tiagojpdias / pepe-as-a-service Goto Github PK

Problem:
Currently, users can invoke the bot as many times as they want, causing a flood if the command is abused.

Solution:
Implement a throttling mechanism so that each user has to wait N seconds until the bot stops ignoring him.

Workflow:

1. A user invokes the bot for the first time;
2. The bot records the user, the current timestamp and sends a random image;
3. The same user invokes the bot again;
4. The bot checks if there's a record of the user;
5. If so, it then checks if N seconds have passed since the last invocation;
6. If true, a random image is sent and the timestamp gets updated;
7. If not, the user is ignored;

Deploying and running Pepe-as-a-service is too hard for pepe.

Problem
Spreading default values when env variables aren't set

Solution
Create config files as middleware between env variables and the ones we use throughout the code. Something like Laravel's config files.

Improve code coverage

Problem

• When adding new tags to the JSON file we need to restart the bot so it can load the new map

Solution

• Check if the files checksum has changed and if so then load again the map

Problem:

• Whenever someones writes a similar tag the bot doesn't reply.

Solution:

• Add an aliases property to accommodate all the possible tags related

We're missing the 🤷‍♂️ tag and I bet there are a lot of shrugging pepes out there for us to use.

We currently have text strings hard coded in the code base.

As a first step into solving #9, we should yank them from the code and load them from language files, instead.

• Fill README
• Add instalation guide
• Add configuration guide
• Add a contribution guide
• Add license

• Fill the feeder with tags and images

Example:

{
  "feelsBadMan": [
    "url_for_image1",
    "url_for_image2",
  ],
}

So when we trigger the bot with "feelsBadMan" it will randomly fetch an image from the array