-MM-dd HH:mm:ss.fff") },
};

// the data
IEnumerable<MyEntity> rows = someDataSource.AsEnumerable<MyEntity>();

// the result
using (var writer = new StringWriter())
{
  // perhaps write out some additional notes or headers
  // writer.WriteLine("# -- bof --");

  // write out some tab delimited data
  fields.Write(writer, rows, "\t");

  // writer.WriteLine("# -- eof --");
  return writer.ToString();
}

At its heart, CsvDefinition<T> is a wrapper around List<KeyValuePair<string, Func<T, string>>> List<CsvField<T>> with a .Write method. You can call .Add or .Remove to modify your base definition after it's created.

if (!User.IsInRole("Admin")) {
  fields.Remove("Spend ($)");
}

Alternatively, from version 2.0.0, you can also define an include function when creating the definition. The function which will be evaluated during rendering.

var fields = new CsvDefinition<MyEntity>
{
  { "Username",  row => row.Username },
  { "Email",     row => row.Email },
  { "Spend ($)", row => row.Spend,   key => User.IsInRole("Admin") },
  { "My Field",  row => row.MyField, key => AllowedFields.Contains(key) },
  ...
};

Customising field labels can be achieved by providing another string argument at second position, which makes it easier to work with more complex scenarios, say filtering keys based on prefix or different sources.

var fields = new CsvDefinition<MyEntity>
{
  { "user.name",     "Username", row => row.Username },
  { "user.email",    "Email",    row => row.Email },
  { "company.name",  "Company",  row => row.Company.Name },
  { "company.addr1", "Address",  row => row.Address.Line1, key => AllowedFields.Contains(key) },
  { "company.addr2", "",         row => row.Address.Line2, key => AllowedFields.Contains(key) },
  { "company.addr3", "",         row => row.Address.City,  key => AllowedFields.Contains(key) },
  ...
};

Or have the definition created for you, from a class

var fields = new CsvDefinitionFactory().CreateFromModel<MyEntity>();

... get data, create writer, etc...

fields.Write(writer, rows, ",");

If you want more control over the production of the results, you can specify Func<PropertyInfo, object, string> converters for the type.

var factory = new CsvDefinitionFactory();

// optionally limit what's included
factory.ShouldInclude = (prop) => AllowFields.Contains(prop.Name);

// convert header names to lower case
factory.NameConverter = (prop) => prop.Name.ToLowerInvariant();

// make bools convert to Y/N
factory.ValueConverters.Add(typeof(bool), (prop, value) => (bool)value ? "Y" : "N");

// use the factory to create multiple definitions
var def1 = factory.CreateFromModel<MyEntity>();

var def2 = factory.CreateFromModel<AnotherClass>();

Usage: PlainTextTable

A PlainTextTable allows padding of tabular data so it can be displayed easily as text.

• Works best with small to medium length strings of data
• Does not support multi-line text (newlines) or tabs
• Since columns are padded using spaces, output from this class is best presented using a fixed-width font

Example 1: classic

var tab = new PlainTextTable();

// header
tab.AddRow("Name", "Enabled?", "Title");

// lines of data
foreach (var row in someDataSource) {
  tab.AddRow(row.Name, row.IsEnabled ? "Y" : "N", row.Title);
}

tab.Align('L', 'C');

return tab.ToString();

which might return

Name    Enabled? Job
Dave       Y     Developer
Sarah      Y     Designer
Mustafa    N     Data Analysis

Example 2: fluid

Using the fluid style for a quick dump of data...

return new PlainTextTable()
  .AddRow("Name", "Size", "Status") // optionally add a header
  .ImportRows(someDataSource.Select(x => new [] { x.Name, x.Size, x.Status }))
  .SeparateBy(" | ")
  .Align('L', 'R', 'C')
  .TrimTrailingSpace()
  .ToString();

which might produce something like this

Name               |    Size |  State
TPS Report rev 362 |   42 MB | started
readme.md          |    6 KB |  late
TPS Report final   | 0 bytes |  late

Dividers can also be injected to separate headers, content and footers, e.g.

return new PlainTextTable()
  .Align('L', 'R')
  .SeparateBy("   ")
  .AddRow("Name", "Size")
  .Divide('-')
  .ImportRows(someDataSource.Select(x => new [] { x.Name, x.Size.ToString() }))
  .Divide('-')
  .AddRow(null, someDataSource.Sum(x => x.Size).ToString())
  .ToString();

resulting in...

Name             Size
---------------------
File1.txt        5 KB
File2.config    12 KB
File3.exe      137 KB
---------------------
               154 KB

Happy coding!