emailparser 0.0.1

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 3e6873cbc10f2ca98f4c1eb06998a03723006291
+  data.tar.gz: bf6276064de2d634ae0b295584c6ec17983e07b9
+SHA512:
+  metadata.gz: 08d5b376e97ee1d0af8cd75351f38b20a7d52df576e5d9dbafa824a1ee57a53fd9b95d9d29d7e3af9cc369ca5320fc1b469a5cf70ef5fb388453fcb866e4cd6f
+  data.tar.gz: 6a6a4435d7d2dd4056b36cd3fa3e83b7a96a7bfe2d2ed00d72d453ab6fc4a9082fcfebea793039b862c4bf168eab4a0d9cc461da25d82c6e8dd5b59c5cb1c4b3

data/emailparser.rb

@@ -0,0 +1,100 @@
+require 'pry'
+require 'json'
+require 'mail'
+
+class Emailparser
+
+	def initialize(message, attachment_dir)
+		@message = message
+		@attachment_dir = attachment_dir
+	end
+
+	# Voodoo to fix nasty encoded strings
+	def fix_encode(str)
+		if str.is_a?(String)
+		  return str.unpack('C*').pack('U*')
+		else
+		  return str
+		end
+	end
+
+	# Accepts a message
+	def parse_message
+
+		print "loading file: " + @message
+
+		email = Mail.read(@message)
+
+		# Defaults
+		source_file = @message.split("/").last
+
+		if email.subject
+			subject = fix_encode(email.subject)
+		else 
+			subject = "No Subject"
+		end
+		
+		body_plain	= ""
+		body_html	= ""
+		attachments = []
+
+		# Check for Multipart
+		if email.multipart?
+
+			# Parse Parts
+			email.parts.map do |part|
+				if (part.content_type.start_with?('text/plain'))
+					body_plain = fix_encode(part.body.decoded)
+				elsif (part.content_type.start_with?('text/html'))
+					body_html = fix_encode(part.body.decoded)
+				elsif (part.content_type.start_with?('multipart/alternative'))
+					print "has multipart/alternative\n"
+				else
+					print "has part: " + part.content_type + "\n"
+				end
+				print "-------------------------------------------------------\n"
+			end
+
+			body_plain = fix_encode(email.text_part.body.decoded)
+			body_html = fix_encode(email.html_part.body.decoded)
+
+		else
+			print "found single part email\n"
+			body_plain = fix_encode(email.body.decoded)
+			body_html = fix_encode(email.body.decoded)
+		end
+
+		# Handle Attachments
+		email.attachments.each do | attachment |
+			if (attachment.content_type.start_with?('image/'))
+				filename = fix_encode(attachment.filename)
+				attachments.push(filename)
+				print "found attachment " + filename + "\n"
+				#begin
+				#	File.open(@attachment_dir + filename, "w+b", 0644) do |f|
+				#		f.write attachment.body.decoded 
+				#	end
+				#rescue => e
+				#	puts "Unable to save data for #{filename} because #{e.message}"
+				#end
+			end
+		end
+
+		# Structure Data
+		email_data = {
+			source_file: source_file,
+			message_id: email.message_id,
+			from: email.from,
+			to: email.to,
+			cc: email.cc,
+			subject: subject,
+			body_plain: body_plain,
+			body_html: body_html,
+			attachments: attachments
+		}
+
+		email_json = JSON.pretty_generate(email_data)
+		return email_json
+	end
+
+end

metadata

@@ -0,0 +1,44 @@
+--- !ruby/object:Gem::Specification
+name: emailparser
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+platform: ruby
+authors:
+- Brennan Novak
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2016-05-16 00:00:00.000000000 Z
+dependencies: []
+description: Parses a signle email file to JSON
+email: hi@brennannovak.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- emailparser.rb
+homepage: https://github.com/TransparencyToolkit/EmailParser
+licenses:
+- GPL
+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: []
+rubyforge_project: 
+rubygems_version: 2.4.8
+signing_key: 
+specification_version: 4
+summary: Parses a single email file to JSON
+test_files: []