colander 0.0.1

data/.gitignore

@@ -0,0 +1,4 @@
+*.gem
+.bundle
+Gemfile.lock
+pkg/*

data/.rvmrc

@@ -0,0 +1 @@
+rvm --create ruby-1.8.7-p302@email_extractor

data/Gemfile

@@ -0,0 +1,4 @@
+source "http://rubygems.org"
+
+# Specify your gem's dependencies in colander.gemspec
+gemspec

data/README.md

@@ -0,0 +1,15 @@
+Colander
+==============
+Colander is a gem to collect emails from various file formats.
+
+Supported formats
+
+  * "Old" Ms Excel files, ending in .xls
+  * Ms Excel files ending with .xlsx
+
+Usage
+-----
+
+    parser = Colander.parse("/path/to/file.xls")
+    parser.emails # => ["foo@bar.com"]
+

data/Rakefile

@@ -0,0 +1,2 @@
+require 'bundler'
+Bundler::GemHelper.install_tasks

data/colander.gemspec

@@ -0,0 +1,26 @@
+# -*- encoding: utf-8 -*-
+$:.push File.expand_path("../lib", __FILE__)
+require "colander/version"
+
+Gem::Specification.new do |s|
+  s.name        = "colander"
+  s.version     = Colander::VERSION
+  s.platform    = Gem::Platform::RUBY
+  s.authors     = ["Kristian Hellquist", "Jonas Forsberg"]
+  s.email       = ["dev@mynewsdesk.com"]
+  s.homepage    = "http://devcorner.mynewsdesk.com"
+  s.summary     = %q{Exctract an array of emails from various file formats}
+  s.description = %q{See summary. lol}
+
+  # s.rubyforge_project = "mnd_colander"
+
+  s.files         = `git ls-files`.split("\n")
+  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
+  s.require_paths = ["lib"]
+
+  s.add_dependency "nokogiri", "~> 1.4.4"
+  s.add_dependency "roo", "~> 1.9.5"
+  s.add_dependency "zip", "~> 2.0.2"
+  s.add_development_dependency "rspec"
+end

data/lib/colander/invalid_file.rb

@@ -0,0 +1,4 @@
+module Colander
+  class InvalidFile < StandardError
+  end
+end

data/lib/colander/parser/base.rb

@@ -0,0 +1,15 @@
+module Colander
+  module Parser
+    class Base
+      attr_reader :emails
+
+      def initialize(file_path)
+        @file_path = file_path
+      end
+
+      def parse
+        raise "plz implement me in"
+      end
+    end
+  end
+end

data/lib/colander/parser/null.rb

@@ -0,0 +1,14 @@
+require 'colander/parser/base'
+
+module Colander
+  module Parser
+    class Null
+      def parse
+      end
+      def emails
+        []
+      end
+    end
+  end
+end
+

data/lib/colander/parser/xls.rb

@@ -0,0 +1,30 @@
+require 'colander/invalid_file'
+require 'colander/parser/base'
+require 'roo'
+
+module Colander
+  module Parser
+    class Xls < Base
+      def parse
+        spreadsheet = parse_file
+        @emails = collect_emails spreadsheet
+      rescue TypeError, IOError
+        raise InvalidFile
+      end
+
+      protected
+
+      def parse_file
+        Excel.new(@file_path,nil,:ignore)
+      end
+
+      def collect_emails(spreadsheet)
+        spreadsheet.sheets.map do |sheet|
+          spreadsheet.default_sheet = sheet
+          spreadsheet.to_yaml.scan(/\b[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,4}\b/)
+        end.flatten
+      end
+
+    end
+  end
+end

data/lib/colander/parser/xlsx.rb

@@ -0,0 +1,13 @@
+require 'colander/parser/base'
+require 'roo'
+
+module Colander
+  module Parser
+    class Xlsx < Xls
+      protected
+      def parse_file
+        Excelx.new(@file_path,nil,:ignore)
+      end
+    end
+  end
+end

data/lib/colander/version.rb

@@ -0,0 +1,3 @@
+module Colander
+  VERSION = "0.0.1"
+end

data/lib/colander.rb

@@ -0,0 +1,18 @@
+require 'colander/invalid_file'
+require 'colander/parser/xls'
+require 'colander/parser/xlsx'
+
+module Colander
+  def self.parse(file_path, file_name = nil)
+    parser = case (file_name || file_path).split(".").last
+    when "xls"
+      Parser::Xls.new(file_path)
+    when "xlsx"
+      Parser::Xlsx.new(file_path)
+    else
+      raise InvalidFile
+    end
+    parser.parse
+    parser
+  end
+end

data/spec/base_spec.rb

@@ -0,0 +1,10 @@
+require 'spec_helper'
+
+describe Colander::Parser::Base do
+  describe "#emails" do
+    it "should responses to it" do
+      Colander::Parser::Base.any_instance.stub(:parse)
+      Colander::Parser::Base.new("foo/bar").should respond_to :emails
+    end
+  end
+end

data/spec/colander_spec.rb

@@ -0,0 +1,45 @@
+require 'spec_helper'
+
+describe Colander do
+  describe ".parse" do
+    it "should raise error if no file path is passed" do
+      lambda{
+        Colander.parse
+      }.should raise_exception(ArgumentError)
+    end
+
+    it "should raise error if file is unsupported" do
+      lambda {
+        Colander.parse("foo.bar")
+      }.should raise_error(Colander::InvalidFile)
+    end
+
+    it "should not raise error if a file path is passed" do
+      lambda{
+        Colander.parse("/file/path")
+      }.should_not raise_exception(ArgumentError)
+    end
+
+    it "accepts an option file name and returns correct parser" do
+      Colander::Parser::Xls.any_instance.stub(:parse)
+      Colander.parse("/file/path", "apa.xls").should be_a Colander::Parser::Xls
+    end
+
+    it "should recognize a xls file and return correct parser" do
+      path = "/foo/apa.xls"
+      Colander::Parser::Xls.any_instance.stub(:parse)
+      Colander.parse(path).should be_a Colander::Parser::Xls
+    end
+
+    it "should recognize a xlsx file and return correct parser" do
+      path = "/foo/apa.xlsx"
+      Colander::Parser::Xlsx.any_instance.stub(:parse)
+      Colander.parse(path).should be_a Colander::Parser::Xlsx
+    end
+
+    it "should invoke parse method on the parser" do
+      Colander::Parser::Xlsx.any_instance.should_receive(:parse)
+      Colander.parse("/foo/bar.xlsx")
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1,3 @@
+$:.push File.expand_path("../lib", __FILE__)
+
+require 'colander'

data/spec/xls_spec.rb

@@ -0,0 +1,36 @@
+require 'spec_helper'
+
+describe Colander::Parser::Xls do
+  describe "#parse" do
+    it "stores found emails" do
+      Excel.stub(:new)
+      parser = Colander::Parser::Xls.new("file/path")
+      parser.should_receive(:collect_emails).and_return(["bruce@wayne.com"])
+      parser.parse
+      parser.emails.sort.should eql(["bruce@wayne.com"])
+    end
+
+    it "raises exception if file is invalid" do
+      parser = Colander::Parser::Xls.new("./#{__FILE__}")
+      lambda {
+        parser.parse
+      }.should raise_error(Colander::InvalidFile)
+    end
+
+    it "retreives emails from an 95-excel spreadsheet" do
+      parser = Colander::Parser::Xls.new("spec/fixtures/excel95.xls")
+      parser.parse
+      parser.emails.should eql(["markus.nordin@mynewsdesk.com", "markus@hej.se", "sven@bertil.se", "Adam.A@hotmail.com", "apa@elabs.se", "liam@neeson.net", "david@mynewsdesk.com"])
+    end
+    it "retreives emails from an xls spreadsheet" do
+      parser = Colander::Parser::Xls.new("spec/fixtures/old-format.xls")
+      parser.parse
+      parser.emails.should eql(["markus.nordin@mynewsdesk.com", "markus@hej.se", "sven@bertil.se", "Adam.A@hotmail.com", "apa@elabs.se", "liam@neeson.net", "david@mynewsdesk.com"])
+    end
+    it "retreives emails from an 95-excel spreadsheet without file suffix" do
+      parser = Colander::Parser::Xls.new("spec/fixtures/excel95-without-file-suffix")
+      parser.parse
+      parser.emails.should eql(["markus.nordin@mynewsdesk.com", "markus@hej.se", "sven@bertil.se", "Adam.A@hotmail.com", "apa@elabs.se", "liam@neeson.net", "david@mynewsdesk.com"])
+    end
+  end
+end

data/spec/xlsx_spec.rb

@@ -0,0 +1,14 @@
+require 'spec_helper'
+describe Colander::Parser::Xlsx do
+  it "is subclass of Colander::Parser::Xls" do
+     Colander::Parser::Xlsx.new("foo/bar").should be_a Colander::Parser::Xls
+  end
+
+  describe "#parse" do
+    it "retreives emails from an xlsx spreadsheet" do
+      parser = Colander::Parser::Xlsx.new("spec/fixtures/new-format.xlsx")
+      parser.parse
+      parser.emails.should eql(["markus.nordin@mynewsdesk.com", "markus@hej.se", "sven@bertil.se", "Adam.A@hotmail.com", "apa@elabs.se", "liam@neeson.net", "david@mynewsdesk.com"])
+    end
+  end
+end

metadata

@@ -0,0 +1,157 @@
+--- !ruby/object:Gem::Specification 
+name: colander
+version: !ruby/object:Gem::Version 
+  hash: 29
+  prerelease: 
+  segments: 
+  - 0
+  - 0
+  - 1
+  version: 0.0.1
+platform: ruby
+authors: 
+- Kristian Hellquist
+- Jonas Forsberg
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2011-08-29 00:00:00 Z
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: nokogiri
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 15
+        segments: 
+        - 1
+        - 4
+        - 4
+        version: 1.4.4
+  type: :runtime
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: roo
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 57
+        segments: 
+        - 1
+        - 9
+        - 5
+        version: 1.9.5
+  type: :runtime
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: zip
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 11
+        segments: 
+        - 2
+        - 0
+        - 2
+        version: 2.0.2
+  type: :runtime
+  version_requirements: *id003
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id004 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  type: :development
+  version_requirements: *id004
+description: See summary. lol
+email: 
+- dev@mynewsdesk.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- .gitignore
+- .rvmrc
+- Gemfile
+- README.md
+- Rakefile
+- colander.gemspec
+- lib/colander.rb
+- lib/colander/invalid_file.rb
+- lib/colander/parser/base.rb
+- lib/colander/parser/null.rb
+- lib/colander/parser/xls.rb
+- lib/colander/parser/xlsx.rb
+- lib/colander/version.rb
+- spec/base_spec.rb
+- spec/colander_spec.rb
+- spec/fixtures/excel95-without-file-suffix
+- spec/fixtures/excel95.xls
+- spec/fixtures/new-format.xlsx
+- spec/fixtures/old-format.xls
+- spec/spec_helper.rb
+- spec/xls_spec.rb
+- spec/xlsx_spec.rb
+homepage: http://devcorner.mynewsdesk.com
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.8.10
+signing_key: 
+specification_version: 3
+summary: Exctract an array of emails from various file formats
+test_files: 
+- spec/base_spec.rb
+- spec/colander_spec.rb
+- spec/fixtures/excel95-without-file-suffix
+- spec/fixtures/excel95.xls
+- spec/fixtures/new-format.xlsx
+- spec/fixtures/old-format.xls
+- spec/spec_helper.rb
+- spec/xls_spec.rb
+- spec/xlsx_spec.rb