sisimai 4.16.0-java

data/Rakefile

@@ -0,0 +1,12 @@
+require 'rspec/core/rake_task'
+require 'bundler/gem_helper'
+if RUBY_PLATFORM =~ /java/
+  filename = 'sisimai-java'
+else
+  filename = 'sisimai'
+end
+Bundler::GemHelper.install_tasks :name => filename
+RSpec::Core::RakeTask.new(:spec)
+
+task :default => :spec
+

data/Repository.mk

@@ -0,0 +1,69 @@
+# rb-Sisimai/Repository.mk
+#  ____                      _ _                               _    
+# |  _ \ ___ _ __   ___  ___(_) |_ ___  _ __ _   _   _ __ ___ | | __
+# | |_) / _ \ '_ \ / _ \/ __| | __/ _ \| '__| | | | | '_ ` _ \| |/ /
+# |  _ <  __/ |_) | (_) \__ \ | || (_) | |  | |_| |_| | | | | |   < 
+# |_| \_\___| .__/ \___/|___/_|\__\___/|_|   \__, (_)_| |_| |_|_|\_\
+#           |_|                              |___/                  
+# -----------------------------------------------------------------------------
+SHELL := /bin/sh
+GIT   := /usr/bin/git
+CP    := cp
+B      = master
+V      = neko
+RSYNC  = rsync -var
+EMAILS = set-of-emails
+
+
+.DEFAULT_GOAL = git-status
+
+# -----------------------------------------------------------------------------
+.PHONY: clean
+
+git-status:
+	$(GIT) status
+
+git-push:
+	@ for v in `$(GIT) remote show | grep -v origin`; do \
+		printf "[%s]\n" $$v; \
+		$(GIT) push --tags $$v $(B); \
+	done
+
+git-tag-list:
+	$(GIT) tag -l
+
+git-diff:
+	$(GIT) diff -w
+
+git-branch:
+	$(GIT) branch -a
+
+git-commit-amend:
+	$(GIT) commit --amend
+
+git-follow-log:
+	$(GIT) log --follow -p $(V) || \
+		printf "\nUsage:\n %% make -f Repository.mk $@ V=<filename>\n"
+
+git-branch-tree:
+	$(GIT) log --graph \
+		--pretty='format:%C(yellow)%h%Creset %s %Cgreen(%an)%Creset %Cred%d%Creset'
+
+git-rm-cached:
+	$(GIT) rm -f --cached $(V) || \
+		printf "\nUsage:\n %% make -f Repository.mk $@ V=<filename>\n"
+
+git-reset-soft:
+	$(GIT) reset --soft HEAD^
+
+import-set-of-emails:
+	test -d $(EMAILS)
+	$(CP) -vRp ../$(EMAILS)/README.md $(EMAILS)/
+	$(RSYNC) ../$(EMAILS)/mailbox/ ./$(EMAILS)/mailbox/
+	$(RSYNC) ../$(EMAILS)/maildir/ ./$(EMAILS)/maildir/
+	$(RSYNC) ../$(EMAILS)/private/ ./$(EMAILS)/private/
+	$(RSYNC) ../$(EMAILS)/to-be-debugged-because/ ./$(EMAILS)/to-be-debugged-because/
+	$(RSYNC) ../$(EMAILS)/to-be-parsed-for-test/  ./$(EMAILS)/to-be-parsed-for-test/
+
+clean:
+

data/appveyor.yml

@@ -0,0 +1,23 @@
+version: '{build}'
+
+install:
+  - SET PATH=C:\Ruby%ruby_version%\bin;%PATH%
+  - "%devkit%\\devkitvars.bat"
+  - ruby --version
+  - gem --version
+  - bundle install
+build: off
+test_script:
+  - bundle exec rake spec TESTOPTS=-v
+
+environment:
+  matrix:
+    - ruby_version: "22-x64"
+      devkit: C:\Ruby21-x64\DevKit
+    - ruby_version: "22"
+      devkit: C:\Ruby21\DevKit
+    - ruby_version: "21-x64"
+      devkit: C:\Ruby21-x64\DevKit
+    - ruby_version: "21"
+      devkit: C:\Ruby21\DevKit
+

data/bin/console

@@ -0,0 +1,14 @@
+#!/usr/bin/env ruby
+
+require "bundler/setup"
+require "sisimai"
+
+# You can add fixtures and/or initialization code here to make experimenting
+# with your gem easier. You can also use a different console, if you like.
+
+# (If you use this, don't forget to add pry to your Gemfile!)
+# require "pry"
+# Pry.start
+
+require "irb"
+IRB.start

data/bin/setup

@@ -0,0 +1,7 @@
+#!/bin/bash
+set -euo pipefail
+IFS=$'\n\t'
+
+bundle install
+
+# Do any other automated setup that you need to do here

data/lib/sisimai.rb

@@ -0,0 +1,95 @@
+require 'sisimai/version'
+
+# Sisimai is the system formerly known as bounceHammer 4, is a Ruby module for
+# analyzing bounce mails and generate structured data in a JSON format (YAML is
+# also available if "YAML" module is installed on your system) from parsed bounce
+# messages. Sisimai is a coined word: Sisi (the number 4 is pronounced "Si" in
+# Japanese) and MAI (acronym of "Mail Analyzing Interface").
+module Sisimai
+  # Imported from p5-Sisimail/lib/Sisimai.pm
+  class << self
+    def version(); return Sisimai::VERSION; end
+    def sysname(); return 'bouncehammer';   end
+    def libname(); return 'Sisimai';        end
+
+    # Wrapper method for parsing mailbox/maidir
+    # @param         [String] path       Path to mbox or Maildir/
+    # @param         [Hash]  argvs       Parser options(delivered=false)
+    # @options argvs [Boolean] delivered true: Include "delivered" reason
+    # @return        [Array]             Parsed objects
+    # @return        [nil]               nil if the argument was wrong or an empty array
+    def make(path, **argvs)
+      return nil unless path
+
+      require 'sisimai/mail'
+      mail = Sisimai::Mail.new(path)
+      list = []
+      opts = argvs[:delivered] || false
+
+      return nil unless mail
+      require 'sisimai/data'
+      require 'sisimai/message'
+
+      while r = mail.read do
+        # Read and parse each mail file
+        mesg = Sisimai::Message.new(data: r)
+        next if mesg.void
+        data = Sisimai::Data.make(data: mesg, delivered: opts)
+        next unless data
+        list.concat(data) if data.size > 0
+      end
+
+      return nil if list.size == 0
+      return list
+    end
+
+    # Wrapper method to parse mailbox/Maildir and dump as JSON
+    # @param         [String] path       Path to mbox or Maildir/
+    # @param         [Hash] argvs        Parser options
+    # @options argvs [Integer] delivered true: Include "delivered" reason
+    # @return        [String]            Parsed data as JSON text
+    def dump(path, **argvs)
+      return nil unless path
+
+      parseddata = Sisimai.make(path, argvs) || []
+      if RUBY_PLATFORM =~ /java/
+        # java-based ruby environment like JRuby.
+        require 'jrjackson'
+        jsonstring = JrJackson::Json.dump(parseddata)
+      else
+        require 'oj'
+        jsonstring = Oj.dump(parseddata, :mode => :compat)
+      end
+      return jsonstring
+    end
+
+    # Parser engine list (MTA/MSP modules)
+    # @return   [Hash]     Parser engine table
+    def engine
+      names = %w|MTA MSP ARF RFC3464 RFC3834|
+      table = {}
+
+      names.each do |e|
+        r = 'Sisimai::' + e
+        require r.gsub('::', '/').downcase
+
+        if e == 'MTA' || e == 'MSP'
+          # Sisimai::MTA or Sisimai::MSP
+          Module.const_get(r).send(:index).each do |ee|
+            # Load and get the value of "description" from each module
+            rr = sprintf('Sisimai::%s::%s', e, ee)
+            require rr.gsub('::', '/').downcase
+            table[rr.to_sym] = Module.const_get(rr).send(:description)
+          end
+        else
+          # Sisimai::ARF, Sisimai::RFC3464, and Sisimai::RFC3834
+          table[r.to_sym] = Module.const_get(r).send(:description)
+        end
+      end
+
+      return table
+    end
+
+  end
+
+end

data/lib/sisimai/address.rb

@@ -0,0 +1,224 @@
+module Sisimai
+  # Sisimai::Address provide methods for dealing email address.
+  class Address
+    # Imported from p5-Sisimail/lib/Sisimai/Address.pm
+    require 'sisimai/rfc5322'
+
+    # Return pseudo recipient or sender address
+    # @param    [String] argv1  Address type: 'r' or 's'
+    # @return   [String, Nil]   Pseudo recipient address or sender address or
+    #                           Undef when the $argv1 is neither 'r' nor 's'
+    def self.undisclosed(argv1)
+      return nil unless argv1
+      return nil unless %w|r s|.index(argv1)
+
+      local = argv1 == 'r' ? 'recipient' : 'sender'
+      return sprintf('undisclosed-%s-in-headers@dummy-domain.invalid', local)
+    end
+
+
+    # Email address parser
+    # @param    [Array] argvs   List of strings including email address
+    # @return   [Array, Nil]    Email address list or Undef when there is no
+    #                           email address in the argument
+    # @example  Parse email address
+    #   parse( [ 'Neko <neko@example.cat>' ] )  #=> [ 'neko@example.cat' ]
+    def self.parse(argvs)
+      addrs = []
+      argvs.each do |e|
+        # Parse each element in the array
+        #   1. The element must include '@'.
+        #   2. The element must not include character except from 0x20 to 0x7e.
+        next unless e
+        next unless e =~ /[@]/
+        next if e =~ /[^\x20-\x7e]/
+
+        v = Sisimai::Address.s3s4(e)
+        if v.size > 0
+          # The element includes a valid email address
+          addrs << v
+        end
+      end
+
+      return nil unless addrs.size > 0
+      return addrs
+    end
+
+    # Runs like ruleset 3,4 of sendmail.cf
+    # @param    [String] email  Text including an email address
+    # @return   [String]        Email address without comment, brackets
+    # @example  Parse email address
+    #   s3s4( '<neko@example.cat>' ) #=> 'neko@example.cat'
+    def self.s3s4(input)
+      unless input =~ /[ ]/
+        # no space character between " and < .
+        # no space character between " and < .
+        input = input.sub(/\A(.+)"<(.+)\z/, '\1" <\2')      # "=?ISO-2022-JP?B?....?="<user@example.jp>,
+        input = input.sub(/\A(.+)[?]=<(.+)\z/, '\1?= <\2')  # =?ISO-2022-JP?B?....?=<user@example.jp>
+
+        # comment-part<localpart@domainpart>
+        input = input.sub(/[<]/, ' <') unless input =~ /\A[<]/
+        input = input.sub(/[>]/, '> ') unless input =~ /[>]\z/
+      end
+
+      canon = ''
+      addrs = []
+      token = input.split(' ')
+
+      token.map! do |e|
+        # Convert character entity; "&lt;" -> ">", "&gt;" -> "<".
+        e = e.gsub(/&lt;/, '<')
+        e = e.gsub(/&gt;/, '>')
+        e = e.gsub(/,\z/, '')
+      end
+
+      if token.size == 1
+        addrs << token[0]
+
+      else
+        token.each do |e|
+          e.chomp
+          next unless e =~ /\A[<]?.+[@][-.0-9A-Za-z]+[.]?[A-Za-z]{2,}[>]?\z/
+          addrs << e
+        end
+      end
+
+      if addrs.size > 1
+        # Get the first element which is <...> format string from @addrs array.
+        canon = addrs.detect { |e| e =~ /\A[<].+[>]\z/ } || ''
+        canon = addrs[0] if canon.size < 1
+
+      else
+        canon = addrs.shift
+      end
+
+      return '' if !canon || canon == ''
+      canon = canon.delete('<>[]():;')  # Remove brackets, colons
+
+      if canon =~ /\A["].+["][@].+\z/
+        canon = canon.delete(%q|{}'`|)  # "localpart..."@example.jp
+      else
+        canon = canon.delete(%q|{}'"`|) # Remove brackets, quotations
+      end
+
+      return canon
+    end
+
+    # Expand VERP: Get the original recipient address from VERP
+    # @param    [String] email  VERP Address
+    # @return   [String]        Email address
+    # @example  Expand VERP address
+    #   expand_verp('bounce+neko=example.jp@example.org') #=> 'neko@example.jp'
+    def self.expand_verp(email)
+      local = email.split('@', 2).first
+      verp0 = ''
+
+      if cv = local.match(/\A[-_\w]+?[+](\w[-._\w]+\w)[=](\w[-.\w]+\w)\z/)
+        verp0 = cv[1] + '@' + cv[2]
+        return verp0 if Sisimai::RFC5322.is_emailaddress(verp0)
+
+      else
+        return ''
+      end
+    end
+
+    # Expand alias: remove from '+' to '@'
+    # @param    [String] email  Email alias string
+    # @return   [String]        Expanded email address
+    # @example  Expand alias
+    #   expand_alias('neko+straycat@example.jp') #=> 'neko@example.jp'
+    def self.expand_alias(email)
+      return '' unless Sisimai::RFC5322.is_emailaddress(email)
+
+      local = email.split('@')
+      value = ''
+      if cv = local[0].match(/\A([-_\w]+?)[+].+\z/)
+        value = sprintf('%s@%s', cv[1], local[1])
+      end
+      return value
+    end
+
+    @@roaccessors = [
+      :address, # [String] Email address
+      :user,    # [String] local part of the email address
+      :host,    # [String] domain part of the email address
+      :verp,    # [String] VERP
+      :alias,   # [String] alias of the email address
+    ]
+    @@roaccessors.each { |e| attr_reader   e }
+
+    # Constructor of Sisimai::Address
+    # @param <str>  [String] email          Email address
+    # @return       [Sisimai::Address, Nil] Object or Undef when the email
+    #                                       address was not valid
+    def initialize(email)
+      return nil unless email
+
+      if cv = email.match(/\A([^@]+)[@]([^@]+)\z/)
+        # Get the local part and the domain part from the email address
+        lpart = cv[1]
+        dpart = cv[2]
+
+        # Remove MIME-Encoded comment part
+        lpart = lpart.sub(/\A=[?].+[?]b[?].+[?]=/, '')
+        lpart = lpart.delete(%q|`'"<>|) unless lpart =~ /\A["].+["]\z/
+
+        aflag = false
+        addr0 = sprintf('%s@%s', lpart, dpart)
+        addr1 = Sisimai::Address.expand_verp(addr0)
+
+        if addr1.size < 1
+          addr1 = Sisimai::Address.expand_alias(addr0)
+          aflag = true if addr1.size > 0
+        end
+
+        if addr1.size > 0
+          # The email address is VERP or alias
+          addrs = addr1.split('@')
+          if aflag
+            # The email address is an alias
+            @alias = addr0
+
+          else
+            # The email address is a VERP
+            @verp = addr0
+          end
+          @user = addrs[0]
+          @host = addrs[1]
+
+        else
+          # The email address is neither VERP nor alias.
+          @user = lpart
+          @host = dpart
+
+        end
+        @address = sprintf('%s@%s', @user, @host)
+        @alias ||= ''
+        @verp  ||= ''
+
+      else
+        return nil
+      end
+    end
+
+    # Check whether the object has valid content or not
+    # @return        [True,False]   returns true if the object is void
+    def void
+      return true unless @address
+      return false
+    end
+
+    # Returns the value of address as String
+    # @return [String] Email address
+    def to_json(*)
+      return to_s
+    end
+
+    # Returns the value of address as String
+    # @return [String] Email address
+    def to_s
+      return self.address.to_s
+    end
+
+  end
+end