#!/usr/bin/python

import os, argparse, hashlib, email, email.header, glob, sqlite3

def getheader(header_text, default="ascii"):
	headers = email.header.decode_header(header_text)
	header_sections = [unicode(text, charset or default) for text, charset in headers]
	return u"".join(header_sections)
	
def find_submessages(message):
	if message.is_multipart():
		return [find_submessages(part) for part in message.get_payload()]
	else:
		return message

def flatten(x):
	# http://kogs-www.informatik.uni-hamburg.de/~meine/python_tricks
	result = []
	for el in x:
		if hasattr(el, "__iter__") and not isinstance(el, basestring):
			result.extend(flatten(el))
		else:
			result.append(el)
	return result

def get_charset(part):
	if part.get_content_charset():
		return part.get_content_charset()
	elif part.get_charset():
		return part.get_charset()
	else:
		return "ascii"

print sqlite3.version
parser = argparse.ArgumentParser(description='Parses emails into an SQLite database, and optionally renders static HTML files.')

parser.add_argument('-p', '--pattern', dest='pattern', action='store', default='*',
                   help='glob pattern (including path) that has to be matched for a file to be parsed')
                   
parser.add_argument('-r', '--render', dest='render', action='store_true',
                   help='render static HTML files using the template files in templates/')

args = parser.parse_args()
options = vars(args)

database = sqlite3.connect('emails.db')
cursor = database.cursor()

try:
	# Try to create the table
	cursor.execute("CREATE TABLE emails (`message_id`, `from`, `to`, `subject`, `date`, `body`, `html`, `hash`)")
except sqlite3.OperationalError:
	# Table already exists
	pass

file_list = glob.glob(options['pattern'])

for email_file in file_list:
	message = email.message_from_file(open(email_file))
	
	if message['message-id'] is None:
		print "%s is not a valid e-mail file." % email_file
	else:
		if 'subject' not in message or message['subject'] is None:
			subject = ""
		else:
			subject = message['subject']
		
		textbody = ""
		htmlbody = ""
		
		sha1_hash = hashlib.sha1("%s/%s/%s/%s" % (message['from'], message['to'], message['message-id'], subject)).hexdigest()
		
		message_parts = [find_submessages(message)]
		message_parts = flatten(message_parts)
		print message_parts
		
		for part in message_parts:
			if part.get_filename() is None:
				# Part of the message
				if part.get_content_type() == "text/plain":
					if textbody == "":
						textbody = part.get_payload(decode=True).decode(get_charset(part))
				elif part.get_content_type() == "text/html" or part.get_content_type == "text/xhtml+xml" or part.get_content_type == "application/xhtml+xml":
					if htmlbody == "":
						htmlbody = part.get_payload(decode=True).decode(get_charset(part))
				else:
					print "WARNING: Unknown message format encountered in %s, message may be incomplete." % sha1_hash
			else:
				# Attachment
				print "Attachment found of type %s: %s" % (part.get_content_type(), part.get_filename())

		
		timestamp = 0
		
		new_row = (getheader(message['message-id']), getheader(message['from']), getheader(message['to']), getheader(subject), timestamp, textbody, htmlbody, sha1_hash)
		cursor.execute("INSERT INTO emails VALUES (?, ?, ?, ?, ?, ?, ?, ?)", new_row)
		print "Successfully parsed and inserted e-mail with SHA1 hash %s." % sha1_hash

database.commit()
print "Changes successfully committed to database, exiting..."