import re
import json

def parse_header(header):
    regex = re.compile(r"(\w+)(?:{(?:(\d+),)?(\d+)}(?:::(\w+))?)?")
    capture_list = regex.findall(header.strip())

    header_order = []
    lists = {}
    funcs = {}

    for id, min, max, func in capture_list:
        header_order.append(id)
        if max != "":
            lists[id] = (int(min) if min else max, int(max))
        if func != "":
            funcs[id] = func
    return header_order, lists, funcs


def parse_data(lines, header_order, lists, funcs):
    data_regex = ""
    for id in header_order:
        if id in lists:
            min, max = lists[id]
            if min == max:
                num = '{' + str(max) + '}'
            else:
                num = '{' + str(min) + ',' + str(max) + '}'
            data_regex +=  rf"(?P<{id}>([^,]+,?){num}),?"
        else:
            data_regex += rf"(?P<{id}>[^,]+),?"
    data_regex = re.compile(data_regex)
    
    data = []
    for line in lines:
        for match in data_regex.finditer(line.strip()):
            data.append(match.groupdict())
    
    for elem in data:
        for id in header_order:
            if id in lists:
                elem[id] = [int(num) for num in re.findall(r"\d+",elem[id])]
            if id in funcs:
                if funcs[id] == "sum":
                    elem[id] = sum(elem[id])
                elif funcs[id] == "media":
                    elem[id] = sum(elem[id]) / len(elem[id])
    
    return data


def main():
    for i in range(2,6):
        test_file = f"alunos{i}"
        f = open(f"{test_file}.csv", "r")
        lines = f.readlines()
        header_order, lists, funcs = parse_header(lines[0])

        data = parse_data(lines[1:], header_order, lists, funcs)
        json_file = open(f"{test_file}.json","w")
        json.dump(data,json_file,indent=4,ensure_ascii=False)

if __name__ == "__main__":
    main()