gmail-logger

Log metadata about your Gmail inbox.

Motivation

I wrote a blog post, Three Years of Logging My Inbox Count, in which I explore what this data says about my email habits. You can use my gmail-graphs code to generate the same graphs from your own data.

Quick start

• Add a file called secret.py which looks like this. If your Google account has 2-step verification, create a new app-specific password for this.

  email = 'your-email@gmail.com'
  password = 'your-password'
  

• Make sure you have pip and virtualenv.

• Set up a virtualenv, activate it, and install requirements:

  virtualenv venv
  . venv/bin/activate
  pip install -r requirements.txt
  

• Test it out:

  python log_inbox.py
  

• Schedule it to run regularly. This is the line from my crontab which schedules it at the top of every hour:

  0 * * * * cd /path/to/log/to; /path/to/gmail-logger/venv/bin/python /path/to/gmail-logger/log_inbox.py >/dev/null
  

Output

If you were to run it hourly, inbox_count.log would look something like this:

1428130801	2015-04-04 00:00:01	48
1428134401	2015-04-04 01:00:01	49
1428138001	2015-04-04 02:00:01	49
1428141601	2015-04-04 03:00:01	50
...

...and every hour, a JSON file named something like 2015-04-04_00.00.01.json would be produced:

[
  {
    "date": "Mon, 30 Mar 2015 14:06:20 -0700",
    "thread_id": "1497103965428563028",
    "unread": false,
    "from": [
      "Barack Obama <b@whitehouse.gov>"
    ],
    "subject": "Supreme Court nomination"
  },
  ...
]

Notes

I wrote this hastily, for my personal needs, and with the goal of logging as much data as reasonably possible. Eventually I'll update it to log in a saner format and offer options for what gets logged. It's not the best format, but it's unlikely to change, since gmail-graphs expects this format to compute summary data and generate graphs.

Add to my inbox count: mark@warkmilson.com