enplotz / mensa_ukon Goto Github PK

Ideally, we would like stdout to really just contain errors (or warnings?) and everything else go to the log file(s). pygogo seems like a nice to use library.

Maybe it's a good idea to divide the log into multiple files:

• common.err.log <- all libraries except mensa and the bot
• mensa.err.log <- mensa library errors (warnings?)
• mensa.log <- mensa info (or debug)
• bot.err.log <- bot errors
• bot.log <- bot info (or debug)

Find out if we can respond to a command msg that was edited and edit our response accordingly.

Works:
/mensa
/wok today
/grill 2016-07-14
/grill banana

Does not work:
/mensa today
/mensa 2016-07-14
/mensa banana

Order meals by custom order.

Options:

• globally
• per user account
• per chat

• Upgrade python-telegram-bot to latest version!

Overview over the whole week for specific meals would be nice. For now we could just send the weekly plan as PDF. The plan for next week is also only available as PDF.

We could use something like:

bot.sendDocument(chat_id=chat_id, document=plan.pdf)

Feature to subscribe to daily/weekly messages directly from the bot.

• configurable interval
• configurable message time (morning, right before lunch time)
• per chat/user

Possible commands:

    /subscribe giessberg @daily 8:00
    /subscribe giessberg @weekly 8pm
    /unsubscribe

Currently, the bot polls for new updates. I find this kinda stupid, as the bot framework offers a web hook API where the Telegram server pushes updates to. However until now I could not figure out how to make this work, as one needs to submit a certificate matching the hooks url (only HTTPS is allowed).

The bot could use some administrative capabilities (like configure logging levels or reload the logging config). We could look into ownbot for that.

Persistent settings for

• bot message locale (de, en)
• canteen plan listings locale (de & en are available)
  (- daily canteen plan subscriptions)

Cache responses.

• locally as pickles?
• per day? hourly?

• create necessary templates
• deploy on live instance