email_reply_parser 0.4.0 → 0.5.0

data/email_reply_parser.gemspec

@@ -13,7 +13,7 @@ Gem::Specification.new do |s|
   ## If your rubyforge_project name is different, then edit it and comment out
   ## the sub! line in the Rakefile
   s.name              = 'email_reply_parser'
-  s.version           = '0.4.0'
+  s.version           = '0.5.0'
   s.date              = '2012-03-01'
   s.rubyforge_project = 'email_reply_parser'
 
@@ -72,6 +72,12 @@ Gem::Specification.new do |s|
     test/emails/email_1_4.txt
     test/emails/email_1_5.txt
     test/emails/email_1_6.txt
+    test/emails/email_2_1.txt
+    test/emails/email_BlackBerry.txt
+    test/emails/email_bullets.txt
+    test/emails/email_iPhone.txt
+    test/emails/email_multi_word_sent_from_my_mobile_device.txt
+    test/emails/email_sent_from_my_not_signature.txt
   ]
   # = MANIFEST =
 

data/lib/email_reply_parser.rb

@@ -30,9 +30,9 @@ require 'strscan'
 #
 # [mail]: https://github.com/mikel/mail
 class EmailReplyParser
-  VERSION = "0.4.0"
+  VERSION = "0.5.0"
 
-  # Splits an email body into a list of Fragments.
+  # Public: Splits an email body into a list of Fragments.
   #
   # text - A String email body.
   #
@@ -41,6 +41,15 @@ class EmailReplyParser
     Email.new.read(text)
   end
 
+  # Public: Get the text of the visible portions of the given email body.
+  #
+  # text - A String email body.
+  #
+  # Returns a String.
+  def self.parse_reply(text)
+    self.read(text).visible_text
+  end
+
   ### Emails
 
   # An Email instance represents a parsed body String.
@@ -52,6 +61,13 @@ class EmailReplyParser
       @fragments = []
     end
 
+    # Public: Gets the combined text of the visible fragments of the email body.
+    #
+    # Returns a String.
+    def visible_text
+      fragments.select{|f| !f.hidden?}.map{|f| f.to_s}.join("\n").rstrip
+    end
+
     # Splits the given text into a list of Fragments.  This is roughly done by
     # reversing the text and parsing from the bottom to the top.  This way we
     # can check for 'On <date>, <author> wrote:' lines above quoted blocks.
@@ -105,8 +121,8 @@ class EmailReplyParser
 
   private
     EMPTY = "".freeze
-    SIG_REGEX = /(\s--|__|\w-)$/
-
+    SIG_REGEX = /(--|__|\w-$)|(^(\w+\s*){1,3} #{"Sent from my".reverse}$)/
+    
     ### Line-by-Line Parsing
 
     # Scans the given line of text and figures out which fragment it belongs
@@ -240,3 +256,4 @@ class EmailReplyParser
     end
   end
 end
+

data/test/email_reply_parser_test.rb

@@ -101,6 +101,47 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal "The Quick Brown Fox Jumps Over The Lazy Dog", original
   end
 
+  def test_returns_only_the_visible_fragments_as_a_string
+    reply = email(:email_2_1)
+    assert_equal reply.fragments.select{|r| !r.hidden?}.map{|r| r.to_s}.join("\n").rstrip, reply.visible_text
+  end
+
+  def test_parse_out_just_top_for_outlook_reply
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_2_1.txt").to_s
+    assert_equal "Outlook with a reply", EmailReplyParser.parse_reply(body)
+  end
+
+  def test_parse_out_sent_from_iPhone
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_iPhone.txt").to_s
+    assert_equal "Here is another email", EmailReplyParser.parse_reply(body)
+  end
+
+  def test_parse_out_sent_from_BlackBerry
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_BlackBerry.txt").to_s
+    assert_equal "Here is another email", EmailReplyParser.parse_reply(body)
+  end
+
+  def test_parse_out_send_from_multiword_mobile_device
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_multi_word_sent_from_my_mobile_device.txt").to_s
+    assert_equal "Here is another email", EmailReplyParser.parse_reply(body)
+  end
+
+  def test_do_not_parse_out_send_from_in_regular_sentence
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_sent_from_my_not_signature.txt").to_s
+    assert_equal "Here is another email\n\nSent from my desk, is much easier then my mobile phone.", EmailReplyParser.parse_reply(body)
+  end
+
+  def test_retains_bullets
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_bullets.txt").to_s
+    assert_equal "test 2 this should list second\n\nand have spaces\n\nand retain this formatting\n\n\n   - how about bullets\n   - and another", 
+      EmailReplyParser.parse_reply(body)
+      end
+
+  def test_parse_reply
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_1_2.txt").to_s
+    assert_equal EmailReplyParser.read(body).visible_text, EmailReplyParser.parse_reply(body)
+  end
+
   def email(name)
     body = IO.read EMAIL_FIXTURE_PATH.join("#{name}.txt").to_s
     EmailReplyParser.read body

data/test/emails/email_2_1.txt

@@ -0,0 +1,25 @@
+Outlook with a reply
+
+
+ ------------------------------
+
+*From:* Google Apps Sync Team [mailto:mail-noreply@google.com]
+*Sent:* Thursday, February 09, 2012 1:36 PM
+*To:* jow@xxxx.com
+*Subject:* Google Apps Sync was updated!
+
+
+
+Dear Google Apps Sync user,
+
+Google Apps Sync for Microsoft Outlook® was recently updated. Your computer
+now has the latest version (version 2.5). This release includes bug fixes
+to improve product reliability. For more information about these and other
+changes, please see the help article here:
+
+http://www.google.com/support/a/bin/answer.py?answer=153463
+
+Sincerely,
+
+The Google Apps Sync Team.
+

data/test/emails/email_BlackBerry.txt

@@ -0,0 +1,3 @@
+Here is another email
+
+Sent from my BlackBerry

data/test/emails/email_bullets.txt

@@ -0,0 +1,22 @@
+test 2 this should list second
+
+and have spaces
+
+and retain this formatting
+
+
+   - how about bullets
+   - and another
+
+
+On Fri, Feb 24, 2012 at 10:19 AM, <examples@email.goalengine.com> wrote:
+
+> Give us an example of how you applied what they learned to achieve
+> something in your organization
+
+
+
+
+-- 
+
+*Joe Smith | Director, Product Management*

data/test/emails/email_iPhone.txt

@@ -0,0 +1,3 @@
+Here is another email
+
+Sent from my iPhone

data/test/emails/email_multi_word_sent_from_my_mobile_device.txt

@@ -0,0 +1,3 @@
+Here is another email
+
+Sent from my Verizon Wireless BlackBerry

data/test/emails/email_sent_from_my_not_signature.txt

@@ -0,0 +1,3 @@
+Here is another email
+
+Sent from my desk, is much easier then my mobile phone.

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: email_reply_parser
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.5.0
   prerelease: 
 platform: ruby
 authors:
@@ -32,6 +32,12 @@ files:
 - test/emails/email_1_4.txt
 - test/emails/email_1_5.txt
 - test/emails/email_1_6.txt
+- test/emails/email_2_1.txt
+- test/emails/email_BlackBerry.txt
+- test/emails/email_bullets.txt
+- test/emails/email_iPhone.txt
+- test/emails/email_multi_word_sent_from_my_mobile_device.txt
+- test/emails/email_sent_from_my_not_signature.txt
 homepage: http://github.com/github/email_reply_parser
 licenses: []
 post_install_message: