Class Cucumber::Ast::Table
In: lib/cucumber/ast/table.rb
Parent: Object

Step Definitions that match a plain text Step with a multiline argument table will receive it as an instance of Table. A Table object holds the data of a table parsed from a feature file and lets you access and manipulate the data in different ways.

For example:

  Given I have:
    | a | b |
    | c | d |

And a matching StepDefinition:

  Given /I have:/ do |table|
    data = table.raw
  end

This will store [[‘a’, ‘b’], [‘c’, ‘d’]] in the data variable.

Methods

diff!   dup   hashes   map_column!   map_headers   map_headers!   new   raw   rows   rows_hash   transpose  

Included Modules

Enumerable

Constants

NULL_CONVERSIONS = Hash.new(lambda{ |cell_value| cell_value }).freeze
TO_S_PREFIXES = Hash.new(' ')

Attributes

file  [RW] 

Public Class methods

Creates a new instance. raw should be an Array of Array of String. You don‘t typically create your own Table objects - Cucumber will do it internally and pass them to your Step Definitions.

Public Instance methods

Compares other_table to self. If other_table contains columns and/or rows that are not in self, new columns/rows are added at the relevant positions, marking the cells in those rows/columns as surplus. Likewise, if other_table lacks columns and/or rows that are present in self, these are marked as missing.

surplus and missing cells are recognised by formatters and displayed so that it‘s easy to read the differences.

Cells that are different, but look identical (for example the boolean true and the string "true") are converted to their Object#inspect representation and preceded with (i) - to make it easier to identify where the difference actually is.

Since all tables that are passed to StepDefinitions always have String objects in their cells, you may want to use map_column! before calling diff!. You can use map_column! on either of the tables.

An exception is raised if there are missing rows or columns, or surplus rows. An error is not raised for surplus columns. Whether to raise or not raise can be changed by setting values in options to true or false:

  • missing_row : Raise on missing rows (defaults to true)
  • surplus_row : Raise on surplus rows (defaults to true)
  • missing_col : Raise on missing columns (defaults to true)
  • surplus_col : Raise on surplus columns (defaults to false)

The other_table argument can be another Table, an Array of Array or an Array of Hash (similar to the structure returned by hashes).

Calling this method is particularly useful in Then steps that take a Table argument, if you want to compare that table to some actual values.

Creates a copy of this table, inheriting any column mappings. registered with map_headers!

Converts this table into an Array of Hash where the keys of each Hash are the headers in the table. For example, a Table built from the following plain text:

  | a | b | sum |
  | 2 | 3 | 5   |
  | 7 | 9 | 16  |

Gets converted into the following:

  [{'a' => '2', 'b' => '3', 'sum' => '5'}, {'a' => '7', 'b' => '9', 'sum' => '16'}]

Use map_column! to specify how values in a column are converted.

Change how hashes converts column values. The column_name argument identifies the column and conversion_proc performs the conversion for each cell in that column. If strict is true, an error will be raised if the column named column_name is not found. If strict is false, no error will be raised. Example:

  Given /^an expense report for (.*) with the following posts:$/ do |table|
    posts_table.map_column!('amount') { |a| a.to_i }
    posts_table.hashes.each do |post|
      # post['amount'] is a Fixnum, rather than a String
    end
  end

Returns a new Table where the headers are redefined. See map_headers!

Redefines the table headers. This makes it possible to use prettier and more flexible header names in the features. The keys of mappings are Strings or regular expressions (anything that responds to #=== will work) that may match column headings in the table. The values of mappings are desired names for the columns.

Example:

  | Phone Number | Address |
  | 123456       | xyz     |
  | 345678       | abc     |

A StepDefinition receiving this table can then map the columns with both Regexp and String:

  table.map_headers!(/phone( number)?/i => :phone, 'Address' => :address)
  table.hashes
  # => [{:phone => '123456', :address => 'xyz'}, {:phone => '345678', :address => 'abc'}]

You may also pass in a block if you wish to convert all of the headers:

  table.map_headers! { |header| header.downcase }
  table.hashes.keys
  # => ['phone number', 'address']

When a block is passed in along with a hash then the mappings in the hash take precendence:

  table.map_headers!('Address' => 'ADDRESS') { |header| header.downcase }
  table.hashes.keys
  # => ['phone number', 'ADDRESS']

Gets the raw data of this table. For example, a Table built from the following plain text:

  | a | b |
  | c | d |

gets converted into the following:

  [['a', 'b], ['c', 'd']]

Same as raw, but skips the first (header) row

Converts this table into a Hash where the first column is used as keys and the second column is used as values

  | a | 2 |
  | b | 3 |

Gets converted into the following:

  {'a' => '2', 'b' => '3'}

The table must be exactly two columns wide

Returns a new, transposed table. Example:

  | a | 7 | 4 |
  | b | 9 | 2 |

Gets converted into the following:

  | a | b |
  | 7 | 9 |
  | 4 | 2 |

[Validate]