Extract attachments from EML files in the current dir, and write them to the output subdir

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 glob
import os
import email
from email import policy
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, policy=policy.default)
            for attachment in msg.iter_attachments():
                try:
                    output_filename = attachment.get_filename()
                except AttributeError:
                    print("Got string instead of filename for %s. Skipping." % f.name)
                    continue
                # If no attachments are found, skip this file
                if output_filename:
                    with open(os.path.join(od, output_filename), "wb") as of:
                        try:
                            of.write(attachment.get_payload(decode=True))
                            output_count += 1
                        except TypeError:
                            print("Couldn't get payload for %s" % output_filename)
            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))

Works like a charm!
Thank you :)

Nice, thanks!

Hi @urschrei thank you very much !! very useful

Hi @e-tplus

I had the same problem which happened with .eml embedded in .eml.
For some shady reason, the parser is unable to process adequately .eml attachments.

But at least you can open it as a string or bytes, and you notice that there is kind of weird header like:
"Content-Type: message/rfc822
Content-Disposition: attachment;
creation-date="Fri, 03 Sep 2021 07:45:44 GMT";
modification-date="Fri, 03 Sep 2021 07:45:44 GMT""

So I tried to remove it, by simply removing the first 162 characters of attachments when the attachment is detected with NoneType:
attachment.as_bytes()[162:]

Also, I added a recursive call to extract, to extract the attachments in .eml embedded in .eml

Please find below the reviewed function (working on my side):

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, policy=policy.default)
            for attachment in msg.iter_attachments():
                try:
                    output_filename = attachment.get_filename()
                except AttributeError:
                    print("Got string instead of filename for %s. Skipping." % f.name)
                    continue
                # If no attachments are found, skip this file
                if output_filename:
                    with open(os.path.join(od, output_filename), "wb") as of:
                        try:
                            of.write(attachment.get_payload(decode=True))
                            output_count += 1
                        except TypeError:
                            print("Couldn't get payload for %s" % output_filename)
                **#for EML embedded in EML
                else:
                    output_filename = filename.split('.')[:-1][0].split('\\')[-1]+'_void_'+str(output_count)+'.eml'
                    with open(os.path.join(od, output_filename), "wb") as of:
                        try:
                            of.write(attachment.as_bytes()[162:])
                            output_count += 1
                            extract(os.path.join(od, output_filename))
                        except TypeError:
                            print("Couldn't get payload for %s" % output_filename)**
            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

only works with python3

got this error when running using python ImportError: cannot import name policy

Thankyou so much this helped me process all those realestate emails and get the rent receipts thankyou

It fails when there are files with the same name and ext. for this I added the date of the email eg. "15 Apr 2021".

if output_filename:
with open(os.path.join(od, str(msg['date'])[5:16] + " " + output_filename), "wb") as of:
try:
of.write(attachment.get_payload(decode=True))
output_count += 1
except TypeError:
print("Couldn't get payload for %s" % output_filename)

EXTREMELY useful, much thanks!

@urschrei
Many thanks.
But I find this code cannot extract .msg files which are from outlook app. When I attach one eml and one msg, only eml file is saved.

Worked great! Thank you for your hard work

Didn't work for me for embedded eml inside eml, with pdfs in, this did:
(Note that it only seeks out pdf files)

#!/usr/bin/python3
import os
import sys
import email
from email import policy
from email.parser import BytesParser
from email.iterators import typed_subpart_iterator

def extract_attachments(email_message, output_folder):
    for part in typed_subpart_iterator(email_message, 'application', 'pdf'):
        filename = part.get_filename()
        if not filename:
            continue
        filepath = os.path.join(output_folder, filename)
        with open(filepath, 'wb') as f:
            f.write(part.get_payload(decode=True))

def parse_email(file_path, output_folder):
    with open(file_path, 'rb') as f:
        msg = BytesParser(policy=policy.default).parse(f)
        if msg.is_multipart():
            for payload in msg.iter_parts():
                if payload.get_content_type() == 'message/rfc822':
                    extract_attachments(payload.get_payload(0), output_folder)
                elif payload.get_content_type() == 'application/pdf':
                    extract_attachments(msg, output_folder)
        else:
            extract_attachments(msg, output_folder)

if __name__ == "__main__":
    file_path = sys.argv[1]
    output_folder = sys.argv[2]
    os.makedirs(output_folder, exist_ok=True)
    parse_email(file_path, output_folder)

First argument is the .eml and second is where you want the files extracted.

I wanted to take a moment to express my gratitude for the Python code you wrote for extracting attachments from EML files. Your solution has been incredibly helpful and efficient for my needs. The way you handled the EML format and attachment extraction is impressive and much appreciated.

Thank you so much for your time and effort. It’s clear that a lot of thought went into creating this code, and I’m genuinely grateful for your contribution.

Just worked. Thank you ! (Probably GitHub should add a comment box to repositories as well. Sometimes you just want to say thanks and there is no simple way to do it.)