Search.IO

A Discord bot search engine implemented in Python

To use:

Install requirements with:

pip install -r requirements.txt

Obtain Discord bot credentials and place in a .env file in the same directory as main.py

DISCORD_TOKEN = [YOUR TOKEN]

Download and add to PATH: geckodriver and firefox browser

Run main.py

To develop:

The bot uses a modularised system were individual search modules are hooked up to main.py. Each module consists of a class named [Search Engine]Search, with a constructor:

def __init__(
    self,
    bot: commands.Bot,
    ctx: commands.Context,
    server_settings: dict,
    user_settings: dict,
    message: discord.Message,
    args: list,
    query: str
):
    self.bot = bot
    self.ctx = ctx
    self.serverSettings = server_settings
    self.userSettings = user_settings
    self.message = message
    self.args = args
    self.query = query
    return

and an async __call__ method.

Modules are self sufficient, handling all search engine API calls and results displays. Before the module is called, a discord.Message instance is created, containing a loading animation. All attempts should be made to contain search results to that one message instance. After the search concludes and there is an associated single URL with the search (i.e. a Google search page or website URL), the result is logged as follows:

from src.utils import Log
Log.append_to_log(self.ctx, f"{self.ctx.command} result", [url variable])

All modules display their results in a discord.Embed, with the title linking to the relevant url and the discription displaying a brief summary of the results. Any additional information is displayed in embed fields. The footer always contains "Requested by: [Discord User and Discriminator]" The embeds are immediately followed by a 🗑️ reaction to allow the user to delete their search results.

If any Exceptions occur outside of the normal ones (asyncio.TimeoutError, etc), they are sent to a handler as follows:

from src.utils import error_handler
except Exception as e:
  await error_handler(self.bot, self.ctx, e, self.query)
  return

Single-option modules, such as youtube.py, immediately display the results to the user, and are limited to a maximum of 10 results. Each result is contained within a separate embed and all results are precached as a list before results are sent to the user. The user is allowed to navigate between results using ◀️ and ▶️ reactions.

Multi-option modules, such as wikipedia.py, display a multipage listing of all the results to the user to choose from. This listing is organised into a newline-separated string of [index number]: [result name]. Users reply to the result with their chosen index number and the listing is replaced with further information on the result of their choice.

Both servers and users have settings that must be saved in between bot run instances. These settings are in the form of dicts and are saved to .yaml files whenever their are changed and every hour.

The server dict is structured as follows:

  self.serverSettings = {
    guildID=str: { #hex code of  the guildID
      adminrole: roleID=int OR null,
      blacklist: [userID=int],
      sudoer: [userID=int],
      safesearch: bool,
      commandprefix: char,
      searchEngines: {
        #search engine modules: bool
      }
    }
  }

The user dict is structured as follows:

  self.userSettings = {
    userID=int: {
      downloadquota: {
          dailyDownload: float
          lifetimeDownload: float
          updateTime: ISO-8601 datetime str
        }
      level: {
        rank: int
        xp: int
      }
      locale: str OR null
      searchAlias: str OR null
    }
  }