email_reply_parser 0.5.9 → 0.5.11

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 18ed91d7b72ad0b331359d9719a6f69f02366bdd
-  data.tar.gz: b97b2b17d548dfc86066001d936b49bf6035a0e1
+SHA256:
+  metadata.gz: 19456eb75469b983f8db4a0807f6691213264309223f37d02c7fcf44748e2d7c
+  data.tar.gz: 0a67068492fd8c2fe7221420e383f69f7ad0f6e7f435882830d8553e6f917bcf
 SHA512:
-  metadata.gz: fb73afe46d2313905f2cf542d5adf134439bed105922272f53bc6a1e93a3a786e2ab18445dcb0f6712c62251ef0a0632c8cdf1dca441b0d6b6413cd6b83cc8ba
-  data.tar.gz: 74370b59c92a813864096e4436c8c606bdc350d1ad43af12eded545d4eea38ea76764a98c5bbb3864ef813694f51eefb3895c3622cdd78844ebfff2354be32bf
+  metadata.gz: e6b065c74ce3eec383d570824c3c22a0ff9fc1cbc17a99cbbac230a92a0eb9c90adcc96f6288891a676ff65da1a791d11b7c1a3cce9a22bacb3752c6fab6abf0
+  data.tar.gz: 72e65bff3a0850354268451e89ea352a8ab59c5bb8fe1c27ff7558df740ecd233915daaf5e35830a4c9c07617bc5ce6634a0c50aa5d3bd3464821c3328824e7e

data/README.md

@@ -11,7 +11,7 @@ See the [Ruby docs][rubydocs] for more information.
 
 [rubydocs]: http://rubydoc.info/gems/email_reply_parser/
 
-##Usage
+## Usage
 
 To parse reply body:
 

data/email_reply_parser.gemspec

@@ -1,60 +1,33 @@
 $LOAD_PATH.unshift '.'
 require 'lib/email_reply_parser'
 
-## This is the rakegem gemspec template. Make sure you read and understand
-## all of the comments. Some sections require modification, and others can
-## be deleted if you don't need them. Once you understand the contents of
-## this file, feel free to delete any comments that begin with two hash marks.
-## You can find comprehensive Gem::Specification documentation, at
-## http://docs.rubygems.org/read/chapter/20
 Gem::Specification.new do |s|
   s.specification_version = 2 if s.respond_to? :specification_version=
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.rubygems_version = '1.3.5'
+  s.license = 'MIT'
 
-  ## Leave these as is they will be modified for you by the rake gemspec task.
-  ## 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           = EmailReplyParser::VERSION
   s.date              = Time.now.strftime('%Y-%m-%d')
-  s.rubyforge_project = 'email_reply_parser'
 
-  ## Make sure your summary is short. The description may be as long
-  ## as you like.
-  s.summary     = "Short description used in Gem listings."
-  s.description = "Long description. Maybe copied from the README."
+  s.summary     = "EmailReplyParser is a small library to parse plain text " \
+                  "email content."
+  s.description = "EmailReplyParser is a small library to parse plain text " \
+                  "email content. This is what GitHub uses to display comments " \
+                  "that were created from email replies."
 
-  ## List the primary authors. If there are a bunch of authors, it's probably
-  ## better to set the email to an email list or something. If you don't have
-  ## a custom homepage, consider using your GitHub URL or the like.
   s.authors  = ["Rick Olson"]
   s.email    = 'technoweenie@gmail.com'
   s.homepage = 'http://github.com/github/email_reply_parser'
 
-  ## This gets added to the $LOAD_PATH so that 'lib/NAME.rb' can be required as
-  ## require 'NAME.rb' or'/lib/NAME/file.rb' can be as require 'NAME/file.rb'
   s.require_paths = %w[lib]
 
-  ## This sections is only necessary if you have C extensions.
-  #s.require_paths << 'ext'
-  #s.extensions = %w[ext/extconf.rb]
-
-  ## If your gem includes any executables, list them here.
-  #s.executables = ["name"]
-  #s.default_executable = 'name'
-
-  ## Specify any RDoc options here. You'll want to add your README and
-  ## LICENSE files to the extra_rdoc_files list.
   s.rdoc_options = ["--charset=UTF-8"]
   s.extra_rdoc_files = %w[README.md LICENSE]
 
-  ## List your runtime dependencies here. Runtime dependencies are those
-  ## that are needed for an end user to actually USE your code.
   #s.add_dependency('DEPNAME', [">= 1.1.0", "< 2.0.0"])
 
-  ## List your development dependencies here. Development dependencies are
-  ## those that are only needed during development
   #s.add_development_dependency('DEVDEPNAME', [">= 1.1.0", "< 2.0.0"])
 
   ## Leave this section as-is. It will be automatically generated from the
@@ -81,9 +54,11 @@ Gem::Specification.new do |s|
     test/emails/email_1_8.txt
     test/emails/email_2_1.txt
     test/emails/email_2_2.txt
+    test/emails/email_2_3.txt
     test/emails/email_BlackBerry.txt
     test/emails/email_bullets.txt
     test/emails/email_iPhone.txt
+    test/emails/email_long_quote.txt
     test/emails/email_multi_word_sent_from_my_mobile_device.txt
     test/emails/email_one_is_not_on.txt
     test/emails/email_sent_from_my_not_signature.txt
@@ -93,8 +68,5 @@ Gem::Specification.new do |s|
   ]
   # = MANIFEST =
 
-  ## Test files will be grabbed from the file list. Make sure the path glob
-  ## matches what you actually use.
   s.test_files = s.files.select { |path| path =~ /^test\/.*_test\.rb/ }
 end
-

data/lib/email_reply_parser.rb

@@ -30,7 +30,7 @@ require 'strscan'
 #
 # [mail]: https://github.com/mikel/mail
 class EmailReplyParser
-  VERSION = "0.5.9"
+  VERSION = "0.5.11"
 
   # Public: Splits an email body into a list of Fragments.
   #
@@ -132,14 +132,8 @@ class EmailReplyParser
 
   private
     EMPTY = "".freeze
-    SIGNATURE = '(?m)(--\s*$|__\s*$|\w-$)|(^(\w+\s*){1,3} ym morf tneS$)'
-
-    begin
-      require 're2'
-      SIG_REGEX = RE2::Regexp.new(SIGNATURE)
-    rescue LoadError
-      SIG_REGEX = Regexp.new(SIGNATURE)
-    end
+    SIGNATURE = '(?m)(--\s*$|__\s*$|\w-$)|(^(\w+\s+){1,3}ym morf tneS$)'
+    SIG_REGEX = Regexp.new(SIGNATURE)
 
     ### Line-by-Line Parsing
 
@@ -153,9 +147,9 @@ class EmailReplyParser
       line.chomp!("\n")
       line.lstrip! unless SIG_REGEX.match(line)
 
-      # We're looking for leading `>`'s to see if this line is part of a
+      # We're looking for a leading `>` to see if this line is part of a
       # quoted Fragment.
-      is_quoted = !!(line =~ /(>+)$/)
+      is_quoted = !!(line =~ /(>)$/)
 
       # Mark the current Fragment as a signature if the current line is empty
       # and the Fragment starts with a common signature indicator.
@@ -188,7 +182,7 @@ class EmailReplyParser
     #
     # Returns true if the line is a valid header, or false.
     def quote_header?(line)
-      line =~ /^:etorw.*nO$/
+      line =~ /^:etorw.*nO$/ || line =~ /^.*:(morF|tneS|oT|tcejbuS)$/
     end
 
     # Builds the fragment string and reverses it, after all lines have been

data/test/email_reply_parser_test.rb

@@ -2,6 +2,7 @@ require 'rubygems'
 require 'test/unit'
 require 'pathname'
 require 'pp'
+require 'timeout'
 
 dir = Pathname.new File.expand_path(File.dirname(__FILE__))
 require dir + '..' + 'lib' + 'email_reply_parser'
@@ -50,10 +51,10 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal [false, true, false, false, true],
       reply.fragments.map { |f| f.signature? }
 
-    assert_match /^Oh thanks.\n\nHaving/, reply.fragments[0].to_s
-    assert_match /^-A/, reply.fragments[1].to_s
-    assert_match /^On [^\:]+\:/, reply.fragments[2].to_s
-    assert_match /^_/, reply.fragments[4].to_s
+    assert_match(/^Oh thanks.\n\nHaving/, reply.fragments[0].to_s)
+    assert_match(/^-A/, reply.fragments[1].to_s)
+    assert_match(/^On [^\:]+\:/, reply.fragments[2].to_s)
+    assert_match(/^_/, reply.fragments[4].to_s)
   end
 
   def test_reads_bottom_post
@@ -68,10 +69,10 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
       reply.fragments.map { |f| f.hidden? }
 
     assert_equal "Hi,", reply.fragments[0].to_s
-    assert_match /^On [^\:]+\:/, reply.fragments[1].to_s
-    assert_match /^You can list/, reply.fragments[2].to_s
-    assert_match /^> /, reply.fragments[3].to_s
-    assert_match /^_/, reply.fragments[5].to_s
+    assert_match(/^On [^\:]+\:/, reply.fragments[1].to_s)
+    assert_match(/^You can list/, reply.fragments[2].to_s)
+    assert_match(/^> /, reply.fragments[3].to_s)
+    assert_match(/^_/, reply.fragments[5].to_s)
   end
 
   def test_reads_inline_replies
@@ -85,11 +86,11 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal [false, false, false, false, true, true, true],
       reply.fragments.map { |f| f.hidden? }
 
-    assert_match /^On [^\:]+\:/, reply.fragments[0].to_s
-    assert_match /^I will reply/, reply.fragments[1].to_s
+    assert_match(/^On [^\:]+\:/, reply.fragments[0].to_s)
+    assert_match(/^I will reply/, reply.fragments[1].to_s)
     assert_match "okay?", reply.fragments[2].to_s
-    assert_match /^and under this./, reply.fragments[3].to_s
-    assert_match /inline/, reply.fragments[4].to_s
+    assert_match(/^and under this./, reply.fragments[3].to_s)
+    assert_match(/inline/, reply.fragments[4].to_s)
     assert_equal "\n", reply.fragments[5].to_s
     assert_equal "--\nHey there, this is my signature\n", reply.fragments[6].to_s
   end
@@ -97,9 +98,9 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
   def test_recognizes_date_string_above_quote
     reply = email :email_1_4
 
-    assert_match /^Awesome/, reply.fragments[0].to_s
-    assert_match /^On/,      reply.fragments[1].to_s
-    assert_match /Loader/,   reply.fragments[1].to_s
+    assert_match(/^Awesome/, reply.fragments[0].to_s)
+    assert_match(/^On/,      reply.fragments[1].to_s)
+    assert_match(/Loader/,   reply.fragments[1].to_s)
   end
 
   def test_a_complex_body_with_only_one_fragment
@@ -115,23 +116,28 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal [false, false], reply.fragments.map { |f| f.quoted? }
     assert_equal [false, true], reply.fragments.map { |f| f.signature? }
     assert_equal [false, true], reply.fragments.map { |f| f.hidden? }
-    assert_match /^-- \nrick/, reply.fragments[1].to_s
+    assert_match(/^-- \nrick/, reply.fragments[1].to_s)
   end
 
   def test_deals_with_multiline_reply_headers
     reply = email :email_1_6
 
-    assert_match /^I get/,   reply.fragments[0].to_s
-    assert_match /^On/,      reply.fragments[1].to_s
-    assert_match /Was this/, reply.fragments[1].to_s
+    assert_match(/^I get/,   reply.fragments[0].to_s)
+    assert_match(/^On/,      reply.fragments[1].to_s)
+    assert_match(/Was this/, reply.fragments[1].to_s)
   end
 
   def test_deals_with_windows_line_endings
     reply = email :email_1_7
 
-    assert_match /:\+1:/,     reply.fragments[0].to_s
-    assert_match /^On/,       reply.fragments[1].to_s
-    assert_match /Steps 0-2/, reply.fragments[1].to_s
+    assert_match(/:\+1:/,     reply.fragments[0].to_s)
+    assert_match(/^On/,       reply.fragments[1].to_s)
+    assert_match(/Steps 0-2/, reply.fragments[1].to_s)
+  end
+
+  def test_handles_non_ascii_characters
+    non_ascii_body = "Here’s a test."
+    assert_equal non_ascii_body, EmailReplyParser.parse_reply(non_ascii_body)
   end
 
   def test_does_not_modify_input_string
@@ -155,6 +161,11 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal "Outlook with a reply directly above line", EmailReplyParser.parse_reply(body)
   end
 
+  def test_parse_out_just_top_for_outlook_with_no_line
+    body = IO.read EMAIL_FIXTURE_PATH.join("email_2_3.txt").to_s
+    assert_equal "Outlook with a reply directly above line", 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)
@@ -188,14 +199,14 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
 
   def test_one_is_not_on
     reply = email("email_one_is_not_on")
-    assert_match /One outstanding question/, reply.fragments[0].to_s
-    assert_match /^On Oct 1, 2012/, reply.fragments[1].to_s
+    assert_match(/One outstanding question/, reply.fragments[0].to_s)
+    assert_match(/^On Oct 1, 2012/, reply.fragments[1].to_s)
   end
 
   def test_mulitple_on
     reply = email("greedy_on")
-    assert_match /^On your remote host/, reply.fragments[0].to_s
-    assert_match /^On 9 Jan 2014/, reply.fragments[1].to_s
+    assert_match(/^On your remote host/, reply.fragments[0].to_s)
+    assert_match(/^On 9 Jan 2014/, reply.fragments[1].to_s)
     assert_equal [false, true, false], reply.fragments.map { |f| f.quoted? }
     assert_equal [false, false, false], reply.fragments.map { |f| f.signature? }
     assert_equal [false, true, true], reply.fragments.map { |f| f.hidden? }
@@ -203,8 +214,8 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
 
   def test_pathological_emails
     t0 = Time.now
-    reply = email("pathological")
-    assert (Time.now - t0) < 1, "Took too long, upgrade to re2 gem."
+    email("pathological")
+    assert (Time.now - t0) < 1, "Took too long, upgrade to re2 gem. See https://rubygems.org/gems/re2"
   end
 
   def test_doesnt_remove_signature_delimiter_in_mid_line
@@ -212,6 +223,12 @@ I am currently using the Java HTTP API.\n", reply.fragments[0].to_s
     assert_equal 1, reply.fragments.size
   end
 
+  def test_long_quote_processing_completes
+    reply = Timeout.timeout(1) { email(:email_long_quote) }
+
+    assert_equal 5, reply.fragments.size
+  end
+
   def email(name)
     body = IO.read EMAIL_FIXTURE_PATH.join("#{name}.txt").to_s
     EmailReplyParser.read body

data/test/emails/email_2_3.txt

@@ -0,0 +1,10 @@
+Outlook with a reply directly above line
+
+From: CRM Comments [crm-comment@example.com]
+Sent: Friday, 23 March 2012 5:08 p.m.
+To: John S. Greene
+Subject: [contact:106] John Greene
+
+> A new comment has been added to the Contact named 'John Greene':
+>
+> I am replying to a comment.