import numpy
import glob
import csv

from parse_cases import analysis
from compare_analysis import get_dbname


def filter_all_cases_checked_result():
    # for i in glob.glob("/home/jiajun/Downloads/cases20230802/cases_new*_st_sqlite.csv"):
    for i in glob.glob("./cases_new*_st_sqlite.csv"):
    # for i in glob.glob("./cases_new*_st_with_equal.csv"):
        yield i


def analysis_all(output_file, gen):
    m = {}

    with open(output_file, "w") as f:
        writer = csv.writer(f)
        row = ("dbname\tdifficulty\ttotal\tsql_equal_check_failed\traw_query_failed\tsql_run_failed\tgen_sql_failed\thuman_check\tsucceed\tsuccess_rate\twith_human_check_success_rate").split("\t")
        writer.writerow(row)

        for i in gen:
            dbname = get_dbname(i)

            for result in analysis(i):
                difficulty = result[0]
                current_value = m.get(difficulty, [0] * len(result[1:]))
                current_value = numpy.add(current_value, result[1:])
                m[difficulty] = current_value

                writer.writerow([dbname, *result])

            print()

        for difficulty in ("easy", "medium", "hard", "extra", "total"):
            if difficulty not in m:
                continue

            total, failed, raw_query_failed, result_failed, sql_not_found, human_check, succeed, _, _ = m[difficulty]
            writer.writerow(["total", difficulty, total, failed, raw_query_failed, result_failed, sql_not_found, human_check, succeed, succeed/total*100, (succeed + human_check)/total*100])


if __name__ == "__main__":
    output_file = "total_report.csv"
    analysis_all(output_file, filter_all_cases_checked_result())
    # dbnames = ["new_car_1", "new_baseball_1", "new_allergy_1", "new_activity_1", "new_singer"]
    # dbnames = ["new_music_1"]
    # cases_files = ["cases_{}_st_sqlite.csv".format(i) for i in dbnames]
    # analysis_all(output_file, cases_files)