A few months ago at a random Chinatown bar in New York City, Tyler Richards and I were having a conversation about journaling. I told him that I had over 4 years worth of consistent journal entries and Tyler, being the data-minded person he is, gave me an idea to reinforce the lessons and insights from these journaling sessions.
The Idea: A system that emails me a random journal entry every day so I could see exactly what was going on in my mind at that given time.
The Data: ~600 journal entries with content length ranging from cool 1-liners to 3 pages of word vomit
My journal entries are very free-form and stream-of-consciousness so I’m not totally sure if I’ll end up learning anything useful, but I was curious to find out. The first step? Create a database to hold all of my journal entries.
1. Creating the Database
Holding all of my entries in a way that could be easily & efficiently accessed while also providing me the flexibility to add new entries if needed, SQLite was the tool for the job. This tutorial provides a good introduction to understanding how to create and query a local SQLite database with Python.
2. Convert my Google Doc files into usable data
The next step was aggregating all of my Google Doc files so that I could go through all of them with a script to extract all entries. Thankfully, I had some semblance of organization with which I separated entries — before each entry, I have a date. Sometimes it’s mm/dd/yy, sometimes dd/mm/yy if I thought I was cool for being in Europe, but there was always a date.
Once I had all of my files in a folder, I had to figure out how to go through each file and extract every entry. After reading this lovely post, I decided to go with the docx Python library to read each file paragraph-by-paragraph and create a new entry in my database every time I ran into a new date.
3. Send out a random entry every day
The next step was figuring how to send myself an email with an entry that was fetched from the database created earlier. This guide went through that process, which is surprisingly easy!
The final step was scheduling a recurring task that would run this code every morning automatically. Thankfully, Mac has a built-in system for this called crontab. One issue I ran into was that crontab tasks only run when the computer is active. This was fixed by waking my laptop a minute before the 10am slot when the job was scheduled to run, detailed here.
All in all, I learned how to work with local SQLite databases, use Python to leverage data, and schedule recurring tasks on Mac. Cool. So what now? I’m going to spend the next couple of months reading a random journal entry daily while traveling to see if I can glean any unique insights from my ramblings.
See you on the other side and stay tuned for my analysis 🤙🏽
My full code can be found here.