chronicle-email 0.1.1 → 0.2.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 85cc6d5cd7ef25487ecb960c67f12e287c5a2c28894c6cba0b075bd7f0511725
-  data.tar.gz: 39969678f5abf0104df8dce01dd6e30aeee2878fbe9082e7055f9ffd94c7ade5
+  metadata.gz: e37ee5bdabd254e03a06962503f5875ab75d2aa0b271aff0b946c73f673d5cc3
+  data.tar.gz: c351c35f1b5545f0d0f9e6b4375076b159dab88eb454cf3e946fdab84ddc03a4
 SHA512:
-  metadata.gz: 50ac6a07fdc6324430a54a6739e7493b72583f2e35220b6a92c5949cc5731b6276aaea9fbebcac012f4f32e57ef34f10ad069bdda43afad9f3b9ee15a6e832ed
-  data.tar.gz: 32f9fb63a319f0b37745dc297d229dcf1113f68bef528d13a4dd21dfb4dd83d74a41ff38e906ff6dda284fb8d34b066cb7bfbcafa642ad7e190f56173fb6f018
+  metadata.gz: fedb95cf29c21d0dac17d6cbeca81da0a5318ab41ea8a5eaf355e8996606b03398ba70c4233e4591a78ceaef5ce5242b59c5e227e79a6310ba4474c245476e57
+  data.tar.gz: 7a06834ea16425d456da704b3272c96d6e0f427044e46e60a477dbce1cab7f9621056f4f18f7b90a964f83e3e85281a981fdf155155ebf4cb8b06871cd9884ab

data/.gitignore

@@ -7,6 +7,8 @@
 /spec/reports/
 /tmp/
 
+Gemfile.lock
+
 # rspec failure tracking
 .rspec_status
 .DS_Store

data/README.md

@@ -1,16 +1,72 @@
 # Chronicle::Email
+[![Gem Version](https://badge.fury.io/rb/chronicle-email.svg)](https://badge.fury.io/rb/chronicle-email)
 
-Email importer for [chronicle-etl](https://github.com/chronicle-app/chronicle-etl)
-
-## Available classes
-- `email:mbox` - Extractor for processing .mbox files
-- `email:chronicle` - Transformer that converts an email into a chronicle schema
+Extract and work with your email using the command line with this plugin for [chronicle-etl](https://github.com/chronicle-app/chronicle-etl).
 
 ## Usage
 
-```bash
-gem install chronicle-etl
-gem install chronicle-email
+```sh
+# Install chronicle-etl and this plugin
+$ gem install chronicle-etl
+$ chronicle-etl plugins:install email
+```
+
+### Extracting email from IMAP
+
+For Gmail accounts, you can create an [app password](https://myaccount.google.com/apppasswords); your email address is your username.
+
+```sh
+# Save username and password
+$ chronicle-etl secrets:set imap username foo@gmail.com
+$ chronicle-etl secrets:set imap password APPPASSWORD
+
+# Then, retrieve your email from the last five days
+$ chronicle-etl --extractor email:imap --transformer email --since 5d --loader json
+
+# If you don't want to save your credentials as a secret, you can just pass
+# them to the extractor directly
+$ chronicle-etl --extractor email:imap --transformer email --since 5d --loader json \
+    --extractor-opts username:foo@gmail.com --password:APPPASSWORD
+```
+
+### Processing email from an .mbox file
+The MBOX format is used to archive an email mailbox. [Google Takeout](https://takeout.google.com/settings/takeout) exports emails from gmail in this format.
+
+```sh
+# Retrieve the subject lines of all emails in test.mbox
+$ chronicle-etl --extractor email:mbox --input inbox.mbox --transformer email --fields subject
+```
+
+## Available Connectors
+### Extractors
+
+#### `imap`
+Extractor for importing recent emails from an IMAP server.
+
+##### Settings
+
+- `since`: Retrieve emails since this date
+- `until`: Retrieve emails until this date
+- `username`
+- `password`
+- `host`: (default: imap.gmail.com)
+- `port`: (default: 993) Use 143 for unencrypted connections
+- `mailbox`: (default: "[Gmail]/All Mail")
+- `search_query`: When using Gmail, you can pass in a search query (`from:foo has:attachment`) to filter messages by
+
+For accessing Gmail, you can create a one-time [app password](https://myaccount.google.com/apppasswords). Your email address is your username.
+
+#### `mbox`
+Extractor for importing emails from an MBOX file
+
+##### Settings
+- `input`: A path to an .mbox file
+
+### Transformers
+
+#### `email`
+Transform an email (in the form of a string) into Chronicle Schema
 
-chronicle-etl --extractor email:mbox --extractor-opts filename:"./mail.mbox" --transformer email:chronicle --loader stdout
-```
+##### Settings
+- `body_as_markdown`: (default: false) Whether to convert the email body into markdown
+- `remove_signature`: (default: true) Whether to attempt to strip out the email signature (using the [`email_reply_parser`](https://github.com/github/email_reply_parser) gem)

data/chronicle-email.gemspec

@@ -36,12 +36,12 @@ Gem::Specification.new do |spec|
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
 
-  spec.add_dependency "chronicle-etl", "~> 0.2.2"
+  spec.add_dependency "chronicle-etl", "~> 0.5"
   spec.add_dependency "mail", "~> 2.7"
   spec.add_dependency 'email_reply_parser', '~> 0.5'
+  spec.add_dependency 'reverse_markdown', '~> 2.0'
 
   spec.add_development_dependency "bundler", "~> 2.1"
   spec.add_development_dependency "rake", "~> 13.0"
   spec.add_development_dependency "rspec", "~> 3.9"
-  spec.add_development_dependency "pry-byebug", "~> 3.9"
 end

data/lib/chronicle/email/email_transformer.rb

@@ -0,0 +1,122 @@
+require 'chronicle/etl'
+require 'mail'
+require 'timeout'
+require 'email_reply_parser'
+require 'reverse_markdown'
+
+module Chronicle
+  module Email
+    class EmailTransformer < Chronicle::ETL::Transformer
+      register_connector do |r|
+        r.description = 'an email object'
+        r.provider = 'email'
+        r.identifier = 'email'
+      end
+
+      setting :body_as_markdown, default: false
+      setting :remove_signature, default: true
+
+      def transform
+        build_messaged
+      end
+
+      def id
+        message.message_id || raise(Chronicle::ETL::UntransformableRecordError, "Email doesn't have an ID")
+      end
+
+      def timestamp
+        message.date&.to_time || raise(Chronicle::ETL::UntransformableRecordError, "Email doesn't have a timestamp")
+      end
+
+      private
+
+      def message
+        @message ||= Mail.new(@extraction.data[:email])
+      end
+
+      def build_messaged
+        record = ::Chronicle::ETL::Models::Activity.new
+        record.verb = 'messaged'
+        record.provider = 'email'
+        record.provider_id = id
+        record.end_at = timestamp
+
+        record.dedupe_on << [:verb, :provider, :provider_id]
+
+        record.actor = build_actor
+        record.involved = build_message
+        record
+      end
+
+      def build_actor
+        # sometimes From: fields are malformed and we can't build an
+        # actor out of it.
+        raise(Chronicle::ETL::UntransformableRecordError, "Can't determine email sender") unless message[:from]&.addrs&.any?
+
+        record = ::Chronicle::ETL::Models::Entity.new
+        record.represents = 'identity'
+        record.provider = 'email'
+        record.slug = message[:from].addrs.first.address
+        record.title = message[:from].addrs.first.display_name
+
+        record.dedupe_on << [:represents, :provider, :slug]
+
+        record
+      end
+
+      def build_message
+        record = ::Chronicle::ETL::Models::Entity.new
+        record.represents = 'message'
+        record.title = clean_subject(message.subject)
+        record.body = clean_body(message)
+        record.provider = 'email'
+        record.provider_id = id
+
+        # TODO: handle consumer
+        # TODO: handle email references
+        # TODO: handle email account owner
+        # TODO: handle attachments
+
+        record
+      end
+
+      def clean_subject(subject)
+        subject&.encode("UTF-8", invalid: :replace, undef: :replace)
+      end
+
+      def clean_body message
+        # FIXME: this all needs to be refactored        
+        if message.multipart?
+          body = message.text_part&.decoded rescue Mail::UnknownEncodingType
+        else
+          body = message.body&.decoded rescue Mail::UnknownEncodingType
+          body = body_to_markdown if @config.body_as_markdown
+        end
+
+        return if body == Mail::UnknownEncodingType
+        return unless body && body != ""
+
+        body = body_without_signature(body) if @config.remove_signature
+
+        # Force UTF-8 encoding
+        body.encode("UTF-8", invalid: :replace, undef: :replace)
+      end
+
+      def body_to_markdown(body)
+        ReverseMarkdown.convert(body)
+      rescue StandardError
+        # Fall back to unparsed body? Raise Untransformable error?
+      end
+
+      def body_without_signature(body)
+        # FIXME: regex in EmailReplyParse gem seems to get into infinite loops
+        #   with certain long bodies that have binary data
+        parsed_body = Timeout::timeout(5) do
+          EmailReplyParser.parse_reply(body)
+        end
+      rescue Timeout::Error, StandardError => e
+        return body
+      end
+    end
+  end
+end

data/lib/chronicle/email/imap_extractor.rb

@@ -0,0 +1,89 @@
+require 'net/imap'
+
+module Chronicle
+  module Email
+    class IMAPExtractor < Chronicle::ETL::Extractor
+      register_connector do |r|
+        r.provider = 'email'
+        r.description = 'imap server'
+        r.identifier = 'imap'
+      end
+
+      setting :host, required: true, default: 'imap.gmail.com'
+      setting :port, type: :numeric, required: true, default: 993
+      setting :mailbox, required: true, default: '[Gmail]/All Mail'
+      setting :username, required: true
+      setting :password, required: true
+      setting :search_query
+
+      def prepare
+        @connection = create_connection
+        @message_ids = fetch_message_ids
+      end
+
+      def results_count
+        @message_ids.count
+      end
+
+      def extract
+        @message_ids.each do |message_id|
+          message = fetch_message(message_id)
+          yield Chronicle::ETL::Extraction.new(data: { email: message.attr["BODY[]"]} )
+        end
+      end
+
+      private
+
+      def create_connection
+        connection = Net::IMAP.new(@config.host, @config.port, true)
+        connection.login(@config.username, @config.password)
+        connection.select(@config.mailbox)
+        connection
+      rescue Net::IMAP::NoResponseError => e
+        raise(Chronicle::ETL::ExtractionError, "Error connecting to IMAP server. Please check username and password")
+      end
+
+      def fetch_message_ids
+        keys = gmail_mode? ? search_keys_gmail : search_keys_default
+        @connection.search(keys)
+      rescue Net::IMAP::BadResponseError => e
+        raise(Chronicle::ETL::ExtractionError, "Error searching IMAP server for messages")
+      end
+
+      def fetch_message(message_id)
+        response = @connection.fetch(3100020, "BODY.PEEK[]")
+        raise(Chronicle::ETL::ExtractionError, "Error loading message") unless response
+
+        return response[0]
+      end
+
+      def search_keys_gmail
+        # Gmail offers an extension to IMAP that lets us use gmail queries
+        q = ""
+
+        # First, we ignore drafts beacuse they break a lot of assumptions we
+        # make when when processing emails (lack of timestamps, ids, etc)
+        q = "-label:draft"
+
+        # We use UNIX timestamps in gmail filters which let us do more precise
+        # since/until compared with date-based imap filters
+        q += " after:#{@config.since.to_i}" if @config.since
+        q += " before:#{@config.until.to_i}" if @config.until
+        q += " #{@config.search_query}" if @config.search_query
+
+        ["X-GM-RAW", q]
+      end
+
+      def search_keys_default
+        keys = []
+        # TODO: test out non-gmail IMAP searching (for @config.search_query)
+        keys += ['SINCE', Net::IMAP.format_date(@config.since)] if @config.since
+        keys += ['BEFORE', Net::IMAP.format_date(@config.until)] if @config.until
+      end
+
+      def gmail_mode?
+        @config.host == 'imap.gmail.com'
+      end
+    end
+  end
+end

data/lib/chronicle/email/mbox_extractor.rb

@@ -5,35 +5,55 @@ require 'tempfile'
 module Chronicle
   module Email
     class MboxExtractor < Chronicle::ETL::Extractor
+      register_connector do |r|
+        r.provider = 'email'
+        r.description = 'an .mbox file'
+        r.identifier = 'mbox'
+      end
+
+      setting :input, required: true
+
       # mbox format is a bunch of emails concatanated together, separated
       # by a line that starts with "From "
       NEW_EMAIL_REGEX = Regexp.new('^From [^\s]+ .{24}')
 
       def results_count
-        file = File.open(@options[:filename])
-        count = 0
-        file.each do |line|
-          count += 1 if line =~ NEW_EMAIL_REGEX
+        File.foreach(@filename).sum do |line|
+          line.scan(NEW_EMAIL_REGEX).count
         end
-        return count
+      end
+
+      def prepare
+        @filename = @config.input.first
       end
 
       def extract
-        file = File.open(@options[:filename])
-        tmp = Tempfile.new('chronicile-mbox')
+        file = File.open(@filename)
+        tmp = Tempfile.new('chronicle-mbox')
 
+        # Read the .mbox file line by line and look for a header that indicates
+        # the start of a new email. As we read line by line, we save to a tmp
+        # file and then read it back when we notice the next header.
+        # Doing it this way is a lot faster than saving each line to a
+        # a variable, especially when we're reading emails with large binary
+        # attachments.
+        #
+        # TODO: make this thread-safe (one tmp file per email?)
         file.each do |line|
           if line =~ NEW_EMAIL_REGEX
             if File.size(tmp) > 0
               tmp.rewind
               email = tmp.read
-              yield email
+              yield Chronicle::ETL::Extraction.new(data: { email: email} )
               tmp.truncate(0)
               tmp.rewind
             end
           end
           tmp.write(line)
         end
+      ensure
+        tmp.close
+        tmp.unlink
         file.close
       end
     end

data/lib/chronicle/email/version.rb

@@ -1,5 +1,5 @@
 module Chronicle
   module Email
-    VERSION = "0.1.1"
+    VERSION = "0.2.2"
   end
 end

data/lib/chronicle/email.rb

@@ -1,9 +1,9 @@
 require "chronicle/email/version"
-require "chronicle/email/chronicle_transformer"
+require "chronicle/email/email_transformer"
 require "chronicle/email/mbox_extractor"
+require "chronicle/email/imap_extractor"
 
 module Chronicle
   module Email
-    PROVIDER_NAME = "email"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: chronicle-email
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.2.2
 platform: ruby
 authors:
 - Andrew Louis
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2020-08-29 00:00:00.000000000 Z
+date: 2022-03-31 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: chronicle-etl
@@ -16,14 +16,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.2
+        version: '0.5'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.2
+        version: '0.5'
 - !ruby/object:Gem::Dependency
   name: mail
   requirement: !ruby/object:Gem::Requirement
@@ -53,49 +53,49 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0.5'
 - !ruby/object:Gem::Dependency
-  name: bundler
+  name: reverse_markdown
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.1'
-  type: :development
+        version: '2.0'
+  type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.1'
+        version: '2.0'
 - !ruby/object:Gem::Dependency
-  name: rake
+  name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '13.0'
+        version: '2.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '13.0'
+        version: '2.1'
 - !ruby/object:Gem::Dependency
-  name: rspec
+  name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.9'
+        version: '13.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.9'
+        version: '13.0'
 - !ruby/object:Gem::Dependency
-  name: pry-byebug
+  name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
@@ -117,11 +117,9 @@ extra_rdoc_files: []
 files:
 - ".gitignore"
 - ".rspec"
-- ".ruby-version"
 - ".travis.yml"
 - CODE_OF_CONDUCT.md
 - Gemfile
-- Gemfile.lock
 - LICENSE.txt
 - README.md
 - Rakefile
@@ -129,7 +127,8 @@ files:
 - bin/setup
 - chronicle-email.gemspec
 - lib/chronicle/email.rb
-- lib/chronicle/email/chronicle_transformer.rb
+- lib/chronicle/email/email_transformer.rb
+- lib/chronicle/email/imap_extractor.rb
 - lib/chronicle/email/mbox_extractor.rb
 - lib/chronicle/email/version.rb
 homepage: https://github.com/chronicle-app/chronicle-email
@@ -154,7 +153,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
+rubygems_version: 3.3.3
 signing_key:
 specification_version: 4
 summary: Email importer for Chronicle

data/.ruby-version

@@ -1 +0,0 @@
-2.7.1

data/Gemfile.lock

@@ -1,85 +0,0 @@
-PATH
-  remote: .
-  specs:
-    chronicle-email (0.1.1)
-      chronicle-etl (~> 0.2.2)
-      email_reply_parser (~> 0.5)
-      mail (~> 2.7)
-
-GEM
-  remote: https://rubygems.org/
-  specs:
-    byebug (11.1.3)
-    chronicle-etl (0.2.2)
-      colorize (~> 0.8.1)
-      thor (~> 0.20)
-      tty-progressbar (~> 0.17)
-      tty-table (~> 0.11)
-    coderay (1.1.3)
-    colorize (0.8.1)
-    diff-lcs (1.4.4)
-    email_reply_parser (0.5.10)
-    equatable (0.6.1)
-    mail (2.7.1)
-      mini_mime (>= 0.1.1)
-    method_source (1.0.0)
-    mini_mime (1.0.2)
-    necromancer (0.6.0)
-    pastel (0.7.4)
-      equatable (~> 0.6)
-      tty-color (~> 0.5)
-    pry (0.13.1)
-      coderay (~> 1.1)
-      method_source (~> 1.0)
-    pry-byebug (3.9.0)
-      byebug (~> 11.0)
-      pry (~> 0.13.0)
-    rake (13.0.1)
-    rspec (3.9.0)
-      rspec-core (~> 3.9.0)
-      rspec-expectations (~> 3.9.0)
-      rspec-mocks (~> 3.9.0)
-    rspec-core (3.9.2)
-      rspec-support (~> 3.9.3)
-    rspec-expectations (3.9.2)
-      diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.9.0)
-    rspec-mocks (3.9.1)
-      diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.9.0)
-    rspec-support (3.9.3)
-    strings (0.1.8)
-      strings-ansi (~> 0.1)
-      unicode-display_width (~> 1.5)
-      unicode_utils (~> 1.4)
-    strings-ansi (0.1.0)
-    thor (0.20.3)
-    tty-color (0.5.2)
-    tty-cursor (0.7.1)
-    tty-progressbar (0.17.0)
-      strings-ansi (~> 0.1.0)
-      tty-cursor (~> 0.7)
-      tty-screen (~> 0.7)
-      unicode-display_width (~> 1.6)
-    tty-screen (0.8.1)
-    tty-table (0.11.0)
-      equatable (~> 0.6)
-      necromancer (~> 0.5)
-      pastel (~> 0.7.2)
-      strings (~> 0.1.5)
-      tty-screen (~> 0.7)
-    unicode-display_width (1.7.0)
-    unicode_utils (1.4.0)
-
-PLATFORMS
-  ruby
-
-DEPENDENCIES
-  bundler (~> 2.1)
-  chronicle-email!
-  pry-byebug (~> 3.9)
-  rake (~> 13.0)
-  rspec (~> 3.9)
-
-BUNDLED WITH
-   2.1.4

data/lib/chronicle/email/chronicle_transformer.rb

@@ -1,150 +0,0 @@
-require 'chronicle/etl'
-require 'mail'
-require 'timeout'
-require 'email_reply_parser'
-
-module Chronicle
-  module Email
-    class ChronicleTransformer < Chronicle::ETL::Transformer
-      def transform
-        message = Mail.new(@data.b)
-        build_messaged(message)
-      end
-
-      def build_messaged message
-        {
-          type: 'activities',
-          attributes: {
-            verb: 'messaged',
-            end_at: message.date,
-            provider: 'email',
-            provider_id: message.message_id,
-          },
-          meta: { dedupe_on: 'verb,provider,provider_id'},
-          relationships: {
-            actor: { data: build_actor(message) },
-            involved: { data: build_message(message) }
-          }
-        }
-      end
-
-      def build_actor message
-        # sometimes From: fields are malformed and we can't build an
-        # actor out of it.
-        return unless message[:from] && message[:from].addrs && message[:from].addrs.any?
-
-        {
-          type: 'entities',
-          attributes: {
-            represents: 'identity',
-            provider: 'email',
-            slug: message[:from].addrs.first.address,
-            title: message[:from].addrs.first.display_name
-          },
-          meta: { dedupe_on: 'represents,provider,slug'}
-        }
-      end
-
-      def build_message message
-        {
-          type: 'entities',
-          attributes: {
-            represents: 'message',
-            title: clean_subject(message.subject),
-            body: clean_body(message),
-            provider: 'email',
-            provider_id: message.message_id
-          },
-          meta: { dedupe_on: 'represents,provider,provider_id'},
-          relationships: {
-            consumers: { data: build_consumers(message) },
-            antecedents: { data: build_references(message) },
-            owners: { data: build_account(message) },
-            # contains: { data: build_attachments(message) }
-          }
-        }
-      end
-
-      def build_account message
-        return unless account_email = [message.header['delivered-to']].flatten[0]&.value
-
-        {
-          type: 'entities',
-          attributes: {
-            represents: 'identity',
-            provider: 'email',
-            slug: account_email
-          },
-          meta: { dedupe_on: 'provider,slug,represents' }
-        }
-      end
-
-      def build_consumers(message)
-        to = []
-        to += message[:to].addrs if message[:to]
-        to += message[:cc].addrs.flatten.compact if message[:cc]
-
-        to.collect do |consumer|
-          {
-            type: 'entities',
-            attributes: {
-              represents: 'identity',
-              provider: 'email',
-              slug: consumer.address,
-              title: consumer.display_name
-            },
-            meta: { dedupe_on: 'provider,slug' }
-          }
-        end
-      end
-
-      def build_references(message)
-        references = [message.references].flatten.compact
-        references.collect{|reference|
-          {
-            type: 'entities',
-            attributes: {
-              represents: 'message',
-              provider: 'email',
-              provider_id: reference
-            },
-            meta: { dedupe_on: 'represents,provider,provider_id' }
-          }
-        }
-      end
-
-      def clean_subject(subject)
-        subject&.encode("UTF-8", invalid: :replace, undef: :replace)
-      end
-
-      def clean_body message
-        # FIXME: this all needs to be refactored
-
-        if message.multipart?
-          body = message.text_part&.decoded rescue Mail::UnknownEncodingType
-        else
-          body = message.body&.decoded rescue Mail::UnknownEncodingType
-        end
-
-        if body && body != ""
-          begin
-            # regex in EmailReplyParse gem seems to get into infinite loops with
-            # certain long bodies that have binary data
-            parsed_body = Timeout::timeout(5) do
-              EmailReplyParser.parse_reply(body)
-            end
-          rescue Timeout::Error => e
-            return nil
-          rescue StandardError => e  # Whackamole game with these parsing / encoding problems
-            return nil
-          end
-
-          # Force UTF-8 encoding
-          return parsed_body.encode("UTF-8", invalid: :replace, undef: :replace)
-        else
-          return nil
-        end
-      end
-    end
-  end
-end