mail 2.6.3 → 2.7.0

data/lib/mail/elements/received_element.rb

@@ -1,26 +1,21 @@
 # encoding: utf-8
+# frozen_string_literal: true
+require 'mail/parsers/received_parser'
+require 'date'
+
 module Mail
   class ReceivedElement
-    
     include Mail::Utilities
-    
-    def initialize( string )
-      received = Mail::Parsers::ReceivedParser.new.parse(string)
+    attr_reader :date_time, :info
+
+    def initialize(string)
+      received = Mail::Parsers::ReceivedParser.parse(string)
       @date_time = ::DateTime.parse("#{received.date} #{received.time}")
       @info = received.info
     end
-    
-    def date_time
-      @date_time
-    end
-    
-    def info
-      @info
-    end
-    
+
     def to_s(*args)
-      "#{@info}; #{@date_time.to_s(*args)}"
+      "#{info}; #{date_time.to_s(*args)}"
     end
-    
   end
 end

data/lib/mail/elements.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Mail
   register_autoload :Address, 'mail/elements/address'
   register_autoload :AddressList, 'mail/elements/address_list'

data/lib/mail/encodings/7bit.rb

@@ -1,31 +1,22 @@
 # encoding: utf-8
+# frozen_string_literal: true
 require 'mail/encodings/8bit'
 
 module Mail
   module Encodings
+    # 7bit and 8bit are equivalent. 7bit encoding is for text only.
     class SevenBit < EightBit
       NAME = '7bit'
-    
       PRIORITY = 1
+      Encodings.register(NAME, self)
 
-      # 7bit and 8bit operate the same
-      
-      # Decode the string
       def self.decode(str)
-        super
+        ::Mail::Utilities.binary_unsafe_to_lf str
       end
-    
-      # Encode the string
+
       def self.encode(str)
-        super
-      end
-     
-      # Idenity encodings have a fixed cost, 1 byte out per 1 byte in
-      def self.cost(str)
-        super 
+        ::Mail::Utilities.binary_unsafe_to_crlf str
       end
-
-      Encodings.register(NAME, self) 
     end
   end
 end

data/lib/mail/encodings/8bit.rb

@@ -1,31 +1,18 @@
 # encoding: utf-8
+# frozen_string_literal: true
 require 'mail/encodings/binary'
 
 module Mail
   module Encodings
     class EightBit < Binary
       NAME = '8bit'
-
       PRIORITY = 4
+      Encodings.register(NAME, self)
 
-      # 8bit is an identiy encoding, meaning nothing to do
-      
-      # Decode the string
-      def self.decode(str)
-        str.to_lf
-      end
-    
-      # Encode the string
-      def self.encode(str)
-        str.to_crlf
+      # Per RFC 2821 4.5.3.1, SMTP lines may not be longer than 1000 octets including the <CRLF>.
+      def self.compatible_input?(str)
+        !str.lines.find { |line| line.bytesize > 998 }
       end
-     
-      # Idenity encodings have a fixed cost, 1 byte out per 1 byte in
-      def self.cost(str)
-        1.0
-      end
-
-      Encodings.register(NAME, self) 
     end
   end
 end

data/lib/mail/encodings/base64.rb

@@ -1,33 +1,38 @@
 # encoding: utf-8
+# frozen_string_literal: true
 require 'mail/encodings/7bit'
 
 module Mail
   module Encodings
+    # Base64 encoding handles binary content at the cost of 4 output bytes
+    # per input byte.
     class Base64 < SevenBit
       NAME = 'base64'
-     
       PRIORITY = 3
- 
+      Encodings.register(NAME, self)
+
       def self.can_encode?(enc)
         true
       end
 
-      # Decode the string from Base64
       def self.decode(str)
-        RubyVer.decode_base64( str )
+        RubyVer.decode_base64(str)
       end
-    
-      # Encode the string to Base64
+
       def self.encode(str)
-        RubyVer.encode_base64( str ).to_crlf
+        ::Mail::Utilities.binary_unsafe_to_crlf(RubyVer.encode_base64(str))
       end
 
-      # Base64 has a fixed cost, 4 bytes out per 3 bytes in
+      # 3 bytes in -> 4 bytes out
       def self.cost(str)
-        4.0/3
+        4.0 / 3
       end
 
-      Encodings.register(NAME, self)      
+      # Ruby Base64 inserts newlines automatically, so it doesn't exceed
+      # SMTP line length limits.
+      def self.compatible_input?(str)
+        true
+      end
     end
   end
 end

data/lib/mail/encodings/binary.rb

@@ -1,31 +1,13 @@
 # encoding: utf-8
-require 'mail/encodings/transfer_encoding'
+# frozen_string_literal: true
+require 'mail/encodings/identity'
 
 module Mail
   module Encodings
-    class Binary < TransferEncoding
+    class Binary < Identity
       NAME = 'binary'
-
       PRIORITY = 5
-
-      # Binary is an identiy encoding, meaning nothing to do
-      
-      # Decode the string
-      def self.decode(str)
-        str
-      end
-    
-      # Encode the string
-      def self.encode(str)
-        str
-      end
-     
-      # Idenity encodings have a fixed cost, 1 byte out per 1 byte in
-      def self.cost(str)
-        1.0
-      end
-
-      Encodings.register(NAME, self) 
+      Encodings.register(NAME, self)
     end
   end
 end

data/lib/mail/encodings/identity.rb

@@ -0,0 +1,24 @@
+# encoding: utf-8
+# frozen_string_literal: true
+require 'mail/encodings/transfer_encoding'
+
+module Mail
+  module Encodings
+    # Identity encodings do no encoding/decoding and have a fixed cost:
+    # 1 byte in -> 1 byte out.
+    class Identity < TransferEncoding #:nodoc:
+      def self.decode(str)
+        str
+      end
+
+      def self.encode(str)
+        str
+      end
+
+      # 1 output byte per input byte.
+      def self.cost(str)
+        1.0
+      end
+    end
+  end
+end

data/lib/mail/encodings/quoted_printable.rb

@@ -1,36 +1,42 @@
 # encoding: utf-8
+# frozen_string_literal: true
 require 'mail/encodings/7bit'
 
 module Mail
   module Encodings
     class QuotedPrintable < SevenBit
       NAME='quoted-printable'
-   
+
       PRIORITY = 2
 
-      def self.can_encode?(str)
-        EightBit.can_encode? str
+      def self.can_encode?(enc)
+        EightBit.can_encode? enc
       end
 
       # Decode the string from Quoted-Printable. Cope with hard line breaks
       # that were incorrectly encoded as hex instead of literal CRLF.
       def self.decode(str)
-        str.gsub(/(?:=0D=0A|=0D|=0A)\r\n/, "\r\n").unpack("M*").first.to_lf
+        str.gsub(/(?:=0D=0A|=0D|=0A)\r\n/, "\r\n").unpack("M*").first
       end
 
       def self.encode(str)
-        [str.to_lf].pack("M").to_crlf
+        [str].pack("M")
       end
 
       def self.cost(str)
         # These bytes probably do not need encoding
         c = str.count("\x9\xA\xD\x20-\x3C\x3E-\x7E")
-        # Everything else turns into =XX where XX is a 
+        # Everything else turns into =XX where XX is a
         # two digit hex number (taking 3 bytes)
         total = (str.bytesize - c)*3 + c
         total.to_f/str.bytesize
       end
-        
+
+      # QP inserts newlines automatically and cannot violate the SMTP spec.
+      def self.compatible_input?(str)
+        true
+      end
+
       private
 
       Encodings.register(NAME, self)

data/lib/mail/encodings/transfer_encoding.rb

@@ -1,4 +1,5 @@
 # encoding: utf-8
+# frozen_string_literal: true
 module Mail
   module Encodings
     class TransferEncoding
@@ -6,52 +7,70 @@ module Mail
 
       PRIORITY = -1
 
+      # And encoding's superclass can always transport it since the
+      # class hierarchy is arranged e.g. Base64 < 7bit < 8bit < Binary.
       def self.can_transport?(enc)
-        enc = Encodings.get_name(enc)
-        if Encodings.defined? enc
-          Encodings.get_encoding(enc).new.is_a? self
-        else
-          false
-        end
+        enc && enc <= self
       end
 
+      # Override in subclasses to indicate that they can encode text
+      # that couldn't be directly transported, e.g. Base64 has 7bit output,
+      # but it can encode binary.
       def self.can_encode?(enc)
-        can_transport? enc 
+        can_transport? enc
       end
 
       def self.cost(str)
         raise "Unimplemented"
       end
 
+      def self.compatible_input?(str)
+        true
+      end
+
       def self.to_s
         self::NAME
       end
 
-      def self.get_best_compatible(source_encoding, str)
-        if self.can_transport? source_encoding then
-            source_encoding 
+      def self.negotiate(message_encoding, source_encoding, str, allowed_encodings = nil)
+        message_encoding = Encodings.get_encoding(message_encoding || '8bit')
+        source_encoding  = Encodings.get_encoding(source_encoding)
+
+        if message_encoding && source_encoding && message_encoding.can_transport?(source_encoding) && source_encoding.compatible_input?(str)
+          source_encoding
         else
-            choices = []
-            Encodings.get_all.each do |enc|
-                choices << enc if self.can_transport? enc and enc.can_encode? source_encoding
-            end
-            best = nil 
-            best_cost = 100
-            choices.each do |enc|
-                this_cost = enc.cost str
-                if this_cost < best_cost then
-                    best_cost = this_cost
-                    best = enc
-                elsif this_cost == best_cost then 
-                    best = enc if enc::PRIORITY < best::PRIORITY
-                end
-            end
-            best
+          renegotiate(message_encoding, source_encoding, str, allowed_encodings)
         end
       end
 
-      def to_s
-        self.class.to_s
+      def self.renegotiate(message_encoding, source_encoding, str, allowed_encodings = nil)
+        encodings = Encodings.get_all.select do |enc|
+          (allowed_encodings.nil? || allowed_encodings.include?(enc)) &&
+            message_encoding.can_transport?(enc) &&
+            enc.can_encode?(source_encoding)
+        end
+
+        lowest_cost(str, encodings)
+      end
+
+      def self.lowest_cost(str, encodings)
+        best = nil
+        best_cost = nil
+
+        encodings.each do |enc|
+          # If the current choice cannot be transported safely, give priority
+          # to other choices but allow it to be used as a fallback.
+          this_cost = enc.cost(str) if enc.compatible_input?(str)
+
+          if !best_cost || (this_cost && this_cost < best_cost)
+            best_cost = this_cost
+            best = enc
+          elsif this_cost == best_cost
+            best = enc if enc::PRIORITY < best::PRIORITY
+          end
+        end
+
+        best
       end
     end
   end

data/lib/mail/encodings/unix_to_unix.rb

@@ -1,6 +1,7 @@
+# frozen_string_literal: true
 module Mail
   module Encodings
-    module UnixToUnix
+    class UnixToUnix < TransferEncoding
       NAME = "x-uuencode"
 
       def self.decode(str)
@@ -12,6 +13,7 @@ module Mail
       end
 
       Encodings.register(NAME, self)
+      Encodings.register("uuencode", self)
     end
   end
 end

data/lib/mail/encodings.rb

@@ -1,4 +1,5 @@
 # encoding: utf-8
+# frozen_string_literal: true
 
 module Mail
   # Raised when attempting to decode an unknown encoding type
@@ -6,7 +7,6 @@ module Mail
   end
 
   module Encodings
-
     include Mail::Constants
     extend  Mail::Utilities
 
@@ -18,7 +18,7 @@ module Mail
     #
     # Encodings.register "base64", Mail::Encodings::Base64
     def Encodings.register(name, cls)
-        @transfer_encodings[get_name(name)] = cls
+      @transfer_encodings[get_name(name)] = cls
     end
 
     # Is the encoding we want defined?
@@ -26,8 +26,8 @@ module Mail
     # Example:
     #
     #  Encodings.defined?(:base64) #=> true
-    def Encodings.defined?( str )
-      @transfer_encodings.include? get_name(str)
+    def Encodings.defined?(name)
+      @transfer_encodings.include? get_name(name)
     end
 
     # Gets a defined encoding type, QuotedPrintable or Base64 for now.
@@ -38,16 +38,24 @@ module Mail
     # Example:
     #
     #  Encodings.get_encoding(:base64) #=> Mail::Encodings::Base64
-    def Encodings.get_encoding( str )
-      @transfer_encodings[get_name(str)]
+    def Encodings.get_encoding(name)
+      @transfer_encodings[get_name(name)]
     end
 
     def Encodings.get_all
       @transfer_encodings.values
     end
 
-    def Encodings.get_name(enc)
-      enc = underscoreize(enc).downcase
+    def Encodings.get_name(name)
+      underscoreize(name).downcase
+    end
+
+    def Encodings.transcode_charset(str, from_charset, to_charset = 'UTF-8')
+      if from_charset
+        RubyVer.transcode_charset str, from_charset, to_charset
+      else
+        str
+      end
     end
 
     # Encodes a parameter value using URI Escaping, note the language field 'en' can
@@ -121,7 +129,7 @@ module Mail
       # Split on white-space boundaries with capture, so we capture the white-space as well
       lines.each do |line|
         line.gsub!(ENCODED_VALUE) do |string|
-          case $1
+          case $2
           when *B_VALUES then b_value_decode(string)
           when *Q_VALUES then q_value_decode(string)
           end
@@ -159,32 +167,51 @@ module Mail
 
     def Encodings.address_encode(address, charset = 'utf-8')
       if address.is_a?(Array)
-        # loop back through for each element
         address.compact.map { |a| Encodings.address_encode(a, charset) }.join(", ")
-      else
-        # find any word boundary that is not ascii and encode it
-        encode_non_usascii(address, charset) if address
+      elsif address
+        encode_non_usascii(address, charset)
       end
     end
 
     def Encodings.encode_non_usascii(address, charset)
       return address if address.ascii_only? or charset.nil?
-      us_ascii = %Q{\x00-\x7f}
-      # Encode any non usascii strings embedded inside of quotes
-      address = address.gsub(/(".*?[^#{us_ascii}].*?")/) { |s| Encodings.b_value_encode(unquote(s), charset) }
-      # Then loop through all remaining items and encode as needed
-      tokens = address.split(/\s/)
-      map_with_index(tokens) do |word, i|
-        if word.ascii_only?
-          word
-        else
-          previous_non_ascii = i>0 && tokens[i-1] && !tokens[i-1].ascii_only?
-          if previous_non_ascii #why are we adding an extra space here?
-            word = " #{word}"
+
+      # With KCODE=u we can't use regexps on other encodings. Go ASCII.
+      with_ascii_kcode do
+        # Encode all strings embedded inside of quotes
+        address = address.gsub(/("[^"]*[^\/]")/) { |s| Encodings.b_value_encode(unquote(s), charset) }
+
+        # Then loop through all remaining items and encode as needed
+        tokens = address.split(/\s/)
+
+        map_with_index(tokens) do |word, i|
+          if word.ascii_only?
+            word
+          else
+            previous_non_ascii = i>0 && tokens[i-1] && !tokens[i-1].ascii_only?
+            if previous_non_ascii #why are we adding an extra space here?
+              word = " #{word}"
+            end
+            Encodings.b_value_encode(word, charset)
           end
-          Encodings.b_value_encode(word, charset)
+        end.join(' ')
+      end
+    end
+
+    if RUBY_VERSION < '1.9'
+      # With KCODE=u we can't use regexps on other encodings. Go ASCII.
+      def Encodings.with_ascii_kcode #:nodoc:
+        if $KCODE
+          $KCODE, original_kcode = '', $KCODE
         end
-      end.join(' ')
+        yield
+      ensure
+        $KCODE = original_kcode if original_kcode
+      end
+    else
+      def Encodings.with_ascii_kcode #:nodoc:
+        yield
+      end
     end
 
     # Encode a string with Base64 Encoding and returns it ready to be inserted
@@ -194,12 +221,15 @@ module Mail
     #
     #  Encodings.b_value_encode('This is あ string', 'UTF-8')
     #  #=> "=?UTF-8?B?VGhpcyBpcyDjgYIgc3RyaW5n?="
-    def Encodings.b_value_encode(encoded_str, encoding = nil)
-      return encoded_str if encoded_str.to_s.ascii_only?
-      string, encoding = RubyVer.b_value_encode(encoded_str, encoding)
-      map_lines(string) do |str|
-        "=?#{encoding}?B?#{str.chomp}?="
-      end.join(" ")
+    def Encodings.b_value_encode(string, encoding = nil)
+      if string.to_s.ascii_only?
+        string
+      else
+        Encodings.each_base64_chunk_byterange(string, 60).map do |chunk|
+          str, encoding = RubyVer.b_value_encode(chunk, encoding)
+          "=?#{encoding}?B?#{str.chomp}?="
+        end.join(" ")
+      end
     end
 
     # Encode a string with Quoted-Printable Encoding and returns it ready to be inserted
@@ -240,50 +270,74 @@ module Mail
       RubyVer.q_value_decode(str)
     end
 
-    def Encodings.split_encoding_from_string( str )
-      match = str.match(/\=\?([^?]+)?\?[QB]\?(.*)\?\=/mi)
-      if match
-        match[1]
-      else
-        nil
-      end
-    end
-
     def Encodings.find_encoding(str)
       RUBY_VERSION >= '1.9' ? str.encoding : $KCODE
     end
 
     # Gets the encoding type (Q or B) from the string.
-    def Encodings.split_value_encoding_from_string(str)
-      match = str.match(/\=\?[^?]+?\?([QB])\?(.*)\?\=/mi)
-      if match
-        match[1]
-      else
-        nil
-      end
+    def Encodings.value_encoding_from_string(str)
+      str[ENCODED_VALUE, 1]
     end
 
-    # When the encoded string consists of multiple lines, lines with the same
-    # encoding (Q or B) can be joined together.
+    # Split header line into proper encoded and unencoded parts.
     #
     # String has to be of the format =?<encoding>?[QB]?<string>?=
+    #
+    # Omit unencoded space after an encoded-word.
     def Encodings.collapse_adjacent_encodings(str)
-      lines = str.split(/(\?=)\s*(=\?)/).each_slice(2).map(&:join)
       results = []
-      previous_encoding = nil
+      last_encoded = nil  # Track whether to preserve or drop whitespace
 
-      lines.each do |line|
-        encoding = split_value_encoding_from_string(line)
+      lines = str.split(FULL_ENCODED_VALUE)
+      lines.each_slice(2) do |unencoded, encoded|
+        if last_encoded = encoded
+          if !Utilities.blank?(unencoded) || (!last_encoded && unencoded != EMPTY)
+            results << unencoded
+          end
 
-        if encoding == previous_encoding
-          line = results.pop + line
+          results << encoded
+        else
+          results << unencoded
         end
-
-        previous_encoding = encoding
-        results << line
       end
 
       results
     end
+
+    # Partition the string into bounded-size chunks without splitting
+    # multibyte characters.
+    def Encodings.each_base64_chunk_byterange(str, max_bytesize_per_base64_chunk, &block)
+      raise "size per chunk must be multiple of 4" if (max_bytesize_per_base64_chunk % 4).nonzero?
+
+      if block_given?
+        max_bytesize = ((3 * max_bytesize_per_base64_chunk) / 4.0).floor
+        each_chunk_byterange(str, max_bytesize, &block)
+      else
+        enum_for :each_base64_chunk_byterange, str, max_bytesize_per_base64_chunk
+      end
+    end
+
+    # Partition the string into bounded-size chunks without splitting
+    # multibyte characters.
+    def Encodings.each_chunk_byterange(str, max_bytesize_per_chunk)
+      return enum_for(:each_chunk_byterange, str, max_bytesize_per_chunk) unless block_given?
+
+      offset = 0
+      chunksize = 0
+
+      str.each_char do |chr|
+        charsize = chr.bytesize
+
+        if chunksize + charsize > max_bytesize_per_chunk
+          yield RubyVer.string_byteslice(str, offset, chunksize)
+          offset += chunksize
+          chunksize = charsize
+        else
+          chunksize += charsize
+        end
+      end
+
+      yield RubyVer.string_byteslice(str, offset, chunksize)
+    end
   end
 end

data/lib/mail/envelope.rb

@@ -1,4 +1,5 @@
 # encoding: utf-8
+# frozen_string_literal: true
 # 
 # = Mail Envelope
 # 
@@ -11,7 +12,7 @@ module Mail
   class Envelope < StructuredField
     
     def initialize(*args)
-      super(FIELD_NAME, strip_field(FIELD_NAME, args.last))
+      super(FIELD_NAME, args.last.to_s)
     end
     
     def element