tokenizer.py

"""
Custom Chess Tokenizer for the Chess Challenge.

This tokenizer decomposes each move written in extended UCI notation
into structured sub-tokens.

The dataset format uses:
- W/B prefix for White/Black
- Piece letter: P=Pawn, N=Knight, B=Bishop, R=Rook, Q=Queen, K=King
- Source and destination squares (e.g., e2e4)
- Special suffixes: (x)=capture, (+)=check, (+*)=checkmate, (o)/(O)=castling
"""

from __future__ import annotations

import json
import os
from typing import Dict, List, Optional

from transformers import PreTrainedTokenizer


class ChessTokenizer(PreTrainedTokenizer):
    """
    A custom tokenizer for chess moves using extended UCI notation.

    Each move is decomposed into semantic sub-components rather than
    treated as a single atomic token.
    """

    model_input_names = ["input_ids", "attention_mask"]
    vocab_files_names = {"vocab_file": "vocab.json"}

    PAD_TOKEN = "[PAD]"
    BOS_TOKEN = "[BOS]"
    EOS_TOKEN = "[EOS]"
    UNK_TOKEN = "[UNK]"

    def __init__(
        self,
        vocab_file: Optional[str] = None,
        vocab: Optional[Dict[str, int]] = None,
        **kwargs,
    ):
        self._pad_token = self.PAD_TOKEN
        self._bos_token = self.BOS_TOKEN
        self._eos_token = self.EOS_TOKEN
        self._unk_token = self.UNK_TOKEN

        kwargs.pop("pad_token", None)
        kwargs.pop("bos_token", None)
        kwargs.pop("eos_token", None)
        kwargs.pop("unk_token", None)

        if vocab is not None:
            self._vocab = vocab
        elif vocab_file is not None and os.path.exists(vocab_file):
            with open(vocab_file, "r", encoding="utf-8") as f:
                self._vocab = json.load(f)
        else:
            self._vocab = self._create_default_vocab()

        self._ids_to_tokens = {v: k for k, v in self._vocab.items()}

        super().__init__(
            pad_token=self._pad_token,
            bos_token=self._bos_token,
            eos_token=self._eos_token,
            unk_token=self._unk_token,
            **kwargs,
        )

    def _create_default_vocab(self) -> Dict[str, int]:
        special_tokens = [
            self.PAD_TOKEN,
            self.BOS_TOKEN,
            self.EOS_TOKEN,
            self.UNK_TOKEN,
        ]
        return {tok: i for i, tok in enumerate(special_tokens)}

    @classmethod
    def build_vocab_from_iterator(
        cls,
        iterator,
        min_frequency: int = 1,
    ) -> "ChessTokenizer":
        from collections import Counter

        counter = Counter()
        tokenizer = cls()

        for game in iterator:
            for move in game.strip().split():
                counter.update(tokenizer._tokenize(move))

        tokens = [
            token for token, count in counter.items()
            if count >= min_frequency
        ]

        tokens = sorted(tokens)

        special_tokens = [
            cls.PAD_TOKEN,
            cls.BOS_TOKEN,
            cls.EOS_TOKEN,
            cls.UNK_TOKEN,
        ]
        vocab = {token: idx for idx, token in enumerate(special_tokens + tokens)}

        return cls(vocab=vocab)

    @classmethod
    def build_vocab_from_dataset(
        cls,
        dataset_name: str = "dlouapre/lichess_2025-01_1M",
        split: str = "train",
        column: str = "text",
        min_frequency: int = 500,
        max_samples: Optional[int] = 100000,
    ) -> "ChessTokenizer":
        from datasets import load_dataset

        dataset = load_dataset(dataset_name, split=split)

        if max_samples is not None:
            dataset = dataset.select(range(min(max_samples, len(dataset))))

        def game_iterator():
            for example in dataset:
                yield example[column]

        return cls.build_vocab_from_iterator(
            game_iterator(),
            min_frequency=min_frequency,
        )

    @property
    def vocab_size(self) -> int:
        return len(self._vocab)

    def get_vocab(self) -> Dict[str, int]:
        return dict(self._vocab)

    def _tokenize(self, text: str) -> List[str]:
        """
        Tokenize a string of moves into factorized chess tokens.
        """
        tokens: List[str] = []

        for move in text.strip().split():
            # core components
            tokens.append(move[0])        # side
            tokens.append(move[1])        # piece
            tokens.append(move[2:4])      # from-square
            tokens.append(move[4:6])      # to-square

            # suffix parsing
            suffix = move[6:]
            i = 0
            while i < len(suffix):
                if suffix.startswith("(x)", i):
                    tokens.append("x")
                    i += 3
                elif suffix[i] == "+":
                    tokens.append("+")
                    i += 1
                elif suffix[i] == "*":
                    tokens.append("#")
                    i += 1
                elif suffix[i].lower() == "o":
                    tokens.append("castle")
                    i += 1
                else:
                    i += 1

        return tokens

    def _convert_token_to_id(self, token: str) -> int:
        return self._vocab.get(token, self._vocab[self.UNK_TOKEN])

    def _convert_id_to_token(self, index: int) -> str:
        return self._ids_to_tokens.get(index, self.UNK_TOKEN)

    def convert_tokens_to_string(self, tokens: List[str]) -> str:
        special = {
            self.PAD_TOKEN,
            self.BOS_TOKEN,
            self.EOS_TOKEN,
            self.UNK_TOKEN,
        }
        return " ".join(t for t in tokens if t not in special)

    def save_vocabulary(
        self,
        save_directory: str,
        filename_prefix: Optional[str] = None,
    ) -> tuple:
        if not os.path.isdir(save_directory):
            os.makedirs(save_directory, exist_ok=True)

        vocab_file = os.path.join(
            save_directory,
            (filename_prefix + "-" if filename_prefix else "") + "vocab.json",
        )

        with open(vocab_file, "w", encoding="utf-8") as f:
            json.dump(self._vocab, f, ensure_ascii=False, indent=2)

        return (vocab_file,)


def count_vocab_from_dataset(
    dataset_name: str = "dlouapre/lichess_2025-01_1M",
    split: str = "train",
    column: str = "text",
    max_samples: Optional[int] = 10000,
) -> Dict[str, int]:
    from collections import Counter
    from datasets import load_dataset

    dataset = load_dataset(dataset_name, split=split)

    if max_samples is not None:
        dataset = dataset.select(range(min(max_samples, len(dataset))))

    tokenizer = ChessTokenizer()
    token_counts = Counter()

    for example in dataset:
        for move in example[column].strip().split():
            token_counts.update(tokenizer._tokenize(move))

    return dict(token_counts)