Sean Leavey: Documentation

Downloading tagged photos from my Facebook profile

Searching the internet for a way to download tagged photos was depressingly, but all too predictably, full of scam apps and ancient scripts that had long since stopped working. Facebook don't make it easy to get these photos, which I assume is intentional. Most of the supposed solutions seem to involve running fragile extensions or scripts inside your browser while you have Facebook open which simulate clicks to sequentially load and save your photos. Giving some random person on the internet the power to do stuff in your browser is dangerous, but the one script that I trusted enough to try turned out not to support downloading of tagged photos, only my own albums which I didn't have a problem getting via the Facebook data dump feature. It was at this point that I stumbled upon a 10 year old script that used the Facebook API itself to grab tagged photos. This seemed promising because APIs are clearly defined ways of getting specific data from web services, not subject to the ephemeral nature of Facebook's HTML. The script required a Facebook API token, which meant first registering to become part of the Facebook developer programme (free and easy) and creating an app. Being a Python guy I rewrote the script and tried running it but found some of the API calls were no longer valid. Even APIs change over 10 years. Luckily after reading the API documentation I was able to figure out what was wrong and fix it - and it worked... or at least I thought it did. The first few photos were clearly photos of me but taken by someone else, so I thought it was working - but upon making a more thorough check I realised these photos were still from my account, so I must have obtained these from the photographers and uploaded them myself. To get a copy of others' photos of me, I concluded I had to do it the hard way: opening up the "Photos of Me" lightbox, clicking "Download" on each one, pressing the right arrow key, and... repeating many hundreds of times. In total I downloaded 500 files this way over the course of about an hour... not something I'd like to repeat, but fortunately I don't have to.

I've documented what I did anyway, in case someone finds a way to make it work for all tagged photos. I've written some instructions below for anyone wishing to attempt it.

The Facebook API browser. My API token (expired in this picture) let me grab JSON-encoded data, in this case my tagged photos.

Instructions

Doing what I did involves a few steps and will take short while to set up. You'll need a Facebook account (obviously) but also a recent version of Python (at least 3.6) with a library called requests and a system tool available on Linux and OSX called wget. Search the internet for details on how to install these for your operating system. This will still be possible on Windows but will probably need some changes to the final script, which I've not done but are perfectly possible for others to do. In summary, you have to:

1. Register as a Facebook developer
2. Register a new app
3. Specify that the app needs access to photos, and grant it such access for your profile
4. Generate an API token
5. Run the script using Python

Details for each step below...

Register a new app

Go to https://developers.facebook.com/apps/ and click "Create App". A pop-up appears:

Choose "Something Else" then enter a name for the app. I called mine "Sean App" - this is just temporary, and won't ever be released. Leave the other settings as-is and click "Create App". You might be asked to complete a CAPTCHA.

Specify the permissions your app needs

Once you've made your new app, go to the Graph API Explorer. You'll see a random string of letters and numbers under "Access Token" - this is the special token you need to give to the script below to let you access Facebook's data. But this token does not have permission to see photos - so we need to add this as a permission first. Under "Permissions", search for and select "user_photos":

Generate an API token

Click the "Generate Access Token" button and you'll be presented with a screen that will be familiar to you if you've ever linked an app to your profile before:

Click "Continue as [x]" where "[x]" is you. This will then give you a new API token on the Graph Explorer page that you can use to grab your photos.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70

"""Facebook tagged photos download script.

I, Sean Leavey, hereby release this script into the public domain with
no warranty.

Sean Leavey
https://attackllama.com/2021/01/downloading-tagged-photos-from-facebook/

For help, email me (see https://attackllama.com/contact/).
"""

import os
import pickle
import datetime
import time
import requests

SLEEP = 3600 // 200  # One hour / max requests per hour.
ACCESS_TOKEN = "replace-with-long-API-token-string"
OUTPUT_DIR = "/tmp"
PICKLE_FILE = "/tmp/photos.pickle"

def fetch():
    if os.path.isfile(PICKLE_FILE):
        print("Using existing photo list")
        with open(PICKLE_FILE, "rb") as fobj:
            data = pickle.load(fobj)
    else:
        print("Dumping photo list")
        url = f"https://graph.facebook.com/me/photos"
        result = requests.get(
            url, params={"limit": 10000, "access_token": ACCESS_TOKEN}
        )
        result.raise_for_status()
        data = result.json()["data"]

        with open(PICKLE_FILE, "wb") as fobj:
            pickle.dump(data, fobj)

    ndata = len(data)

    for i, photo in enumerate(data, start=1):
        print(f"{i}/{ndata}")

        imagedate = datetime.datetime.strptime(
            photo["created_time"], "%Y-%m-%dT%H:%M:%S%z"
        )
        imagepath = f"{imagedate.strftime('%Y%m%d%H%M%S')}-{photo['id']}.jpg"
        filename = f"{OUTPUT_DIR}/{imagepath}"

        if os.path.isfile(filename):
            print(f"Skipping already downloaded file {filename}")
            continue

        image_url = f"https://graph.facebook.com/v9.0/{photo['id']}"
        image_result = requests.get(
            image_url,
            params={"fields": ["images"], "access_token": ACCESS_TOKEN}
        )
        image_result.raise_for_status()
        images = image_result.json()["images"]

        os.system(f"wget -O '{filename}' '{images[0]['source']}'")

        print(f"Sleeping for {SLEEP} s")
        time.sleep(SLEEP)


if __name__ == "__main__":
    fetch()