sosmii · August 6, 2020 12:28
diff --git a/parseml.py b/parseml.py
 #!/usr/bin/env python

 """
 2020 update:
 - More iterators, fewer lists
 - Python 3 compatible
 - Processes files in parallel
 (one thread per CPU, but that's not really how it works)
 """

 import email
 import glob
 import os
 from email.header import decode_header
 from multiprocessing import Pool

 EXTENSION = "eml"


 def extract(filename):
    """
    Try to extract the attachments from all files in cwd
    """
    # ensure that an output dir exists
    od = "output"
    os.path.exists(od) or os.makedirs(od)
    output_count = 0
    try:
        with open(filename, "r") as f:
            msg = email.message_from_file(f)
            for attachment in msg.get_payload()[1:]:
                raw_attachment_name = attachment.get_filename()
                # If no attachments are found, skip this file
                if raw_attachment_name:
                    files = decode_header(raw_attachment_name)
                    byte_attachment_name, encoding = files[0]
                    decoded_attachment_name = byte_attachment_name.decode() \
                        if type(byte_attachment_name) == bytes \
                        else byte_attachment_name
                    with open(os.path.join(od, decoded_attachment_name), "wb") as of:
                        of.write(attachment.get_payload(decode=True))
                        output_count += 1
            if output_count == 0:
                print("No attachment found for file %s!" % f.name)
    # this should catch read and write errors
    except IOError:
        print("Problem with %s or one of its attachments!" % f.name)
    return 1, output_count


 if __name__ == "__main__":
    # let's do this in parallel, using cpu count as number of threads
    pool = Pool(None)
    res = pool.map(extract, glob.iglob("*.%s" % EXTENSION))
    # need these if we use _async
    pool.close()
    pool.join()
    # 2-element list holding number of files, number of attachments
    numfiles = [sum(i) for i in zip(*res)]
    print("Done: Processed {} files with {} attachments.".format(*numfiles))
	#!/usr/bin/env python

	"""
	2020 update:
	- More iterators, fewer lists
	- Python 3 compatible
	- Processes files in parallel
	(one thread per CPU, but that's not really how it works)
	"""

	import email
	import glob
	import os
	from email.header import decode_header
	from multiprocessing import Pool

	EXTENSION = "eml"


	def extract(filename):
	"""
	Try to extract the attachments from all files in cwd
	"""
	# ensure that an output dir exists
	od = "output"
	os.path.exists(od) or os.makedirs(od)
	output_count = 0
	try:
	with open(filename, "r") as f:
	msg = email.message_from_file(f)
	for attachment in msg.get_payload()[1:]:
	raw_attachment_name = attachment.get_filename()
	# If no attachments are found, skip this file
	if raw_attachment_name:
	files = decode_header(raw_attachment_name)
	byte_attachment_name, encoding = files[0]
	decoded_attachment_name = byte_attachment_name.decode() \
	if type(byte_attachment_name) == bytes \
	else byte_attachment_name
	with open(os.path.join(od, decoded_attachment_name), "wb") as of:
	of.write(attachment.get_payload(decode=True))
	output_count += 1
	if output_count == 0:
	print("No attachment found for file %s!" % f.name)
	# this should catch read and write errors
	except IOError:
	print("Problem with %s or one of its attachments!" % f.name)
	return 1, output_count


	if __name__ == "__main__":
	# let's do this in parallel, using cpu count as number of threads
	pool = Pool(None)
	res = pool.map(extract, glob.iglob("*.%s" % EXTENSION))
	# need these if we use _async
	pool.close()
	pool.join()
	# 2-element list holding number of files, number of attachments
	numfiles = [sum(i) for i in zip(*res)]
	print("Done: Processed {} files with {} attachments.".format(*numfiles))