Python Markov Chain chatbot running on Telegram

Overview

Hanasubot

Hanasubot (Japanese 話すボット, talking bot) is a Python chatbot running on Telegram. The bot is based on Markov Chains so it can learn your word instantly, unlike neural network chatbots which require training. It uses a modified version of markovify library for that purporse. However, the output may not make sense at all, though it can sometimes generate hilarious replies.

In theory, the bot can learn in any languages, but for some languages word segmentation is required. The bot currently supports Chinese and Japanese word segmentation, with pkuseg, CkipTagger and mecab. Language detection relies on pycld2.

Hanasubot has a permission system so you can easily stop the bot learning from naughty kids in your group, while still reply them. Users with admin right can erase lines from bot corpus as well.

The bot is designed for Chinese Telegram groups so there are a lot of messages written in Chinese. I18n will happen in future and any help is welcome.

Installation

Python 3.6+ is required.

VENV_PATH=/path/to/your/venv  # Change this
python3 -m venv $VENV_PATH
source $VENV_PATH/bin/activate

pip3 install -r requirements.txt

If you are using Python 3.6, dataclasses 0.8 is required as well:

pip3 install dataclasses==0.8

For Python 3.7 and up, dataclasses is included so no need to install it.

To use CkipTagger for Traditional Chinese tokenization, you have to download the model file (see CkipTagger readme for a detailed guide):

python3 -c "from ckiptagger import data_utils; data_utils.download_data_gdown('./')"

Then unzip to a folder named ckipdata, in the same directory as the Python scripts.

Optionally, you can initialize the user dict for pkuseg and CkipTagger, before start running the bot:

touch ./pkuseg_dict.txt
touch ./ckip_dict.json

Configuration

Copy config.example.py and fill it out. Please check the comments in config file.

cp config.example.py config.py

After that, simply start the bot:

python3 tgbot.py

Bot commands and usage

Simply reply to the bot and it will say some random words if you have collected enough corpus. The bot will also learn from your message instantly. Special commands are as follows.

Require root

  • /reload_config - Reload config file without restarting the bot. Some entries cannot be dynamically reloaded though, see config.example.py for details.

Require admin

  • /erase - Remove lines from corpus. (Non-admins can only erase lines sent by themselves.)
  • /userweight - Set user weight.
  • /ban - Set user right to -1.
  • /restrict - Set user right to 1.
  • /grantnormal - Set user right to 2.
  • /granttrusted -Set user right to 3.
  • /grantadmin - Set user right to 4. Admins are able to add/remove other admins with above commands. See also the user right levels section.

Require trusted

  • /addword_cn - Add a word into pkuseg user dictionary.
  • /addword_tw - Add a word into CkipTagger user dictionary.
  • /rmword_cn - Remove a word from pkuseg user dictionary.
  • /rmword_tw - Remove a word from CkipTagger user dictionary.

Other commands

  • /clddbg - Test language detection of some texts.
  • /cutdbg - Test tokenization of some texts.
  • /policy - See what data is collected by the bot and so on.
  • /reload - Claim your admin rights after you get Telegram group admin.
  • /source - See the source code.
  • /start - Start chatting, useful when you can't find the bot messages to reply.

Database

Initialize

CREATE TABLE IF NOT EXISTS chat(
    chat_id integer PRIMARY KEY,
    chat_tgid integer NOT NULL UNIQUE,
    chat_name text
);
CREATE TABLE IF NOT EXISTS user(
    user_id integer PRIMARY KEY,
    user_tgid integer NOT NULL UNIQUE,
    user_name text,
    user_right integer DEFAULT 2,
    user_weight real DEFAULT 1.0
);
CREATE TABLE IF NOT EXISTS corpus(
    corpus_id integer PRIMARY KEY,
    corpus_time integer,
    corpus_line text NOT NULL UNIQUE,
    corpus_raw integer REFERENCES raw,
    corpus_chat integer REFERENCES chat,
    corpus_user integer REFERENCES user,
    corpus_weight real DEFAULT 1.0
);
CREATE TABLE IF NOT EXISTS raw(
    raw_id integer PRIMARY KEY,
    raw_text text UNIQUE
);

User right levels

  • 5 - root.
  • 4 - admin, can change user rights (except root users), can erase a line from corpus, and can set user_weight and corpus_weight (WIP).
  • 3 - trusted user, can feed the bot via private messages, and can add words into dictionary (for tokenization purposes).
  • 2 - normal user.
  • 1 - restricted user, bot will not write their messages into database.
  • -1 - banned user, bot will not reply to their messages.

TODOs

  • Let admins set corpus_weight
  • Batch /erase

License

MIT

Simple script to ban bots at Twitch chats using a text file as a source.

AUTOBAN 🇺🇸 English version Simple script to ban bots at Twitch chats using a text file as a source. How to use Windows Go to releases for further in

And Paiva 5 Feb 06, 2022
allow windows programs to call dssp/mkdssp command from wsl; rework biopython on windows (PDB -> dssp -> fasta)

dssp-wsl Converting PDB (Protein Data Bank) file format to DSSP file format is required for generating datasets of peptides and their secondary struct

Taine Zhao 1 Feb 23, 2022
SimpleDCABot is a simple bot that buys crypto with a dollar-cost averaging strategy.

Simple Open Dollar Cost Averaging (DCA) Bot SimpleDCABot is a simple bot that buys crypto on a selected exchange at regular intervals for a prescribed

4 Mar 28, 2022
Python wrapper for Xeno-canto API 2.0. Enables downloading bird data with one command line

Python wrapper for Xeno-canto API 2.0. Enables downloading bird data with one command line. Supports multithreading

_zza 9 Dec 10, 2022
CnCL - CnCLess it's an Easy to deploy Botnet without CnC/C2

CnCL CnCLess it's an Easy to deploy Botnet without CnC/C2, Harder to track and t

ZSendokame 2 Jan 10, 2022
NekoRobot-2 - Neko is An Anime themed advance Telegram group management bot.

NekoRobot A modular telegram Python bot running on python3 with an sqlalchemy, mongodb database. ╒═══「 Status 」 Maintained Support Group Included Free

Lovely Boy 19 Nov 12, 2022
Weather Tracker, made with Python using Open Weather API

Weather Tracker Weather Tracker, made with Python using Open Weather API

Sahil Kumar 1 Feb 07, 2022
☄️ High performance, easy to use and feature-rich Solana SDK for Python.

Solathon is an high performance, easy to use and feature-rich Solana SDK for Python. Easy for beginners, powerful for real world applications.

Bolt 28 Oct 10, 2022
Build better AWS infrastructure

Sceptre About Sceptre is a tool to drive AWS CloudFormation. It automates the mundane, repetitive and error-prone tasks, enabling you to concentrate o

sceptre 1.4k Jan 04, 2023
Discord bot for Shran development

shranbot A discord bot named Herbert West that will monitor the Shran development discord server. Using dotenv shranbot uses a .env file to load secre

Matt Williams 1 Jul 29, 2022
Detects members having unicode names. Public bot: @scarletwitchprobot

✨ Scarletwitch bot ✨ Detects unicode names members in a tg chat & provides a option to take action on that user ! Public bot: @scarletwitchprobot Supp

ÁÑÑÍHÌLÅTØR SPÄRK 18 Nov 12, 2022
ChannelActionsBot - Channel Actions Bot With Python

ChannelActionsBot Can be found on telegram as @ChannelActionsBot! Features Auto

Aditya 56 Dec 30, 2022
Export Statistics for a Telegram Group Chat

Telegram Statistics Export Statistics for a Telegram Group Chat How to Run First, in main repo directory, run the following code to add src to your PY

Ali Hejazizo 22 Dec 05, 2022
PS4RemotePKGSender - Use with Remote PKG Installer

PS4_Remote_PKG_Sender Used with the remote PKG installer on PS4 Thanks to the au

Teri 4 Sep 23, 2022
Mass-unscrobble Last.fm scrobbles based on artist, track title, or time of day of the scrobble.

Unscrobbler This program is designed to mass-unscrobble Last.fm scrobbles based on artist, track title, or time of day of the scrobble. For example, i

Nathan 6 Nov 04, 2022
Python Bot that attends classes, answers polls, and then again waits for classes to start.

LPU_myclass_Bot LPU_myclass_Bot is a Python bot that waits for class to start, attends class, answers polls, and then again waits for another class to

Saurabh Kumar 6 Apr 07, 2022
Infrastructure template and Jupyter notebooks for running RoseTTAFold on AWS Batch.

AWS RoseTTAFold Infrastructure template and Jupyter notebooks for running RoseTTAFold on AWS Batch. Overview Proteins are large biomolecules that play

AWS Samples 20 May 10, 2022
Netflix Movies and TV Series Downloader Tool including CDM L1 which you guys can Donwload 4K Movies

NFRipper2.0 I could not shared all the code here Because its has lots of files inisde it https://new.gdtot.me/file/86651844 - Downoad File From Here.

Kiran 15 May 06, 2022
Biblioteca Python que extrai dados de mercado do Bacen (Séries Temporais)

Pybacen This library was developed for economic analysis in the Brazilian scenario (Investments, micro and macroeconomic indicators) Installation Inst

42 Jan 05, 2023
PYthon Warframe Market API(pywmapi)

pywmapi PYthon Warframe Market API(pywmapi) API for warframe market, written in Python. For now, the implemented function is listed below: auth sign i

AyajiLin 7 Dec 11, 2022