This repository parses conversation data from Google Hangouts and gives diagnostics on the number of messages in conversations. Two scripts are currently supported: parser.py and visualize.py. The parsing script parses raw JSON data from Google Takeout and creates pickled summary files for each conversation. The visualization script creates a histogram of messages over time using the pickled conversation summaries.

1. Clone this repository
2. Download your hangouts data
   • Navigate to Google Takeout
   • Choose "Select None" and manually select Hangouts
   • Download the data in zip format and move the Hangouts.json file into the raw folder in this repository
3. Install dependencies via pip
   • pip install -r requirements.txt
   • No dependencies are required for the parser.py script, but visualize.py requires the dependencies
4. Run the parser
   • Note: if you did not place your hangouts data as raw/Hangouts.json you can specify the path to the .json file as an argument to the parser.py script via the -f flag

python parser.py

5. Run the visualization
   • The <conversation_id> can be found in the output of the parser.py script

python visualize.py -f output/<conversation_id>.pkl

This code is freely available under the GNU Public License (GPL).

All of the data processing in these scripts happens locally on your computer. The data you provide to the script is NOT uploaded to an external server. Feel free to examine the code if you are concerned.

This repository was inspired by MasterScrat/Chatistics. Chatistics can parse Facebook Messenger and Telegram data, but not Hangouts group messages. I originally intended to contribute to that repository and add Hangouts group message support, but my design drifted far from the existing design in that repository so I created a new project. Shoutout to MasterScrat for great work and thanks for the inspiration!