mail_extract 0.1.2 → 0.1.3

data/.travis.yml

@@ -0,0 +1,4 @@
+rvm:
+  - 1.8.7
+  - 1.9.2
+  - ree

data/Gemfile

@@ -0,0 +1,3 @@
+source 'http://rubygems.org'
+
+gemspec

data/lib/mail_extract.rb

@@ -7,8 +7,8 @@ module MailExtract
     #
     # @return [MailExtract::Parser]
     #
-    def new(body)
-      MailExtract::Parser.new(body)
+    def new(body, options={})
+      MailExtract::Parser.new(body, options)
     end
   end
 end

data/lib/mail_extract/line.rb

@@ -1,17 +1,20 @@
 module MailExtract
   class Line
-    attr_reader :body, :type
+    attr_reader :body, :type, :subtype
     
     PATTERNS = {
-      /^[>]+\s?/    => :quote,
-      /^--/         => :signature,
-      /^-- /        => :signature,
-      /^[_]{2,}\n?/ => :signature,
-      /^[-]{2,}\n?/ => :signature
+      /^[>]+\s?/                     => :quote,
+      /^--/                          => :signature,
+      /^-- /                         => :signature,
+      /^[_]{2,}\n?/                  => :signature,
+      /^[-]{2,}\n?/                  => :signature,
+      /^sent from my (iphone|ipad)/i => :signature 
     }
     
     def initialize(str)
-      @body = str
+      @body    = str
+      @subtype = :none
+      
       detect_type(str)
     end
     
@@ -38,11 +41,13 @@ module MailExtract
     def detect_type(line)
       # Detects the start line of quote text
       if line.strip =~ /^On\s/i && line =~ /at [\d:]+/ || line.strip =~ />? wrote:\z/
-        @type = :quote
+        @type    = :quote
+        @subtype = :start
         return
       end
       
       @type = :text
+      
       PATTERNS.each_pair do |p,t|
         if line =~ p
           @type = t

data/lib/mail_extract/parser.rb

@@ -6,14 +6,20 @@ module MailExtract
     
     # Initialize a new MailExtract::Parser object
     #
-    # text - Email message body
+    # text    - Email message body
+    # options - Parsing options
     #
-    def initialize(text)
+    # Parsing options include:
+    #   :only_head - Skip the rest of the message after quote start (default: false)
+    #
+    def initialize(text, options={})
       @lines     = []
       @text      = text.strip
       @body      = ""
       @last_type = :text
       @type      = :text
+      @options   = options
+      
       parse
     end
     
@@ -22,13 +28,25 @@ module MailExtract
     # Process email message body
     #
     def parse
+      break_after_quote = @options[:only_head] || false
       scanner = StringScanner.new(@text)
+      
+      # Process until message end
       while str = scanner.scan_until(/\n/)
-        parse_line(str)
+        line = parse_line(str)
+        
+        if break_after_quote
+          break if line.quote? && line.subtype == :start
+        end
       end
-      if (last_line = scanner.rest.to_s).size > 0
-        parse_line(last_line)
+      
+      # Process the rest (if any)
+      if !break_after_quote && @last_type != :quote
+        if (last_line = scanner.rest.to_s).size > 0
+          parse_line(last_line)
+        end
       end
+      
       @body = @lines.join("\n").strip
     end
     
@@ -36,6 +54,7 @@ module MailExtract
     #
     def parse_line(str)
       line = MailExtract::Line.new(str)
+      
       if line.quote?
         if @last_type == :text      ; @type = :quote     ; end
       elsif line.text?
@@ -47,6 +66,8 @@ module MailExtract
       end
       @last_type = line.type
       @lines << line.body.strip if @type == :text
+      
+      line
     end
   end
 end

data/lib/mail_extract/version.rb

@@ -1,5 +1,5 @@
 module MailExtract
   unless defined? ::MailExtract::VERSION
-    VERSION = "0.1.2".freeze
+    VERSION = "0.1.3".freeze
   end
 end

data/mail_extract.gemspec

@@ -15,5 +15,6 @@ Gem::Specification.new do |gem|
   gem.executables   = `git ls-files -- bin/*`.split("\n").map{|f| File.basename(f)}
   gem.require_paths = ['lib']
 
+  gem.add_development_dependency 'rake'
   gem.add_development_dependency 'rspec', '~> 2.6'
 end

data/spec/fixtures/iphone.txt

@@ -0,0 +1,3 @@
+This is a shit i sent from my iphone
+
+Sent from my iPhone

data/spec/fixtures/iphone_with_quotes.txt

@@ -0,0 +1,18 @@
+
+Primary reply content
+
+--
+I take full responsibility for any typos and refuse to blame them on my
+cellphone
+
+On Aug 12, 2011, at 6:00 AM, Robot <noreply@foobar.com> wrote:
+
+Hi Dude,
+
+Some text goes here, with no quotes
+Blah blah blah
+Blah!
+Blah!!!!
+
+Thanks,
+Whatever

data/spec/line_spec.rb

@@ -7,12 +7,15 @@ describe 'MailExtract::Line' do
   
   it 'detects quote start by date' do
     line('On Tue, 2011-03-01 at 18:02 +0530, somebody wrote:').type.should == :quote
+    line('On Tue, 2011-03-01 at 18:02 +0530, somebody wrote:').subtype.should == :start
     line('On 2011-03-01 at 18:02 somebody wrote').type.should == :quote
+    line('On 2011-03-01 at 18:02 somebody wrote').subtype.should == :start
     line('On some day somebody wrote').type.should == :text
   end
   
   it 'detects quote' do
     line('> this is a quote').type.should == :quote
+    line('> this is a quote').subtype.should == :none
     line('> >> this is a quote').type.should == :quote
   end
   
@@ -20,7 +23,9 @@ describe 'MailExtract::Line' do
     lines = [
      "--\nUsername",
      "-- \nUsername",
-     "_______\nSome text"
+     "_______\nSome text",
+     "Sent from my iPhone",
+     "Sent from my iPad"
     ]
     
     lines.each do |l|

data/spec/parser_spec.rb

@@ -2,17 +2,27 @@ require 'spec_helper'
 
 describe 'MailExtract::Parser' do
   it 'parses an email' do
-    body = MailExtract.new(fixture('simple.txt')).body
+    body = parse_fixture('simple.txt')
     body.should == result_fixture('simple.txt')
   end
   
   it 'parses an email with quotes' do
-    body = MailExtract.new(fixture('simple_with_quotes.txt')).body
+    body = parse_fixture('simple_with_quotes.txt')
     body.should == result_fixture('simple_with_quotes.txt')
   end
   
   it 'parses a reply email with broken authored line' do
-    body = MailExtract.new(fixture('reply_with_quotes.txt')).body
+    body = parse_fixture('reply_with_quotes.txt')
     body.should == 'This is a first line of the message'
   end
+  
+  it 'parses a message send via iphone' do
+    body = parse_fixture('iphone.txt')
+    body.should == 'This is a shit i sent from my iphone'
+  end
+
+  it 'parses a reply sent via iphone' do
+    body = MailExtract.new(fixture('iphone_with_quotes.txt'), :only_head => true).body
+    body.should == 'Primary reply content'
+  end
 end

data/spec/spec_helper.rb

@@ -15,3 +15,7 @@ end
 def result_fixture(file)
   fixture("result_#{file}")
 end
+
+def parse_fixture(file)
+  MailExtract.new(fixture(file)).body
+end

metadata

@@ -1,35 +1,53 @@
---- !ruby/object:Gem::Specification
+--- !ruby/object:Gem::Specification 
 name: mail_extract
-version: !ruby/object:Gem::Version
-  version: 0.1.2
+version: !ruby/object:Gem::Version 
   prerelease: 
+  version: 0.1.3
 platform: ruby
-authors:
+authors: 
 - Dan Sosedoff
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-08-11 00:00:00.000000000Z
-dependencies:
-- !ruby/object:Gem::Dependency
+
+date: 2011-08-12 00:00:00 -05:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: rake
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+  type: :development
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
   name: rspec
-  requirement: &72663040 !ruby/object:Gem::Requirement
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
     none: false
-    requirements:
+    requirements: 
     - - ~>
-      - !ruby/object:Gem::Version
-        version: '2.6'
+      - !ruby/object:Gem::Version 
+        version: "2.6"
   type: :development
-  prerelease: false
-  version_requirements: *72663040
+  version_requirements: *id002
 description: Email body parser that strips out all quotes and signatures.
 email: dan.sosedoff@gmail.com
 executables: []
+
 extensions: []
+
 extra_rdoc_files: []
-files:
+
+files: 
 - .gitignore
 - .rspec
+- .travis.yml
+- Gemfile
 - README.md
 - Rakefile
 - lib/mail_extract.rb
@@ -37,6 +55,8 @@ files:
 - lib/mail_extract/parser.rb
 - lib/mail_extract/version.rb
 - mail_extract.gemspec
+- spec/fixtures/iphone.txt
+- spec/fixtures/iphone_with_quotes.txt
 - spec/fixtures/reply_with_quotes.txt
 - spec/fixtures/result_simple.txt
 - spec/fixtures/result_simple_with_quotes.txt
@@ -45,28 +65,42 @@ files:
 - spec/line_spec.rb
 - spec/parser_spec.rb
 - spec/spec_helper.rb
+has_rdoc: true
 homepage: https://github.com/sosedoff/mail_extract
 licenses: []
+
 post_install_message: 
 rdoc_options: []
-require_paths:
+
+require_paths: 
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement
+required_ruby_version: !ruby/object:Gem::Requirement 
   none: false
-  requirements:
-  - - ! '>='
-    - !ruby/object:Gem::Version
-      version: '0'
-required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
   none: false
-  requirements:
-  - - ! '>='
-    - !ruby/object:Gem::Version
-      version: '0'
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
 requirements: []
+
 rubyforge_project: 
-rubygems_version: 1.8.5
+rubygems_version: 1.6.2
 signing_key: 
 specification_version: 3
 summary: Extracts email message body
-test_files: []
+test_files: 
+- spec/fixtures/iphone.txt
+- spec/fixtures/iphone_with_quotes.txt
+- spec/fixtures/reply_with_quotes.txt
+- spec/fixtures/result_simple.txt
+- spec/fixtures/result_simple_with_quotes.txt
+- spec/fixtures/simple.txt
+- spec/fixtures/simple_with_quotes.txt
+- spec/line_spec.rb
+- spec/parser_spec.rb
+- spec/spec_helper.rb