mail 1.0.0 → 1.1.0

data/CHANGELOG.rdoc

@@ -0,0 +1,28 @@
+== Tue Nov  3 00:59:45 UTC 2009 Mikel Lindsaar <raasdnil@gmail.com>
+
+* Tested mail against entire Enron set (2.3gb) and the Trec 2005 set (0.5gb), ~ half a million emails without crashing
+* Some headers only can appear once, enforce during header parse assignment. <jlindley>
+* Convert empty bodies into empty arrays instead of nil. <jlindley>
+* Handle blank content dispositions. <jlindley>
+* Mention Trec 2005 Spam Corpus in readme <jlindley>
+* Add 'rake corpus:verify_all' to allow parse checks in bulk. <jlindley>
+* Added handling of multi value parameters, like filename*1*="us-ascii'en'blah" filename*2="bleh" <mikel>
+* Added dependency on ActiveSupport 2.3 or higher <mikel>
+
+== Sun Nov  1 12:00:00 UTC 2009 Mikel Lindsaar <raasdnil@gmail.com>
+
+* handle OpenSSL::SSL::VERIFY_NONE returning 0 <jlindley>
+* doing Mail.new { content_type [text, plain, { charset => UTF-8 }] } is now 
+  possible (content type accepts an array) <mikel>
+
+== Sat Oct 31 11:00:41 UTC 2009 Mikel Lindsaar <raasdnil@gmail.com>
+
+* Fixed attachment handling, so mail can find attachment from a content-type,
+content-disposition or content-location
+* Added content-location field and parser
+* Added message.has_attachments? for ActionMailer friendliness
+* Added attachment.original_filename for ActionMailer friendliness
+
+== Sat Oct 25 13:38:01 UTC 2009 Mikel Lindsaar <raasdnil@gmail.com>
+
+* Birthday, Mail released as a gem... phew

data/Manifest.txt

@@ -1,7 +1,8 @@
 .gitignore
+CHANGELOG.rdoc
 Manifest.txt
-README.rdoc
 Rakefile
+README.rdoc
 lib/mail.rb
 lib/mail/attachment.rb
 lib/mail/body.rb
@@ -13,6 +14,7 @@ lib/mail/core_extensions.rb
 lib/mail/elements/address.rb
 lib/mail/elements/address_list.rb
 lib/mail/elements/content_disposition_element.rb
+lib/mail/elements/content_location_element.rb
 lib/mail/elements/content_transfer_encoding_element.rb
 lib/mail/elements/content_type_element.rb
 lib/mail/elements/date_time_element.rb
@@ -34,9 +36,11 @@ lib/mail/fields/common/common_address.rb
 lib/mail/fields/common/common_date.rb
 lib/mail/fields/common/common_field.rb
 lib/mail/fields/common/common_message_id.rb
+lib/mail/fields/common/parameter_hash.rb
 lib/mail/fields/content_description_field.rb
 lib/mail/fields/content_disposition_field.rb
 lib/mail/fields/content_id_field.rb
+lib/mail/fields/content_location_field.rb
 lib/mail/fields/content_transfer_encoding_field.rb
 lib/mail/fields/content_type_field.rb
 lib/mail/fields/date_field.rb
@@ -71,6 +75,8 @@ lib/mail/parsers/address_lists.rb
 lib/mail/parsers/address_lists.treetop
 lib/mail/parsers/content_disposition.rb
 lib/mail/parsers/content_disposition.treetop
+lib/mail/parsers/content_location.rb
+lib/mail/parsers/content_location.treetop
 lib/mail/parsers/content_transfer_encoding.rb
 lib/mail/parsers/content_transfer_encoding.treetop
 lib/mail/parsers/content_type.rb
@@ -98,9 +104,4 @@ lib/mail/patterns.rb
 lib/mail/utilities.rb
 lib/mail/version.rb
 lib/mail/version_specific/ruby_1_8.rb
-lib/mail/version_specific/ruby_1_8_string.rb
 lib/mail/version_specific/ruby_1_9.rb
-lib/mail/version_specific/multibyte.rb
-lib/mail/version_specific/multibyte/chars.rb
-lib/mail/version_specific/multibyte/exceptions.rb
-lib/mail/version_specific/multibyte/unicode_database.rb

data/README.rdoc

@@ -415,6 +415,25 @@ Of course... Mail will round trip an attachment as well
   @round_tripped_mail.attachments.length #=> 1
   @round_tripped_mail.attachments.first.filename #=> 'myfile.pdf'
 
+== Excerpts from TREC Spam Corpus 2005
+
+The spec fixture files in spec/fixtures/emails/from_trec_2005
+are from the 2005 TREC Public Spam Corpus. They remain copyrighted
+under the terms of that project and license agreement. They are used
+in this project to verify and describe the development of this
+email parser implementation.
+
+http://plg.uwaterloo.ca/~gvcormac/treccorpus/
+
+They are used as allowed by 'Permitted Uses, Clause 3':
+
+    "Small excerpts of the information may be displayed to others
+     or published in a scientific or technical context, solely for
+     the purpose of describing the research and development and
+     related issues."
+     
+     -- http://plg.uwaterloo.ca/~gvcormac/treccorpus/
+
 == License:
 
 (The MIT License)

data/Rakefile

@@ -36,3 +36,6 @@ Rake::RDocTask.new(:rdoc) do |rdoc|
   rdoc.rdoc_files.include('README')
   rdoc.rdoc_files.include('lib/**/*.rb')
 end
+
+# load custom rake tasks
+Dir["#{File.dirname(__FILE__)}/lib/tasks/**/*.rake"].sort.each { |ext| load ext }

data/lib/mail.rb

@@ -5,7 +5,9 @@ module Mail # :doc:
   
   gem "treetop", ">= 1.4"
   require 'treetop'
-
+  gem 'activesupport', ">= 2.3"
+  require 'activesupport'
+  require 'uri'
   require 'net/smtp'
   require 'mime/types'
   require 'tlsmail' if RUBY_VERSION <= '1.8.6'
@@ -17,7 +19,6 @@ module Mail # :doc:
     RubyVer = Mail::Ruby19
   else
     require File.join(dir_name, 'version_specific', 'ruby_1_8.rb')
-    require File.join(dir_name, 'version_specific', 'ruby_1_8_string.rb')
     RubyVer = Mail::Ruby18
   end
 
@@ -50,7 +51,7 @@ module Mail # :doc:
   parsers = %w[ rfc2822_obsolete rfc2822 address_lists phrase_lists
                 date_time received message_ids envelope_from rfc2045 
                 mime_version content_type content_disposition
-                content_transfer_encoding ]
+                content_transfer_encoding content_location ]
 
   parsers.each do |parser|
     begin

data/lib/mail/attachment.rb

@@ -49,7 +49,9 @@ module Mail
     def filename
       @filename
     end
-    
+
+    alias :original_filename :filename
+
     def encoded
       @encoded_data
     end
@@ -70,6 +72,12 @@ module Mail
     
     def set_mime_type(filename, mime_type)
       unless mime_type
+        # Have to do this because MIME::Types is not Ruby 1.9 safe yet
+        if RUBY_VERSION >= '1.9'
+          new_file = String.new(filename).force_encoding(Encoding::BINARY)
+          ext = new_file.split('.'.force_encoding(Encoding::BINARY)).last
+          filename = "file.#{ext}".force_encoding('US-ASCII')
+        end
         @mime_type = MIME::Types.type_for(filename).first
       else
         @mime_type = mime_type

data/lib/mail/body.rb

@@ -53,7 +53,7 @@ module Mail
     end
     
     alias :to_s :encoded
-
+    
     def charset
       @charset
     end
@@ -124,7 +124,7 @@ module Mail
       self.preamble = parts[0].to_s.strip
       # Make the epilogue equal to the epilogue (if any)
       self.epilogue = parts[-1].to_s.sub('--', '').strip
-      @parts = parts[1...-1].map { |part| Mail::Part.new(part) }
+      @parts = parts[1...-1].to_a.map { |part| Mail::Part.new(part) }
       self
     end
     

data/lib/mail/core_extensions/string.rb

@@ -1,14 +1,6 @@
 # encoding: utf-8
 class String #:nodoc:
   
-  if defined?(Mail::Multibyte)
-    include Mail::Multibyte
-  end
-  
-  def blank?
-    self !~ /\S/
-  end
-  
   def to_crlf
     self.gsub(/\n|\r\n|\r/) { "\r\n" }
   end

data/lib/mail/elements/content_location_element.rb

@@ -0,0 +1,25 @@
+# encoding: utf-8
+module Mail
+  class ContentLocationElement # :nodoc:
+    
+    include Mail::Utilities
+    
+    def initialize( string )
+      parser = Mail::ContentLocationParser.new
+      if tree = parser.parse(string)
+        @location = tree.location.text_value
+      else
+        raise Mail::Field::ParseError, "ContentLocationElement can not parse |#{string}|\nReason was: #{parser.failure_reason}\n"
+      end
+    end
+    
+    def location
+      @location
+    end
+    
+    def to_s(*args)
+      location.to_s
+    end
+    
+  end
+end

data/lib/mail/encodings/encodings.rb

@@ -20,5 +20,9 @@ module Mail
       RubyVer.q_encode(str, encoding)
     end
     
+    def Encodings.param_decode(str, encoding)
+      RubyVer.param_decode(str, encoding)
+    end
+    
   end
 end

data/lib/mail/field.rb

@@ -24,8 +24,8 @@ module Mail
     include Comparable
     
     STRUCTURED_FIELDS = %w[ bcc cc content-description content-disposition
-                            content-id content-transfer-encoding content-type 
-                            date from in-reply-to keywords message-id
+                            content-id content-location content-transfer-encoding
+                            content-type date from in-reply-to keywords message-id
                             mime-version received references reply-to
                             resent-bcc resent-cc resent-date resent-from
                             resent-message-id resent-sender resent-to
@@ -46,18 +46,33 @@ module Mail
     class SyntaxError < FieldError #:nodoc:
     end
     
-    # Accepts a text string in the format of:
+    # Accepts a string:
     # 
-    #  "field-name: field data"
+    #  Field.new("field-name: field data")
+    # 
+    # Or name, value pair:
+    # 
+    #  Field.new("field-name", "value")
+    # 
+    # Or a name by itself:
+    # 
+    #  Field.new("field-name")
     # 
     # Note, does not want a terminating carriage return.  Returns
-    # self appropriately parsed
-    def initialize(raw_field_text)
-      if raw_field_text !~ /:/
-        name = raw_field_text
+    # self appropriately parsed.  If value is not a string, then
+    # it will be passed through as is, for example, content-type
+    # field can accept an array with the type and a hash of 
+    # parameters:
+    # 
+    #  Field.new('content-type', ['text', 'plain', {:charset => 'UTF-8'}])
+    def initialize(name, value = nil)
+      case
+      when name =~ /:/ && value.blank?   # Field.new("field-name: field data")
+        name, value = split(name)
+        create_field(name, value)
+      when name !~ /:/ && value.blank?  # Field.new("field-name")
         create_field(name, nil)
-      else
-        name, value = split(raw_field_text)
+      else                              # Field.new("field-name", "value")
         create_field(name, value)
       end
       return self
@@ -112,7 +127,7 @@ module Mail
                       message-id in-reply-to references
                       subject comments keywords
                       mime-version content-type content-transfer-encoding
-                      content-disposition content-description ]
+                      content-location content-disposition content-description ]
     
     private
     
@@ -132,13 +147,13 @@ module Mail
     end
 
     def new_field(name, value)
-      # Could do this with constantize and make it as DRY as, but a simple case is,
-      # well, simpler... 
+      # Could do this with constantize and make it "as DRY as", but a simple case 
+      # statement is, well, simpler... 
       case name.downcase
       when /^to$/
-        ToField.new(name,value)
+        ToField.new(name, value)
       when /^cc$/
-        CcField.new(name,value)
+        CcField.new(name, value)
       when /^bcc$/
         BccField.new(name, value)
       when /^message-id$/
@@ -191,6 +206,8 @@ module Mail
         ContentTypeField.new(name, value)
       when /^content-id$/
         ContentIdField.new(name, value)
+      when /^content-location$/
+        ContentLocationField.new(name, value)
       else 
         OptionalField.new(name, value)
       end

data/lib/mail/fields/common/parameter_hash.rb

@@ -0,0 +1,31 @@
+# encoding: utf-8
+module Mail
+
+  # ParameterHash is an intelligent Hash that allows you to add
+  # parameter values including the mime extension paramaters that
+  # have the name*0="blah", name*1="bleh" keys, and will just return
+  # a single key called name="blahbleh" and do any required un-encoding
+  # to make that happen
+  class ParameterHash < Hash
+
+    include Enumerable
+
+    def [](key_name)
+      pairs = select { |k,v| k =~ /^#{key_name}\*/ }
+      pairs = pairs.to_a if RUBY_VERSION >= '1.9'
+      if pairs.empty? # Just dealing with a single value pair
+        super(key_name)
+      else # Dealing with a multiple value pair or a single encoded value pair
+        string = pairs.sort { |a,b| a.first <=> b.first }.map { |v| v.last }.join('')
+        if mt = string.match(/([\w\d\-]+)'(\w\w)'(.*)/)
+          string = mt[3]
+          encoding = mt[1]
+        else
+          encoding = nil
+        end
+        Mail::Encodings.param_decode(string, encoding)
+      end
+    end
+
+  end
+end

data/lib/mail/fields/content_disposition_field.rb

@@ -25,10 +25,22 @@ module Mail
     end
     
     def parameters
-      @parameters = Hash.new
+      @parameters = ParameterHash.new
       element.parameters.each { |p| @parameters.merge!(p) }
       @parameters
     end
 
+    def filename
+      case
+      when !parameters['filename'].blank?
+        @filename = parameters['filename']
+      when !parameters['name'].blank?
+        @filename = parameters['name']
+      else 
+        @filename = nil
+      end
+      @filename
+    end
+
   end
 end

data/lib/mail/fields/content_location_field.rb

@@ -0,0 +1,28 @@
+# encoding: utf-8
+# 
+# 
+# 
+module Mail
+  class ContentLocationField < StructuredField
+    
+    FIELD_NAME = 'content-location'
+
+    def initialize(*args)
+      super(FIELD_NAME, strip_field(FIELD_NAME, args.last))
+    end
+    
+    def tree
+      @element ||= Mail::ContentLocationElement.new(value)
+      @tree ||= @element.tree
+    end
+    
+    def element
+      @element ||= Mail::ContentLocationElement.new(value)
+    end
+
+    def location
+      element.location
+    end
+
+  end
+end

data/lib/mail/fields/content_type_field.rb

@@ -8,7 +8,17 @@ module Mail
     FIELD_NAME = 'content-type'
 
     def initialize(*args)
-      super(FIELD_NAME, strip_field(FIELD_NAME, args.last))
+      if args.last.class == Array
+        @main_type = args.last[0]
+        @sub_type = args.last[1]
+        @parameters = args.last.last
+        super(FIELD_NAME, args.last)
+      else
+        @main_type = nil
+        @sub_type = nil
+        @parameters = nil
+        super(FIELD_NAME, strip_field(FIELD_NAME, args.last))
+      end
     end
     
     def tree
@@ -21,11 +31,11 @@ module Mail
     end
     
     def main_type
-      element.main_type
+      @main_type ||= element.main_type
     end
 
     def sub_type
-      element.sub_type
+      @sub_type ||= element.sub_type
     end
     
     def content_type
@@ -33,8 +43,10 @@ module Mail
     end
     
     def parameters
-      @parameters = Hash.new
-      element.parameters.each { |p| @parameters.merge!(p) }
+      unless @parameters
+        @parameters = ParameterHash.new
+        element.parameters.each { |p| @parameters.merge!(p) }
+      end
       @parameters
     end
 
@@ -46,5 +58,29 @@ module Mail
       "--==_mimepart_#{Mail.random_tag}"
     end
 
+    def value
+      if @value.class == Array
+        "#{@main_type}/#{@sub_type}; #{stringify(parameters)}"
+      else
+        @value
+      end
+    end
+    
+    def stringify(params)
+      params.map { |k,v| "#{k}=#{v}" }.join("; ")
+    end
+
+    def filename
+      case
+      when parameters['filename']
+        @filename = parameters['filename']
+      when parameters['name']
+        @filename = parameters['name']
+      else 
+        @filename = nil
+      end
+      @filename
+    end
+
   end
 end