new_rfc_2047 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: f4c99d89c3066d41ad5dc5e362d9ae7203ebe68483eb4e5acec0bd2146d1e64a
+  data.tar.gz: f088e8752bbb1b8c6fd06c2c4669896ef9a0236ca093e5be3a9fdc7f613d8a02
+SHA512:
+  metadata.gz: 146f547dedb2052365f618818528ff8a7799a1ba708f15c23ed62bd33c57dbe148ea7f971c8aeaff1a7b0fd98dfa6b8c9e67eaca41146045688d6a3679d67502
+  data.tar.gz: b02fd3440dd5c9e3ab867f128f524b18c6f72b25b6ec674a9ba9c8f82bfca4bdc79f9c97123d0b1c0f495f6994fe9ad7b78c727cfff991b0479b949e9ccffe04

data/lib/rfc_2047.rb

@@ -0,0 +1,97 @@
+# Copyright (c) 2020 Jian Weihang <tonytonyjan@gmail.com>
+# frozen_string_literal: true
+
+module Rfc2047
+  TOKEN = /[\041\043-\047\052\053\055\060-\071\101-\132\134\136\137\141-\176]+/.freeze
+  ENCODED_TEXT = /[\041-\076\100-\176]*/.freeze
+  ENCODED_WORD = /=\?(?<charset>#{TOKEN})\?(?<encoding>[QBqb])\?(?<encoded_text>#{ENCODED_TEXT})\?=/.freeze
+  ENCODED_WORD_SEQUENCE = /#{ENCODED_WORD}(?:\s*#{ENCODED_WORD})*/.freeze
+
+  class << self
+    # example:
+    #
+    #     Rfc2047.encode('己所不欲，勿施於人。')
+    #     # => "=?UTF-8?B?5bex5omA5LiN5qyy77yM5Yu/5pa95pa85Lq644CC?="
+    def encode(input, encoding: :B)
+      return input if input.ascii_only?
+
+      case encoding
+      when :B
+        size = 45
+        chunks = Array.new(((input.bytesize + size - 1) / size)) { input.byteslice(_1 * size, size) }
+        chunks.map! { "=?#{input.encoding}?B?#{[_1].pack('m0')}?=" }.join(' ')
+      when :Q
+        [input]
+          .pack('M').each_line
+          .map { "=?#{input.encoding}?Q?#{_1.chomp!.gsub(' ', '_')}?=" }
+          .join(' ')
+      else raise ":encoding should be either :B or :Q, got #{encoding}"
+      end
+    end
+
+    # example
+    #
+    #     Rfc2047.decode '=?UTF-8?B?5Yu/5Lul5oOh5bCP6ICM54K65LmL77yM5Yu/5Lul5ZaE5bCP6ICM5LiN54K6?= =?UTF-8?B?44CC?='
+    #     # => "勿以惡小而為之，勿以善小而不為。"
+    def decode(input)
+      return input unless input.match?(ENCODED_WORD)
+
+      input.gsub(ENCODED_WORD_SEQUENCE) do |match|
+        result = +''
+        match.scan(ENCODED_WORD) { result << decode_word($&) }
+        if result.encoding == Encoding::UTF_7
+          result.replace(
+            decode_utf7(result.force_encoding(Encoding::BINARY))
+          ).force_encoding(Encoding::UTF_8)
+        else
+          result.encode!(Encoding::UTF_8)
+        end
+        result
+      end
+    end
+
+    private
+
+    def decode_word(input)
+      match_data = ENCODED_WORD.match(input)
+      raise ArgumentError if match_data.nil?
+
+      charset, encoding, encoded_text = match_data.captures
+      charset = 'CP950' if charset == 'MS950'
+
+      decoded =
+        case encoding
+        when 'Q', 'q' then encoded_text.gsub('_', '=20').unpack1('M')
+        when 'B', 'b' then encoded_text.unpack1('m')
+        end
+      found_encoding = find_encoding(charset)
+      found_encoding = Encoding::UTF_8 if found_encoding == Encoding::ASCII_8BIT
+      decoded.force_encoding(found_encoding)
+    end
+
+    def find_encoding(charset)
+      case charset.downcase
+      when 'utf-16' then Encoding::UTF_16BE
+      when 'utf-32' then Encoding::UTF_32BE
+      when 'ks_c_5601-1987' then Encoding::CP949
+      when 'shift-jis' then Encoding::Shift_JIS
+      when 'gb2312' then Encoding::GB18030
+      when 'ms950' then Encoding::CP950
+      when '8bit' then Encoding::ASCII_8BIT
+      when 'latin2' then Encoding::ISO_8859_2
+      else Encoding.find(charset)
+      end
+    end
+
+    # from Net::IMAP
+    def decode_utf7(s)
+      s.gsub(/&([^-]+)?-/n) do
+        if Regexp.last_match(1)
+          (Regexp.last_match(1).tr(',', '/') + '===').unpack1('m').encode(Encoding::UTF_8, Encoding::UTF_16BE)
+        else
+          '&'
+        end
+      end
+    end
+  end
+end

metadata

@@ -0,0 +1,71 @@
+--- !ruby/object:Gem::Specification
+name: new_rfc_2047
+version: !ruby/object:Gem::Version
+  version: 1.0.0
+platform: ruby
+authors:
+- Jian Weihang
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2020-10-18 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '13'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '13'
+description: An implementation of RFC 2047
+email: tonytonyjan@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/rfc_2047.rb
+homepage: https://github.com/tonytonyjan/rfc_2047
+licenses:
+- MIT
+metadata: {}
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.1.2
+signing_key:
+specification_version: 4
+summary: An implementation of RFC 2047
+test_files: []