import os
import datasets

script_path = os.path.abspath(__file__)
src_path = os.path.dirname(script_path)
root_path = os.path.dirname(src_path)
res_path = root_path+"/res"

texts = []
for name in os.listdir(res_path):
    if name.endswith(".txt"):
        with open(res_path+"/"+name, "r") as file:
            text = file.read()
            texts.append(text)

dataset = datasets.Dataset.from_dict({"text": texts})
dataset = dataset.train_test_split(train_size=None, test_size=4, shuffle=False)

train_val_dataset = dataset["train"]
train_val_dataset = train_val_dataset.train_test_split(train_size=None, test_size=2, shuffle=False)

train_dataset = train_val_dataset["train"]
val_dataset = train_val_dataset["test"]
test_dataset = dataset["test"]
dataset = datasets.DatasetDict({
    "train": train_dataset,
    "val": val_dataset,
    "test": test_dataset
})

dataset.push_to_hub("shakespearefirstfolio")