test.py

import os
import datasets
import urllib.request
import csv

_CITATION = """\
@InProceedings{huggingface:dataset,
title = {diffusion train set},
}
"""

_DESCRIPTION = """\
This is a dataset that image data and caption txt
"""
_HOMEPAGE = ""

_LICENSE = ""

_VERSION = "0.0.1"

_URL = "data/"
   
_URLS = {        
            "train": _URL + 'train1_dataset.csv',
            "reg": _URL + 'reg1_dataset.csv',
            }
task_list = [
    "train",
    "reg",
]
class taskConfig(datasets.BuilderConfig):
    def __init__(self, **kwargs):
        super().__init__(version=datasets.Version("1.0.0"), **kwargs)

class imgdataset(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        taskConfig(
            name=task_name,
        )
        for task_name in task_list
    ]
    
    def _info(self):
        features = datasets.Features(
            {
                "folder_name" : datasets.Value("string"),
                "Class_name": datasets.Value("string"),
                "file_name": datasets.Value("string"),
                "file_id": datasets.Value("string")
        
            }
        )
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            supervised_keys=None,
            version=_VERSION,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager: datasets.DownloadManager):
        downloaded_files = dl_manager.download_and_extract(_URLS)
        task_name = self.config.name
        return [
            datasets.SplitGenerator(datasets.Split.TRAIN, {
                "filepath": downloaded_files[task_name]
            }),
            ]
    def _generate_examples(self, filepath):
        """This function returns the examples in the raw (text) form."""
        with open(filepath, encoding="utf-8") as f:
            data = csv.DictReader(f)
            key = 0
            for row in data:
                yield key, {
                    "folder_name" : row['folder_name'],
                    "Class_name": row['Class_name'],
                    "file_name": row['file_name'],
                    "file_id": row['file_id']
                }
                folder_name = row['folder_name']
                class_name =row['Class_name']
                file_name = row['file_name']
                url =row['file_id']
                path = os.path.join('./',folder_name,class_name,file_name)
                folder = os.path.join('./',folder_name,class_name)
                if not os.path.isdir(folder): #폴더가 존재하지 않는다면 폴더 생성
                    os.makedirs(folder)
                urllib.request.urlretrieve(url, path)
                key += 1