tokenizers.py

# Copyright 2024-2025 The Alibaba Wan Team Authors. All rights reserved.
import html
import string

import ftfy
import regex as re
from transformers import AutoTokenizer

__all__ = ['HuggingfaceTokenizer']


def basic_clean(text):
    text = ftfy.fix_text(text)
    text = html.unescape(html.unescape(text))
    return text.strip()


def whitespace_clean(text):
    text = re.sub(r'\s+', ' ', text)
    text = text.strip()
    return text


def canonicalize(text, keep_punctuation_exact_string=None):
    text = text.replace('_', ' ')
    if keep_punctuation_exact_string:
        text = keep_punctuation_exact_string.join(
            part.translate(str.maketrans('', '', string.punctuation))
            for part in text.split(keep_punctuation_exact_string))
    else:
        text = text.translate(str.maketrans('', '', string.punctuation))
    text = text.lower()
    text = re.sub(r'\s+', ' ', text)
    return text.strip()


class HuggingfaceTokenizer:

    def __init__(self, name, seq_len=None, clean=None, **kwargs):
        assert clean in (None, 'whitespace', 'lower', 'canonicalize')
        self.name = name
        self.seq_len = seq_len
        self.clean = clean

        # Initialize the Hugging Face tokenizer.
        self.tokenizer = AutoTokenizer.from_pretrained(name, **kwargs)
        self.vocab_size = self.tokenizer.vocab_size

    def __call__(self, sequence, **kwargs):
        return_mask = kwargs.pop('return_mask', False)

        # Build tokenizer arguments.
        _kwargs = {'return_tensors': 'pt'}
        if self.seq_len is not None:
            _kwargs.update({
                'padding': 'max_length',
                'truncation': True,
                'max_length': self.seq_len
            })
        _kwargs.update(**kwargs)

        # Normalize input and tokenize.
        if isinstance(sequence, str):
            sequence = [sequence]
        if self.clean:
            sequence = [self._clean(u) for u in sequence]
        ids = self.tokenizer(sequence, **_kwargs)

        # Return token ids, and the attention mask if requested.
        if return_mask:
            return ids.input_ids, ids.attention_mask
        else:
            return ids.input_ids

    def _clean(self, text):
        if self.clean == 'whitespace':
            text = whitespace_clean(basic_clean(text))
        elif self.clean == 'lower':
            text = whitespace_clean(basic_clean(text)).lower()
        elif self.clean == 'canonicalize':
            text = canonicalize(basic_clean(text))
        return text