mailparser 0.5.2 → 0.5.3

checksums.yaml

@@ -0,0 +1,7 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: acab11e0f5b50260e17f32e1fea879d8505fff45
+  data.tar.gz: f0ce38c8bb40d11a3332515b4aabb7d835cb692f
+!binary "U0hBNTEy":
+  metadata.gz: 4edf18f3211be9fb2d2114db16265fd728a4f5a69eb1d96ab9a7fc430e0cdcc740f1d3b00ae04290a6203b611cfcd0b22531ac8c5962d7a513798bbdda57c53d
+  data.tar.gz: ee05e15d52db8517094e5c9189787c7e3f557a5c17c9d2dd8a8d3868111e671e086121296b1d13493e9ee6d43fc0ab6576fcd35bd3dda4b39dbfb91451904a95

data/lib/mailparser/rfc2047.rb

@@ -20,53 +20,32 @@ module MailParser::RFC2047
       charset = opt
     end
     charset_converter ||= MailParser::ConvCharset.method(:conv_charset)
-    last_charset = nil
     words = []
     mime_word = false
-    str.gsub(/\r?\n/, '').split(/(\s+)/).each do |s|
-      s, cs, raw, pre, post = decode_word(s)
-      if charset
+    ss = StringScanner.new(str.gsub(/\r?\n/, ''))
+    until ss.eos?
+      if s = ss.scan(/\=\?[^\(\)\<\>\@\,\;\:\"\/\[\]\?\.\=]+\?[QB]\?[^\? ]+\?\=/i)
         begin
-          s = charset_converter.call(cs || charset, charset, s)
+          s = Encoded.new s, charset, charset_converter
+          words.pop if words.length >= 2 and words[-1].is_a? Space and words[-2].is_a? Encoded
         rescue
-          s = raw
-          cs = nil
+          # ignore
         end
+        words.push s
+      elsif s = ss.scan(/\s+/)
+        words.push Space.new(s)
+      elsif s = ss.scan(/[^\s=]+/)
+        words.push s
+      else
+        words.push ss.scan(/./)
       end
-      if cs
-        words.pop if mime_word and words.last =~ /\A\s*\z/
-        mime_word = true
-      elsif s !~ /\A\s*\z/
-        mime_word = false
-      end
-      words.push pre if pre
-      words.push s
-      words.push post if post
     end
     begin
       ret = words.join
     rescue
-      ret = words.map{|s| s.force_encoding('binary')}.join
-    end
-    ret
-  end
-
-  def decode_word(str)
-    charset = nil
-    if str =~ /\=\?([^\(\)\<\>\@\,\;\:\"\/\[\]\?\.\=]+)\?([QB])\?([^\? ]+)\?\=/i
-      pre, post = $`, $'
-      charset, encoding, enc_text = $1.downcase, $2.downcase, $3
-      raw = str
-      str = encoding == "q" ? q_decode(enc_text) : b_decode(enc_text)
-      if String.method_defined? :force_encoding
-        begin
-          str.force_encoding(charset)
-        rescue
-          str.force_encoding('ascii-8bit')
-        end
-      end
+      ret = words.map{|s| s.to_s.force_encoding('binary')}.join
     end
-    [str, charset, raw, pre, post]
+    charset ? charset_converter.call(charset, charset, ret) : ret
   end
 
   def q_decode(str)
@@ -77,4 +56,18 @@ module MailParser::RFC2047
     return str.gsub(/[^A-Z0-9\+\/=]/i,"").unpack("m")[0]
   end
 
+  class Space < String
+  end
+
+  class Encoded
+    def initialize(str, charset, converter)
+      _, cs, encoding, enc_text, = str.split(/\?/)
+      str = encoding.downcase == 'q' ? MailParser::RFC2047.q_decode(enc_text) : MailParser::RFC2047.b_decode(enc_text)
+      @decoded = converter.call(cs, charset||cs, str)
+    end
+
+    def to_s
+      @decoded
+    end
+  end
 end

data/test/test_rfc2047.rb

@@ -23,7 +23,7 @@ class TC_RFC2047 < Test::Unit::TestCase
   end
 
   def test_q_decode_end_equal()
-    assert_equal("abc", MailParser::RFC2047.q_decode("abc="))
+    assert_match(/\Aabc=?\z/, MailParser::RFC2047.q_decode("abc="))
   end
 
   def test_b_decode_ascii()
@@ -38,20 +38,6 @@ class TC_RFC2047 < Test::Unit::TestCase
     assert_equal("とみた", MailParser::RFC2047.b_decode("44Go 44 G/4 4Gf"))
   end
 
-  def test_decode_word()
-    s, cs, raw = MailParser::RFC2047.decode_word("=?us-ascii?q?hoge?=")
-    assert_equal 'hoge', s
-    assert_equal 'us-ascii', cs
-    assert_equal '=?us-ascii?q?hoge?=', raw
-  end
-
-  def test_decode_word_upcase()
-    s, cs, raw = MailParser::RFC2047.decode_word("=?US-ASCII?Q?hoge?=")
-    assert_equal 'hoge', s
-    assert_equal 'us-ascii', cs
-    assert_equal '=?US-ASCII?Q?hoge?=', raw
-  end
-
   def test_decode_q_ascii()
     s = MailParser::RFC2047.decode("=?us-ascii?q?hoge?=")
     assert_equal("hoge", s)
@@ -154,4 +140,9 @@ class TC_RFC2047 < Test::Unit::TestCase
     assert_equal "abc defghijkl mno", s
   end
 
+  def test_decode_invalid_from_docomo
+    s = MailParser::RFC2047.decode("=?us-ascii?q?a?=(=?us-ascii?q?b?=)")
+    assert_equal "a(b)", s
+  end
+
 end

metadata

@@ -1,30 +1,27 @@
 --- !ruby/object:Gem::Specification
 name: mailparser
 version: !ruby/object:Gem::Version
-  version: 0.5.2
-  prerelease: 
+  version: 0.5.3
 platform: ruby
 authors:
 - TOMITA Masahiro
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-11-05 00:00:00.000000000 Z
+date: 2013-01-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mmapscanner
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 description: MailParser is a parser for mail message
@@ -60,27 +57,26 @@ files:
 homepage: http://github.com/tmtm/mailparser
 licenses:
 - Ruby's
+metadata: {}
 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: 
-rubygems_version: 1.8.23
+rubygems_version: 2.0.0.preview3.1
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Mail Parser
 test_files:
 - test.rb
@@ -91,4 +87,3 @@ test_files:
 - test/test_rfc2822.rb
 - test/test_mailparser.rb
 - test/test_rfc2045.rb
-has_rdoc: