, _
 () o  _,        _,  |\    /|/_)o |\  _|   _          _   ,_ _|_
 /\ | / | /|/|  / |  |/     |  \| |/ / |  |/ /\/ |/\_/ \_/  | |
/(_)|/\/|/ | |_/\/|_/|_/    |(_/|/|_/\/|_/|_/ /\/|_/ \_/    |/|_/
       (|                                       (|

Why «Signal Bildexport» had to be done

Ever wanted to save images from Signal messages? It is a chore:

• Every single photo needs to be exported separately
• All metadata is lost – and even with a wrong Timestamp. This messes up your (iCloud) photo timeline!

Unlike Whatsapp, Signal does not provide automatic saving of received images to the camera roll. The discussion is still going on:

• Github Feature Request #1567: Option to auto-save received images/video to camera roll
• Signal Community Discussion Thread: Automatically save attachments to device

Due to this hassle, friends and family were threatening to switch back from Signal to Whatsapp. Things were getting serious! But, fear not:

What «Signal Bildexport» does

Since the images cannot be saved directly from Android/iOS, this script exports photos from Signal Desktop on Mac or Windows to a chosen folder or iCloud Photos

1. Reads Signal Desktop's encrypted sqlite database
2. Collects all messages which were received (since Signal Desktop was installed)
3. Exports every found JPEG Photo to the iCloud Photos folder (Windows) or the Photos App (Mac) with the following metadata annotations:
   • Timestamp (the closed we can get: the instant when the picture was sent)
   • Who send it (Full contact name or Signal Name, if not in contacts)
   • Name of the Chat group
   • The message that came with it as the image's title

Usage

cd signal-bildexport/
python export.py

On the first run, config.yaml will be created. The defaults should be fine, but it allows you to restrict which photos should be exported:

import_photos_from_messages:
  any_with_my_reaction: ['🤩'] # only export images for which your reaction was 🤩
                               # use ['*'] as a wildchar, or an empty list ('[]') to
                               # disable filtering by reactions
  in_conversation:
    exclude: [] # Add conversations from which you don't want to import anyting (work?)
    include:
    - '*' # Every other conversation. Feel free to move the '*' to 'exclude'
          # and specify the desired chats here
last_run: 0 # epoch timestamp of last run. Set to 0 to trigger a full export
output_path: images # output path for the images (different for Mac and Windows)
x-advanced:
  pre_copy: false
  pre_decrypt: false
  signal_path: /Users/matthiasroggo/Library/Application Support/Signal
  sqlcipher_bin: ./sqlcipher/bin/sqlcipher # Only required if pre_decrypt = True (Windows)

Installation on Windows using WSL

Prerequisites

• iCloud for Windows is installed
• Signal Desktop is installed and configured (and some Photos have been received already)

Powershell

# Install Windows Subsystem for Linux
wsl --install

Ubuntu Shell (Bash)

# Clone repo

# Choose a directory for this script.  
# Make sure the path does not contain any spaces! Eg:
cd /mnt/c/users/<username>

git clone https://github.com/bskp/signal-bildexport.git
cd signal-bildexport/

# Install build tools and build sqlcipher
sudo apt-get update  
sudo apt install build-essential libsqlite3-dev tclsh libssl-dev

# Feel free to put sqlcipher anywhere else. Make sure to adapt
# config.yaml (which will be generated) accordingly.
git clone https://github.com/sqlcipher/sqlcipher.git
cd sqlcipher/
./configure --enable-tempstore=yes CFLAGS="-DSQLITE_HAS_CODEC" LDFLAGS="-lcrypto -lsqlite3"
make

# Install required python modules
cd ..
apt-get install python3-pip  
pip install -r requirements.txt

Installation on Mac

Prerequisites

• Signal Desktop is installed and configured (and some Photos have been received already)
• Homebrew is installed

Terminal (zsh or bash)

# Clone repo
git clone https://github.com/bskp/signal-bildexport.git
cd signal-bildexport/

# Install sqlcipher
brew install sqlcipher

# Install required python modules
pip install -r requirements.txt
pip install pysqlcipher3

Thanks

While getting pysqlcipher3 to work was a walk in the park on Mac, it almost got me quitting on Windows. Big thanks to Chris Arderne for his fantastic write-up on solving this for signal-export, where the pre-decrypt approach is taken from!