gmail_extractor 0.0.1

data/.gitignore

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

data/Gemfile

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

data/README.md

@@ -0,0 +1,24 @@
+# GmailExtractor
+
+A small command line tool to extract emails from Google Mail (GMail).
+
+## Usage
+
+    $> gem install gmail_extractor
+    $> gmail_extractor --help
+
+## Examples
+
+To download all emails within the label LABEL and the GMail account test123:
+
+    $> gmail_extractor --user test123 --label LABEL
+
+## Development
+
+This gem is built ontop of the following gems:
+
+* `gmail`
+* `progressbar`
+* `highline`
+
+It has been tested on Ruby 1.9.3-p0!

data/Rakefile

@@ -0,0 +1 @@
+require "bundler/gem_tasks"

data/bin/gmail_extractor

@@ -0,0 +1,4 @@
+#!/usr/bin/env ruby
+# -*- mode: ruby -*-
+require "gmail_extractor"
+GmailExtractor.execute

data/gmail_extractor.gemspec

@@ -0,0 +1,24 @@
+# -*- encoding: utf-8 -*-
+$:.push File.expand_path("../lib", __FILE__)
+require "gmail_extractor/version"
+
+Gem::Specification.new do |s|
+  s.name        = "gmail_extractor"
+  s.version     = GmailExtractor::VERSION
+  s.authors     = ["Simon Harrer"]
+  s.email       = ["simon.harrer@gmail.com"]
+  s.homepage    = 'https://github.com/simonharrer/gmail_extractor'
+  s.summary     = %q{Downloads emails from Google Mail to XML, CSV or HTML}
+  s.description = s.summary
+
+  s.rubyforge_project = "gmail_extractor"
+
+  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_runtime_dependency "gmail"
+  s.add_runtime_dependency "progressbar"
+  s.add_runtime_dependency "highline"
+end

data/lib/gmail_extractor.rb

@@ -0,0 +1,100 @@
+# encoding: utf-8
+require "gmail_extractor/version"
+
+# data structures
+require "gmail_extractor/email"
+
+# formatter
+require "gmail_extractor/email_formatter"
+
+# printer
+require "gmail_extractor/printer"
+
+# printer decorators
+require "gmail_extractor/email_progressbar"
+
+# gmail communication
+require "gmail_extractor/email_extractor"
+
+module GmailExtractor
+
+  def self.execute
+    # Reading from command line
+    require "optparse"
+    options = {}
+    OptionParser.new do |opts|
+      opts.banner = "Usage: get_mails.rb [options]"
+      opts.on("-u","--username USERNAME","Set gmail USERNAME") do |username|
+        options[:username] = username
+      end
+      opts.on("-p","--password PASSWORD","Set gmail PASSWORD") do |password|
+        options[:password] = password
+      end
+      opts.on("-n","--name NAME","Set own NAME; defaults to Self if not given") do |name|
+        options[:name] = name || "Self"
+      end
+      opts.on("-m","--mailbox MAILBOX","Set MAILBOX from where the mails are fetched. Label names can be used too!") do |mailbox|
+        options[:mailbox] = mailbox
+      end
+      opts.on("-l","--limit LIMIT","Set limit LIMIT to how many emails are fetched; defaults to all if not given") do |limit|
+        options[:limit] = limit.to_i
+      end
+      opts.on("-o","--output TYPE","Set output TYPE to xml, csv or html. defaults to csv") do |output|
+        options[:output] = output || "csv"
+      end
+      opts.on("-f","--file FILE","Set output FILE. if not present, only output to STDOUT") do |file|
+        options[:file] = file
+      end
+      opts.on("-pb","--progressbar","Show progressbar on STDOUT") do
+        options[:progressbar] = true
+      end
+      opts.on("-h", "--help", "Show this message") do
+        puts opts
+        exit
+      end
+    end.parse!
+
+    # set printer
+    file = options[:file]
+    if file
+      printer = ConsolePrinter.new(FilePrinter.new(file))
+    else
+      printer = ConsolePrinter.new
+    end
+
+    # set correct formatter
+    if options[:output]
+      output = options[:output].downcase
+    else
+      output = "csv"
+    end
+
+    if output == "xml"
+      formatter = EmailXmlFormatter.new(printer)
+    elsif output == "csv"
+      formatter = EmailCsvFormatter.new(printer)
+    elsif output == "html"
+      formatter = EmailHtmlFormatter.new(printer)
+    else
+      raise "Internal error - no output option given"
+    end
+
+    # handle progressbar
+    if options[:progressbar]
+      formatter = EmailProgressbar.new(formatter)
+    end
+
+    # Asking missing values
+    require "highline/import"
+    user = options[:username] || ask("Enter user: ")
+    password = options[:password] || ask("Enter password: ") { |q| q.echo = false }
+    name = options[:name]
+    extractor = EmailExtractor.new(user,password,name,formatter)
+
+    # Ask missing values and extract
+    label = options[:mailbox] || ask("Enter mailbox/label: ")
+    limit = options[:limit]
+    extractor.extract(label, limit)
+  end
+
+end

data/lib/gmail_extractor/email.rb

@@ -0,0 +1,13 @@
+module GmailExtractor
+
+  class Email
+    attr_accessor :from, :content, :date
+
+    def initialize(from, date, content)
+      @from = from
+      @date = date
+      @content = content
+    end
+  end
+
+end

data/lib/gmail_extractor/email_extractor.rb

@@ -0,0 +1,69 @@
+require "gmail"
+require "date"
+
+module GmailExtractor
+
+  # Controls the extraction
+  class EmailExtractor
+    attr_accessor :user, :password, :printer, :name
+
+    def initialize(user, password, name, printer)
+      @user = user
+      @password = password
+      @printer = printer
+      @name = name
+    end
+
+    def extract(label, limit)
+      connect
+      extract_mails_for label, limit
+      disconnect
+    end
+
+    private
+    def extract_mails_for(label, limit)
+      header
+      body label, limit
+      footer
+    end
+
+    def disconnect
+      @gmail.logout
+    end
+
+    def connect
+      @gmail = Gmail.connect(user, password)
+      raise "password/username wrong" unless @gmail.logged_in?
+    end
+
+    def body(label, limit)
+      label = @gmail.label(label)
+      printer.label(label, limit)
+
+      mails = 0
+      label.emails do |email|
+        from = email.from.first.name || @name
+        date = DateTime.parse(email.date)
+        content = email.body.to_s.force_encoding("UTF-8")
+
+        mail = Email.new(from, date, content)
+
+        printer.email(mail)
+
+        mails += 1
+        if limit and mails == limit
+          break
+        end
+      end
+    end
+
+    def header
+      printer.header
+    end
+
+    def footer
+      printer.footer
+    end
+  end
+
+end

data/lib/gmail_extractor/email_formatter.rb

@@ -0,0 +1,3 @@
+require "gmail_extractor/email_formatter/email_csv_formatter"
+require "gmail_extractor/email_formatter/email_xml_formatter"
+require "gmail_extractor/email_formatter/email_html_formatter"

data/lib/gmail_extractor/email_formatter/email_csv_formatter.rb

@@ -0,0 +1,25 @@
+require "csv"
+
+module GmailExtractor
+
+  class EmailCsvFormatter
+
+    attr_accessor :printer
+
+    def initialize(printer)
+      @printer = printer
+    end
+
+    def header;  end
+
+    def footer;  end
+
+    def email(email)
+      @printer << [email.date, email.from, email.content].to_csv
+    end
+
+    def label(label,limit); end
+
+  end
+
+end

data/lib/gmail_extractor/email_formatter/email_html_formatter.rb

@@ -0,0 +1,48 @@
+module GmailExtractor
+
+  # Prints the output as html
+  class EmailHtmlFormatter
+
+    attr_accessor :printer
+
+    def initialize(printer)
+      @printer = printer
+    end
+
+    def header
+      result = "<!doctype html><html><head><title>E-Mails</title><meta charset='utf-8'></head><body>"
+
+      @printer << result
+    end
+
+    def footer
+      result = "</body></html>"
+
+      @printer << result
+      @printer.close
+    end
+
+    def email(email)
+      result = ""
+      result << "<div class='email'>"
+      result << "<div class='from'>#{email.from}</div>"
+      result << "<div class='date'>#{email.date}</div>"
+      result << "<div class='content'>#{email.content}</div>"
+      result << "</div>"
+
+      @printer << result
+    end
+
+    def label(label,limit)
+      result = ""
+      result << "<div class='label'>"
+      result << "<div class='name'>#{label.name}</div>"
+      result << "<div class='count'>#{label.count}</div>"
+      result << "</div>"
+
+      @printer << result
+    end
+
+  end
+
+end

data/lib/gmail_extractor/email_formatter/email_xml_formatter.rb

@@ -0,0 +1,39 @@
+module GmailExtractor
+
+  class EmailXmlFormatter
+
+    attr_accessor :printer
+
+    def initialize(printer)
+      @printer = printer
+    end
+
+    def header
+      result = "<emails>"
+
+      @printer << result
+    end
+
+    def footer
+      result = "</emails>"
+
+      @printer << result
+      @printer.close
+    end
+
+    def email(email)
+      result = ""
+      result << "<email>"
+      result << "<from>#{email.from}</from>"
+      result << "<date>#{email.date}</date>"
+      result << "<content>#{email.content}</content>"
+      result << "</email>"
+
+      @printer << result
+    end
+
+    def label(label,limit); end
+
+  end
+
+end

data/lib/gmail_extractor/email_progressbar.rb

@@ -0,0 +1,36 @@
+require "progressbar"
+
+module GmailExtractor
+
+  # Simple printer proxy to enable tracking the progress with a progress bar transparently.
+  class EmailProgressbar
+
+    attr_accessor :printer
+
+    def initialize(printer)
+      @printer = printer
+    end
+
+    def header
+      printer.header
+    end
+
+    def footer
+      printer.footer
+      @progress_bar.finish
+    end
+
+    def email(email)
+      printer.email(email)
+      @progress_bar.inc
+    end
+
+    def label(label,limit)
+      printer.label(label,limit)
+      total_count = limit || label.count
+      @progress_bar = ProgressBar.new("emails",total_count)
+    end
+
+  end
+
+end

data/lib/gmail_extractor/printer.rb

@@ -0,0 +1,2 @@
+require "gmail_extractor/printer/console_printer"
+require "gmail_extractor/printer/file_printer"

data/lib/gmail_extractor/printer/console_printer.rb

@@ -0,0 +1,25 @@
+module GmailExtractor
+
+  class ConsolePrinter
+
+    def initialize(printer = nil)
+      @printer = printer
+    end
+
+    def << (string)
+      puts string
+      @printer << string if printer?
+    end
+
+    def close
+      @printer.close if printer?
+    end
+
+    private
+    def printer?
+      @printer
+    end
+
+  end
+
+end

data/lib/gmail_extractor/printer/file_printer.rb

@@ -0,0 +1,19 @@
+module GmailExtractor
+
+  class FilePrinter
+
+    def initialize(output_file)
+      @output_file = File.open(output_file,"w:utf-8")
+      @output_file.sync = true
+    end
+
+    def << (string)
+      @output_file << string
+    end
+
+    def close
+      @output_file.close
+    end
+  end
+
+end

data/lib/gmail_extractor/version.rb

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

metadata

@@ -0,0 +1,97 @@
+--- !ruby/object:Gem::Specification
+name: gmail_extractor
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Simon Harrer
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2012-03-05 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: gmail
+  requirement: &29237616 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: *29237616
+- !ruby/object:Gem::Dependency
+  name: progressbar
+  requirement: &29253564 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: *29253564
+- !ruby/object:Gem::Dependency
+  name: highline
+  requirement: &29254740 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: *29254740
+description: Downloads emails from Google Mail to XML, CSV or HTML
+email:
+- simon.harrer@gmail.com
+executables:
+- gmail_extractor
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- README.md
+- Rakefile
+- bin/gmail_extractor
+- gmail_extractor.gemspec
+- lib/gmail_extractor.rb
+- lib/gmail_extractor/email.rb
+- lib/gmail_extractor/email_extractor.rb
+- lib/gmail_extractor/email_formatter.rb
+- lib/gmail_extractor/email_formatter/email_csv_formatter.rb
+- lib/gmail_extractor/email_formatter/email_html_formatter.rb
+- lib/gmail_extractor/email_formatter/email_xml_formatter.rb
+- lib/gmail_extractor/email_progressbar.rb
+- lib/gmail_extractor/printer.rb
+- lib/gmail_extractor/printer/console_printer.rb
+- lib/gmail_extractor/printer/file_printer.rb
+- lib/gmail_extractor/version.rb
+homepage: https://github.com/simonharrer/gmail_extractor
+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: '0'
+requirements: []
+rubyforge_project: gmail_extractor
+rubygems_version: 1.8.11
+signing_key: 
+specification_version: 3
+summary: Downloads emails from Google Mail to XML, CSV or HTML
+test_files: []