csv_from_table 0.1.0

data/Gemfile

@@ -0,0 +1,2 @@
+gem 'rspec'
+gem 'nokogiri'

data/Gemfile.lock

@@ -0,0 +1,19 @@
+GEM
+  specs:
+    diff-lcs (1.1.3)
+    nokogiri (1.5.0)
+    rspec (2.6.0)
+      rspec-core (~> 2.6.0)
+      rspec-expectations (~> 2.6.0)
+      rspec-mocks (~> 2.6.0)
+    rspec-core (2.6.4)
+    rspec-expectations (2.6.0)
+      diff-lcs (~> 1.1.2)
+    rspec-mocks (2.6.0)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  nokogiri
+  rspec

data/lib/csv_from_table.rb

@@ -0,0 +1,65 @@
+require 'rubygems'
+require 'bundler'
+Bundler.require
+
+class String
+
+  def csv_from_table(options={})
+    @csv_from_table = CSVFromTable.new
+    array = @csv_from_table.ar_from_table(self, options)
+    @csv_from_table.ar_to_csv(array)
+  end
+
+end
+
+class CSVFromTable
+
+  def ar_from_table(table_string, options={})
+    @options = options
+    doc = Nokogiri::HTML(table_string)
+    if !doc.css("table").empty?
+      @headers = table_headers(table_string)
+      lines = []
+      doc.css("tr").each do |node|
+        cells = []
+        node.css("td").each_with_index do |c, index|
+          cells << c.text if include_colunm?(index)
+        end
+        lines << cells
+      end
+      lines.keep_if { |el| !el.empty? }
+    else
+      raise TableNotFoundInString, "The string dosen't contain a valid <table> element."
+    end
+  end
+
+  def include_colunm?(index)
+    if @options[:only]
+      true if @options[:only].include?(@headers[index]) || @options[:only] == @headers[index]
+    elsif @options[:except]
+      true unless @options[:except].include?(@headers[index]) || @options[:except] == @headers[index]
+    else
+      true
+    end
+  end
+
+  def ar_to_csv(array)
+    csv_lines = []
+    array.each do |el|
+      csv_lines << el.join(", ")
+    end
+    csv_lines.join("\n")
+  end
+
+  def table_headers(table_string)
+    doc = Nokogiri::HTML(table_string)
+    headers = []
+    doc.css("th").each do |node|
+      headers << node.text
+    end
+    headers
+  end
+
+end
+
+class TableNotFoundInString < StandardError; end

data/specs/csv_from_table_spec.rb

@@ -0,0 +1,60 @@
+$LOAD_PATH << '../lib'
+require 'csv_from_table'
+
+describe CSVFromTable do
+
+  before do
+    @sample = File.open("sample_table.html").read
+    @csv_from_table = CSVFromTable.new
+  end
+
+  it "should receive a HTML string with a table and return a array of arrays with the table's content" do
+    @csv_from_table.ar_from_table(@sample).length.should == 5
+    @csv_from_table.ar_from_table(@sample).should == [["Baseball bat", "$75.99", "500g", "Probably", "20%"],
+                                    ["Harmonica", "R$ 150,00", "100g", "Sure!", "0%"],
+                                    ["Pencil", "$1", "80g", "Just if there's an eraser", "100%"],
+                                    ["Eraser", "R$ 2,00", "120g", "Sure", "80%"],
+                                    ["Chocolate", "R$ 56", "5 Kg", "Sure", "0%"]]
+  end
+
+  it "should turn a array of arrays into CSV formated string" do
+    sample_ar = @csv_from_table.ar_from_table(@sample)
+    @csv_from_table.ar_to_csv(sample_ar).should == "Baseball bat, $75.99, 500g, Probably, 20%\nHarmonica, R$ 150,00, 100g, Sure!, 0%\nPencil, $1, 80g, Just if there's an eraser, 100%\nEraser, R$ 2,00, 120g, Sure, 80%\nChocolate, R$ 56, 5 Kg, Sure, 0%"
+  end
+
+end
+
+describe String do
+
+  before do
+    @sample_html_table = File.open("sample_table.html").read
+  end
+
+  it "should return the CSV format from a table" do
+    @sample_html_table.csv_from_table.should == "Baseball bat, $75.99, 500g, Probably, 20%\nHarmonica, R$ 150,00, 100g, Sure!, 0%\nPencil, $1, 80g, Just if there's an eraser, 100%\nEraser, R$ 2,00, 120g, Sure, 80%\nChocolate, R$ 56, 5 Kg, Sure, 0%"
+  end
+
+  it "should raise error when there's no table on the given string" do
+    sample_html_table = "No html table here!"
+    lambda { sample_html_table.csv_from_table }.should(raise_error(TableNotFoundInString))
+  end
+
+  describe "should receive options to filter colunms" do
+    it ":only => 'colunm name'" do
+     @sample_html_table.csv_from_table(:only => "Thing").should == "Baseball bat\nHarmonica\nPencil\nEraser\nChocolate"
+    end
+
+    it ":only => 'colunm array'" do
+      @sample_html_table.csv_from_table(:only => ["Thing", "Weight"]).should == "Baseball bat, 500g\nHarmonica, 100g\nPencil, 80g\nEraser, 120g\nChocolate, 5 Kg"
+    end
+
+    it ":except => 'colunm name'" do
+      @sample_html_table.csv_from_table(:except => "Price").should == "Baseball bat, 500g, Probably, 20%\nHarmonica, 100g, Sure!, 0%\nPencil, 80g, Just if there's an eraser, 100%\nEraser, 120g, Sure, 80%\nChocolate, 5 Kg, Sure, 0%"
+    end
+
+    it ":except => 'colunm name'" do
+      @sample_html_table.csv_from_table(:except => ["Price", "Is it worthy?", "Chances I forget it there"]).should == "Baseball bat, 500g\nHarmonica, 100g\nPencil, 80g\nEraser, 120g\nChocolate, 5 Kg"
+    end
+  end
+
+end

data/specs/sample_table.html

@@ -0,0 +1,47 @@
+<h2>What if hter's more markup than the table?</h2>
+<p>Extra texts shouldn't get into the CSV version of the table.</p>
+<table>
+  <caption>Table of thing I would take with me in a trip to the moon!</caption>
+  <tr>
+    <th>Thing</th>
+    <th>Price</th>
+    <th>Weight</th>
+    <th>Is it worthy?</th>
+    <th>Chances I forget it there</th>
+  </tr>
+  <tr>
+    <td>Baseball bat</td>
+    <td>$75.99</td>
+    <td>500g</td>
+    <td>Probably</td>
+    <td>20%</td>
+  </tr>
+  <tr>
+    <td>Harmonica</td>
+    <td>R$ 150,00</td>
+    <td>100g</td>
+    <td>Sure!</td>
+    <td>0%</td>
+  </tr>
+  <tr>Text like this shouldn't be in the CSV
+    <td>Pencil</td>
+    <td>$1</td>
+    <td>80g</td>
+    <td>Just if there's an eraser</td>
+    <td>100%</td>
+  </tr>
+  <tr>
+    <td>Eraser</td>
+    <td>R$ 2,00</td>
+    <td>120g</td>
+    <td>Sure</td>
+    <td>80%</td>
+  </tr>
+  <tr>
+    <td>Chocolate</td>
+    <td>R$ 56</td>
+    <td>5 Kg</td>
+    <td>Sure</td>
+    <td>0%</td>
+  </tr>
+</table>

metadata

@@ -0,0 +1,52 @@
+--- !ruby/object:Gem::Specification
+name: csv_from_table
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+  prerelease: 
+platform: ruby
+authors:
+- Marco Antonio Fogaça Nogueira
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2011-12-24 00:00:00.000000000Z
+dependencies: []
+description: Ruby library that spits out CSV from given HTML table, so you can use
+  the view helpers when formatting the CSV without having to include them on you models
+  or controllers.
+email:
+- marcofognog@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- Gemfile
+- Gemfile.lock
+- lib/csv_from_table.rb
+- specs/csv_from_table_spec.rb
+- specs/sample_table.html
+homepage: http://github.com/marcofognog/csv_from_table
+licenses: []
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: 1.3.7
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.6
+signing_key: 
+specification_version: 3
+summary: Ruby library that spits out CSV from given HTML table.
+test_files: []