Pretokenise on punctuation except hyphens

🤗Tokenizers
1

I need to preprocess some sentences so that an existing module can split them into words based on single spaces. Currently, I have:

import tokenizers.normalizers as tn
import tokenizers.pre_tokenizers as tp

normalizer   = tn.Sequence([tn.NFD(), tn.StripAccents()])
pretokeniser = tp.Whitespace()  # Combines WhitespaceSplit and Punctuation

def preprocess(line: str):
    pretokens = pretokeniser.pre_tokenize_str(normalizer.normalize_str(line))
    return " ".join([w for w,_ in pretokens])

The problem is that my language (Dutch) has hyphenation within words for some compounds, and these are split into separate words by this process. As an example:

Energie-efficiëntie, i.e. zuinig omgaan met stroomverbruik, wordt steeds belangrijker bij het trainen van transformer-architecturen – zoveel is zeker!

now becomes

Energie - efficientie , i . e . zuinig omgaan met stroomverbruik , wordt steeds belangrijker bij het trainen van transformer - architecturen – zoveel is zeker !

Is there a way to exclude hyphens as punctuation mark?