Unified board & card game rules foundation for MoonBit: Go, Xiangqi, International Chess and generic poker behind one engine interface

board-game

card-game

weiqi

baduk

xiangqi

chess

poker

rules-engine

sgf

gtp

Download zip

Author

wbgxiaosu

Version

0.1.0

License

Apache-2.0

Repository

github.com/wbgxiaosu/moonqipai

Last updated

3 days ago

Downloads

#moonqipai

Unified board & card game rules foundation for MoonBit: Go (Weiqi/Baduk), Xiangqi (Chinese chess), International Chess and generic poker behind one engine interface.

This is a reusable infrastructure library, not a game. It gives tool builders — GUI clients, tournament servers, teaching software, AI harnesses, data pipelines — one consistent API for rules adjudication, legal-move enumeration, undo and result reporting across games.

#Components

Package	What it provides	Backed by
core	Unified game model: GameId, Player, Outcome, EngineError and the Engine facade with apply / legal_moves / undo / turn / is_over / outcome / render	wbgxiaosu/moonbaduk (Go), wbgxiaosu/xiangqi (Xiangqi), L76148/moonchess (chess)
cards	Original generic poker: 52-card deck, seeded deterministic shuffle, dealing, 5-card evaluation and best-of-5..7 comparison (high card → straight flush)	— (original)

Each game keeps its native text notation through the unified interface:

Go — GTP vertices (Q16, pass; the letter i is skipped)
Xiangqi — ICCS coordinates (h2e2)
Chess — UCI coordinates (e2e4, promotions as e7e8q)

#Usage

// one interface, four games
let c : @core.Engine = @core.Engine::make(@core.Chess)
let _ = c.apply("e2e4")
let moves : Array[String] = c.legal_moves()

let g : @core.Engine = @core.Engine::make(@core.Go(19, 7.5))
let _ = g.apply("Q16")
let _ = g.apply("pass")

// terminal result in game-conventional notation
let outcome : @core.Outcome? = c.outcome() // e.g. Winner(P2) → "0-1"

// poker
let deck : @cards.Deck = @cards.Deck::shuffled(42UL) // deterministic
let hand : Array[@cards.Card] = deck.deal(5)
let value : @cards.HandValue? = @cards.evaluate_five(hand)

See cmd/main for a runnable example. The example binary is documentation only — the deliverable is the library.

#Unified result strings

Outcome	Go	Xiangqi / Chess
first player wins	B+24.5 (margin)	1-0
second player wins	W+7.5 (margin)	0-1
draw	0	1/2-1/2

#Engine capabilities per game

Capability	Go	Xiangqi	Chess
legal-move enumeration	✓	✓	✓
undo	✓	✓	✓
terminal-state detection	double pass	checkmate / stalemate (a loss)	checkmate / stalemate / fifty-move rule
scoring	Chinese area rules with komi	decisive only	decisive only

Xiangqi repetition adjudication and chess threefold-repetition / insufficient-material draws are not yet wired through the unified interface; the underlying libraries expose the needed primitives (Zobrist hashing, position history) for a future version.

#Building and testing

moon check
moon test
moon run cmd/main

CI runs moon check and moon test on the wasm, wasm-gc and native backends.

#Scope statement

moonqipai is an original MoonBit codebase. The Go, Xiangqi and chess rule cores are delegated to the published mooncakes packages listed above (moonbaduk and xiangqi by the same author; moonchess by L76148, Apache-2.0) — moonqipai's contribution is the unified engine layer, the normalised outcome model and the poker module, all of which are new code written for this project. Rules semantics follow the public SGF FF[4], GTP v2, ICCS, UCI and FIDE/Xiangqi rule documents.

Site source Report issue Packages Build queue Skills Statistics