acosoi/gmail-email-scheduler

Simple Python script for sending GMail drafts at scheduled times

gmail-email-scheduler

Simple Python script for sending GMail drafts at scheduled times. This is a bare-bones implementation, if you want to use this you'll have to set up your own environment - a virtual machine in AWS that runs continuously, or perhaps a Raspberry Pi that hosts the script.

You must have technical knowledge in order to set the script up, this isn't designed to be user-friendly.

How this works:

• The script periodically goes through all saved drafts on a GMail account.
• For every draft, check the schedule.csv file to see if that e-mail is scheduled to be sent at a particular time.
• If a draft is scheduled to be sent and the time is in the past, send the e-mail.

Setup

• Install Python 3 ( tested with Python 3.5.1), including pip.
• Install developer headers (sudo apt-get install python3-dev).
• Follow the steps described in this Google Developer article in order to set up your credentials file (credentials.json).
• Install dateparser for Python (pip install dateparser).

How to use the script

• After performing the steps in the previous section, you should have a new file called credentials.json in your project folder.
• Replace instances of Europe/Bucharest in the script with your desired timezones.
• Start the script and leave it running forever.
• The first time it runs, you'll need to authorize access it by logging into Google and following the wizard.
• Create the draft you want to schedule.
• Add the draft to schedule.csv (lines are separated using |, so add lines such as Draft title|12/25/2020 15:00).

Want to run in AWS?

These were the steps I took to set up an EC2 instance to run the script continuously:

• Created a t3a.nano instance with a persistent 8 GB EBS storage using the Linux 2 AMI. You don't need high performance for the script, so I went with the cheapest option.
• SSH-ed into the virtual machine.
• Installed Python 3 and all of the dependencies mentioned above in the "Setup" section.
• Installed requirements for setting up Dropbox on the machine (yum install mesa-libGL).
• Installed a headless Dropbox instance as described on the official website using cd ~ && wget -O - "https://www.dropbox.com/download?plat=lnx.x86_64" | tar xzf
• Started the Dropbox daemon from ~/.dropbox-dist/dropboxd &. Used the provided link in order to create / login into the Dropbox account on a separate machine (one with a web browser).
• Ensured the path to the schedule is correct in application.py.
• Copied the token.pickle and credentials.json files from my local machine to the EC2 (after running the script locally first).
• Started the application using python3 ~/gmail-scheduler/application.py &.

Differences from Google's scheduling feature

The advantage of Google's solution is that it's beautifully integrated with the GMail clients and it's very easy to use and intuitive for non-technical users. The only advantage of my solution (and why I still use it) is that you can continue editing drafts even after scheduling them to be sent (once you mark an e-mail as scheduled in Google, you can't edit it without removing the schedule).