lpsim 0.0.post1

Lochfolk Prinzessin Simulator, which simulates Genius Invokation TCG

Genius Invokation TCG Simulator

---

Backend of Lochfolk Prinzessin Simulator, which simulates Genius Invokation TCG, written with Python 3.10 using Pydantic and FastAPI.

This project is created for leisure, no guarantee for progress and quality.

This project is under AGPL-3.0 license.

Progress: Charactors 37/54, Cards 133/199

Usage

This project works with Python 3.10, and requires packages listed in requirements.txt. To install the packages, use pip install -r requirements.txt.

A FastAPI server is provided, which can be used to interact with the match. Its requirements are listed in network/requirements.txt.

TODO: package-like usage is suppoted, need to update README

Currently package-like usage is not supported, please follow the following instructions to run the project. This project support a frontend zyr17/GITCG-frontend. To use the frontend, please refer FastAPI section and readme of frontend.

Define decks for both players. Supports text-based or json-based deck definition. Refer to server/deck.py for details.

The outlines:

1. Initialize a fresh server.Match instance.
2. Use the set_deck function to assign decks for players.
3. Modify the Match.config if specific configurations are necessary.
4. Once the decks are set, initiate the match using the Match.start function. This initializes the match according to the configurations and decks.
5. Progress through the match by employing the Match.step function. By default, the step function will continually execute until the match either ends or generates requests requiring responses.

In order to manage responses, the project includes a selection of simple agents within the agents module. These agents can generate responses to various requests. Notably, the InteractionAgent is equipped to interpret command lines and formulate responses. This proves useful both in console environments and within frontend interfaces.

from lpsim import Match, Deck
from lpsim.agents import RandomAgent
deck_string = '''
charactor:Fischl
charactor:Mona
charactor:Nahida
Gambler's Earrings*2
Wine-Stained Tricorne*2
Vanarana
Timmie*2
Rana*2
Covenant of Rock
Wind and Freedom
The Bestest Travel Companion!*2
Changing Shifts*2
Toss-Up
Strategize*2
I Haven't Lost Yet!*2
Leave It to Me!
Clax's Arts*2
Adeptus' Temptation*2
Lotus Flower Crisp*2
Mondstadt Hash Brown*2
Tandoori Roast Chicken
'''
deck0 = Deck.from_str(deck_string)
deck1 = Deck.from_str(deck_string)
match = Match()
match.set_deck([deck0, deck1])
match.start()
match.step()

agent_0 = RandomAgent(player_idx = 0)
agent_1 = RandomAgent(player_idx = 1)

while not match.is_game_end():
    if match.need_respond(0):
        match.respond(agent_0.generate_response(match))
    elif match.need_respond(1):
        match.respond(agent_1.generate_response(match))
    match.step()

print(f'winner is {match.winner}')

FastAPI

Use FastAPI to provide a web server, which can be used to interact with the match. To run a server, use the following command:

uvicorn network:app --reload

It will open a FastAPI server on localhost:8000, and accepts connections on localhost:4000, which is the port of frontend. Currently exception management is chaos, errors may set game state to ERROR, raise Exception on server side, make empty response on agent, return 404, or run JS failed on frontend. Please open console of frontend and check the error message on both sides.

Feature

Pydantic to save & load states of match, exported data is complete to restore from certain state and continue running, and also easy for frontend to render the game states.

High compatible with different version of charactors or cards. (Although currently not implemented,) You can start a match between version 3.8 Itto-Barbara-Noelle and version 3.3 dual-Geo Maguu Kenki.

Interact by request and response. When request list is not empty, agents need to response to one of the request. When multiple players need to response, (e.g. switch card and choose charactor at startup), their requests will be generated simultaneously.

All modifications to the match table are orchestrated through actions. Each action triggers an event and has the potential to activate subsequent actions. These newly activated actions are appended to the top of the existing action list, akin to a stack structure.

Objects integrated into this system introduce two types of triggers: event handlers and value modifiers. Event handlers function to monitor events and produce lists of actions in response. Value modifiers is used on edit mutable values and updating internal states if is necessary. Modifiable values applies to attributes like initial dice color, damage, and cost.

Progress

Charactors Cards

Contribution

Contributions are welcomed, however, currently there is no detailed guide for contribution. To add new charactor related objects (Charactor, Skill, Talent, Summon, Status), please refer to server/charactor/template.py and implemented server/charactor/electro/fischl.py server/charactor/hydro/mona.py and server/charactor/dendro/nahida.py. To add a new card, please refer to existing card implementations in server/card.