运行 python 函数的 Slack 机器人

Question

I am creating a Slack Bot, I have successfully installed it, although I am not 100% sure what does ngrok do to run it when I have node server running locally. I have followed this guide .

I have a python script that:

1. Connects to a database
2. Downloads the entire table and creates a pandas dataframe

In the script, I have declared many variables which takes specific information from the dataframe.

For example:

monday_sales = sum((df['sale_event_time']>=week) & (df['sale_event_time']<=today_real) & ((df['sale_event_time'].apply(lambda x:x.isoweekday()))==1))

Which simply returns how many sales I had on Monday from the dataframe.

My goal is to connect slack bot to my python script, basically being able to run /monday_sales in slack which would print the number that monday_sales variable holds.

I haven't used much of node or js, so linking everything together I find a bit difficult.

First I run ngrok like so:

./ngrok http 4390

Then I run node server locally(?)

var http = require('http');
const PORT=4390;
function handleRequest(request, response){
  response.end('My bot is ALIIIIIIVEEEEE');
}
var server = http.createServer(handleRequest);
server.listen(PORT, function(){
  console.log("Server listening on: http://localhost:%s", PORT);
});

And when I run /ngrok in slack it prints out 'My bot is ALIIIIIIVEEEEE, although only visible to me.

So the question could be split in to few:

1. Why do I need ngrok and node.js server running locally?
2. How can I trigger python functions / variables from slack bot.
3. Is there a better way of having slack bot only ran by python without node, ngrok?

Thank you very much for your suggestions.

Answer 1

I think your best bet is to ditch this tutorial and Node, as it seems you're only using it as a ngrok wrapper. Instead, since the rest of your bot is in Python, just use pyngrok (full disclosure, I am the developer), which is a Python wrapper for ngrok . It gives you programmatic access to the tunnels and everything else.

If this solves your use case, you can check out the docs here .