import os
import json


# Save data
def save_data(dataset, filetype, filename):
    if dataset:
        directory = "./ejde_buffer/" + filetype + "/"
        os.makedirs(directory, exist_ok=True)
        filepath = os.path.join(directory, filename)
        with open(filepath, "w", encoding='utf-8') as json_file:
            json.dump(dataset, json_file, indent=4)
        print(filetype + " data have been added to", filepath)


# Write into output files
def transform_data():
    def read(folder_path, output_files):
        # Create new folders
        os.makedirs('./ejde_buffer/Article_output/', exist_ok=True)
        os.makedirs('./ejde_buffer/Author_output/', exist_ok=True)

        data_oldest = []
        data_2010_2014 = []
        data_2015_2020 = []
        data_newest = []
        data_no_date = []
        data_integrate = []

        for filename in os.listdir(folder_path):
            if filename.endswith('.json'):
                file_path = os.path.join(folder_path, filename)
                with open(file_path, 'r', encoding='utf-8') as file:
                    data = json.load(file)
                    for Dict in data:
                        if Dict.get('volume') is not None or Dict.get('affiliation', [{}])[0].get('year', 0) != '':
                            # Select data
                            if int(Dict.get('volume') or Dict.get('affiliation', [{}])[0].get('year', 0)) <= 2009:
                                data_oldest.append(Dict)
                            elif int(Dict.get('volume') or Dict.get('affiliation', [{}])[0].get('year', 0)) <= 2014:
                                data_2010_2014.append(Dict)
                            elif int(Dict.get('volume') or Dict.get('affiliation', [{}])[0].get('year', 0)) <= 2020:
                                data_2015_2020.append(Dict)
                            else:
                                data_newest.append(Dict)
                        else:
                            data_no_date.append(Dict)

        data_integrate.append(data_oldest)
        data_integrate.append(data_2010_2014)
        data_integrate.append(data_2015_2020)
        data_integrate.append(data_newest)
        data_integrate.append(data_no_date)

        # Transfer
        Data = [data_oldest, data_2010_2014, data_2015_2020, data_newest, data_no_date, data_integrate]

        for index in range(0, 6):
            with open(output_files[index], 'w', encoding='utf-8') as file:
                json.dump(Data[index], file, indent=4)

    # The path of reading
    author_folder_path = './ejde_buffer/Author_TS'
    article_folder_path = './ejde_buffer/Article_TS'

    # The path of storage
    author_output_file = [
        './ejde_buffer/Author_output/EJDE_Author_output_file(oldest).json',
        './ejde_buffer/Author_output/EJDE_Author_output_file(2010-2014).json',
        './ejde_buffer/Author_output/EJDE_Author_output_file(2015-2020).json',
        './ejde_buffer/Author_output/EJDE_Author_output_file(newest).json',
        './ejde_buffer/Author_output/EJDE_Author_output_file(no date).json',
        './ejde_buffer/Author_output/EJDE_Author_output_file(integration).json'
    ]

    article_output_file = [
        './ejde_buffer/Article_output/EJDE_Article_output_file(oldest).json',
        './ejde_buffer/Article_output/EJDE_Article_output_file(2010-2014).json',
        './ejde_buffer/Article_output/EJDE_Article_output_file(2015-2020).json',
        './ejde_buffer/Article_output/EJDE_Article_output_file(newest).json',
        './ejde_buffer/Article_output/EJDE_Article_output_file(no date).json',
        './ejde_buffer/Article_output/EJDE_Article_output_file(integration).json'
    ]

    # Read and write into files
    read(author_folder_path, author_output_file)
    read(article_folder_path, article_output_file)

    # End
    print("\nData has been written into files.")


# Delete files in temporary storage area
def delete_data():
    folder_paths = ['./ejde_buffer/Author_TS', './ejde_buffer/Article_TS']
    for folder_path in folder_paths:
        file_names = os.listdir(folder_path)
        for file_name in file_names:
            file_path = os.path.join(folder_path, file_name)
            if os.path.isfile(file_path):
                os.remove(file_path)
        os.rmdir(folder_path)
    print('\nAttention: The temporary storage files have been deleted!')