a brief info about wikibot and backend functionality

README.md

Wikibot

A simple, fast and scalable solution for an interactive chatbot.

• Bot live at : @wikibot on zulipchat
• Use Terminal : API
• Dashboard : link shared with mentors

Internal Working

How the bot thinks ?

The flow schema followed by the bot:

Scraping the information

The platform uses the following approach to extract useful information from text and provide the best fit response.

Example : What are some Python projects I could work on ? extracted keywords are ['python', 'projects', 'work']

Types of Messages

The following commands are processes by the bot :

Messages without prefix

• Can updated from dashboard (main category) example : Give me a list of Wikimedia's gsoc projects.

• Casual Talk : The bot is pre-programmed to perform casual talk with the users on general topics, such as greetings, activities, likes/dislikes, etc.

Messages with prefix (+)

• Alpha Command : Category names and messages can be created and managed through Alpha section on dashboard.
  • syntax : +<category_name > <query>
  • example : +lang python projects

• Beta Command : Messages with +<name> only. Can be created and managed through beta section on dashboard.
  • syntax : +<name >
  • example : +timeline

• Standard Command : Inbuilt command in code for predefined actions.
  • example : + help for help, + task <task_id> for info about a phab task

Other Features:

• Statistics Page : A statistics page on the dashboard to analyse the user-bot interaction.
• Project Suggestion : suggesting project details (such as project location gerrit/github/phabricator, mentors, workboard etc.) from skills.
• Find status of tasks assigned/created to/by the user.

Dashboard

A single page web app to create, update and manage the bot's knowledge base.

• link : shared with mentors
• dashboard images
• demo video

connect.py

# script to use the bot from terminal
# user parametres can be updated

import time
import requests
import json
import random

# you can update the details here if you want
user_id = '456121'
full_name = 'hesoyam'
short_name = 'hesoyam'
email = 'hesoyam@mymail.com'
# end

message_id = random.randrange(100000000, 999999999, 1)
timestamp = int(time.time())
url = 'https://openinternet.pythonanywhere.com/api'

payload = {
  "data": "",
  "message": {
    "id": message_id,
    "sender_id": user_id,
    "content": "",
    "recipient_id": 433815,
    "timestamp": timestamp,
    "client": "website",
    "subject": "",
    "topic_links": [],
    "rendered_content": "<p>Hi</p>",
    "is_me_message": False,
    "reactions": [],
    "submessages": [],
    "sender_full_name": full_name ,
    "sender_short_name": short_name,
    "sender_email": email,
    "sender_realm_str": "wikimedia",
    "display_recipient": [
      {
        "email": email,
        "full_name": full_name,
        "short_name": short_name,
        "id": user_id,
        "is_mirror_dummy": False
      },
      {
        "id": 274271,
        "email": "wikibot-bot@zulipchat.com",
        "full_name": "wikibot",
        "short_name": "wikibot-bot",
        "is_mirror_dummy": False
      }
    ],
    "type": "private",
    "avatar_url": "https://secure.gravatar.com/avatar/d8f78109444fc5ec90ca174923b580c9?d=identicon&version=1",
    "content_type": "text/x-markdown"
  },
  "bot_email": "wikibot-bot@zulipchat.com",
  "token": "v3BmJRGVB4Ec6pTOTi5gEqghSZKWQ275",
  "trigger": "private_message"
}


print("-> Currently the API is open without any key or authentication (limited sharing)")
print("-> The latency may be higher as compared to using @ wikibot with zulip because of direct connection")
print("-> Use wikibot on zulip for lower latency")
print("-> You can change user_parametres in the file")
print("-> Quit anytime using 'quit' or 'q'")
print("-> Please share you feedback :)")
print("\n")

while True:
  message = input(short_name + ' : ')
  if message == 'quit' or message == 'q':
    print('User closed the chat session')
    break
  
  payload['message']['content'] = message
  payload['data'] = message

  
  headers = {'content-type': 'application/json'}
  start = time.time()
  response = requests.post(url, data=json.dumps(payload), headers = headers)
  end = time.time()
  latency = '(latency : ' + str(int((end-start)*1000)) + ' ms, use wikibot on zulip for low latency)'
  
  response = response.json()
  print('wikibot : '+response['content'])
  print(latency)

wikibot.md

Adding New Category

Delete Category

Add New Entry

Edit Entry

Delete Entry

Beta Page

Error Alert

Home