benlk
commented
7 years ago Doing an array isn't particularly friendly to contributors; what would be neat is a CSV that gets parsed, and the CSV itself is pulled from a Google Sheet that accepts submissions. Periodically dump the Sheet into the CSV, then use something to parse the CSV to get date, text, link.
Parts:
- [x] Google Form feeding into Google Sheet: https://docs.google.com/forms/d/e/1FAIpQLSf5gxPrlEIVNM4d_W5Pu1TwHO_1LJFY-IbTCW7FEaxM25ne0w/viewform
- [x] Google Sheet https://docs.google.com/spreadsheets/d/1Gx3KrCzldP7uYOi3lp6aZSkIZ80C_oZKsaj2CeKSBUM/edit#gid=1275143625
- field for date in ISO 8601
- field for text
- field for link to the thing
- [x] thing to download list of dates of things: #4
- [ ] thing to parse list of dates of things and only keep approved items from the spreadsheet (essentially: only keep form entries that have been approved): #5
- [ ] thing to convert list of dates of things into list of tweets for dates: 18mo out, 12mo out, 6mo out, so on. #6
- [ ] thing to parse tweets-for-dates CSV to find today's messages, send each message into
t: #7
Parsing tweets-for-dates spreadsheet sounds like https://copytext.readthedocs.io/en/stable/
Fetching the spreadsheet is gonna be more complex; might just be a manual task to run for the first few weeks.
This would be a separate script, added to the main cron job script but with a 10s delay so it runs after the main tweet of day.
Stretch goal is to get the tweet ID from the countdown script and have it be the reply-to of this tweet.
Credit to @thebestsophist