Describe the bug
Trying to run after setting up the config.json and installing all of the packages called out in the requirements.txt
Experienced the following traceback in the screenshot below

Screenshots

Desktop (please complete the following information):

• OS: Windows 10 22H2
• Python v3.10.11

I'm getting Notified of appointments by the script. When I login to book them, they don't exist, and the script continues to report their existence.

To Reproduce
Steps to reproduce the behavior:
1: Add a config for location 13321 and date 7th June 2023 (for example)
2: You will receive notification of available appointments from the script
3: Login to https://ttp.cbp.dhs.gov/ and go to reschedule interview
4: for the location configured, note that no appointments are available

Expected behavior
Appointments reported by script should exist in scheduler UI

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: Windows 11
• Browser Tested with Edge, Firefox, Chrome
• Version Latest versions

Smartphone (please complete the following information):

• Device: iPhone 13 Pro
• OS: iOS 16.5
• Browser Safari, Firefox, Edge
• Version Latest versions

Additional context
Could be location specific - I tested this with another location and the reported appointment did exist in the UI.

My config looks like this:

{
  "current_appointment_date": "June 7, 2023",
  "location_ids": [ 13321 ],
  "notification_level": 1,
  "notification_urls": [ "discord://webhookurl" ],
  "retrieval_interval": "5m",
  "start_appointment_time": "00:00",
  "end_appointment_time": "23:59"
}

I am running the script via Docker. Container image ecsouthwick/trusted-traveler-scheduler:latest retrieved 2023-05-24. I have also run the script locally cloned from this repo with the same results.

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• chore(deps): update dependency requests to v2.32.2 [security]
• chore(deps): update dependency apprise to v1.8.0
• chore(deps): update docker/build-push-action action to v6
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

Currently the code only ever adds (location_id, start_time) pairs to the sql database, and uses this to determine if an available appointment is newly-seen or not. This results in missing the time slot if the appointment goes through a process like:

1. appointment is newly seen and reported
2. user is beaten to the appointment by someone else / appointment is unavailable for some reason
3. appointment slot becomes available again
   At 3) above, the appointment, which should be treated as new availability, is not reported to user again.

It's probably impossible to foolproof this (unless the dhs.gov API exposes some more functionality), but it would at least be nice to clear out db entries if they are missing from a newly returned API request in _get_schedule.