import pandas as pd
import os
import shutil
import math
import sys

def split_csv(input_folder, output_folder, max_chunck_size):
    unit_path = input_folder + "/units.csv"
    soil_path = input_folder + "/soil.csv"
    crop_path = input_folder + "/crop.csv"

    os.mkdir(output_folder)
    
    dfs = pd.read_csv(soil_path)
    dfc = pd.read_csv(crop_path)

    dfs['ID'] = dfs['ID'].astype('int')
    dfc['ID'] = dfc['ID'].astype('int')

    IDs = dfs['ID'].unique()
    
    max_ID = len(IDs)

    step = max_chunck_size
    
    nb_split = math.ceil(max_ID / max_chunck_size)   

    for i in range(1, nb_split + 1):
        
        firstID = IDs[(i-1) * step]
        lastID = IDs[min((i * step) - 1 , max_ID - 1)]

        splitIDs = IDs[(i-1) * step : min((i * step) , max_ID)]
        

        dfSplit = pd.DataFrame()
        dfSplit = dfs[dfs['ID'].isin(splitIDs)]

        dfcSplit = pd.DataFrame()
        dfcSplit = dfc[dfc['ID'].isin(splitIDs)]
        
        os.mkdir(os.path.join(output_folder, f"{input_folder}_{i}"))
        shutil.copy(unit_path, os.path.join(output_folder, f"{input_folder}_{i}"))
        dfSplit.to_csv(os.path.join(output_folder, f"{input_folder}_{i}", "soil.csv"), index=False)
        dfcSplit.to_csv(os.path.join(output_folder, f"{input_folder}_{i}", "crop.csv"), index=False)
        shutil.make_archive( os.path.join(output_folder,f"{input_folder}_{i}"), 'zip',  os.path.join(output_folder, f"{input_folder}_{i}"))

# Example usage
input_folder = "input"
output_folder = "outputs"

input_file = sys.argv[1]
max_chunck_size = int(sys.argv[2])

shutil.unpack_archive(input_file, input_folder, 'zip')

split_csv(input_folder, output_folder, max_chunck_size)