#!/usr/bin/env python3

import argparse
from pathlib import Path
from pprint import pprint
import pandas as pd

import utils

argparser = argparse.ArgumentParser(description="OCR-Logfile evaluation")
argparser.add_argument("vp_dir", help="Directory with all VPs")

args = argparser.parse_args()

data_path = Path(args.vp_dir)
all_vp = [x for x in data_path.iterdir() if x.is_dir()]

vp_results = []
for vp_path in all_vp:
    video_path = next(vp_path.glob("*.mkv"))
    ocr_path = vp_path / "analysis_results.csv"
    log_path = vp_path / f"{vp_path.stem}.csv"

    df = utils.combine_ocr_logs(video_path, ocr_path, log_path)
    df = df.fillna('')
    df["vp_code"] = vp_path.stem

    df = utils.calc_levenshtein_distance(df)

    url_groups = utils.group_urls(list(df["url"].values))
    pprint(len(url_groups))

    df.to_csv(f"{vp_path}/metrics.csv")
    utils.write_grouped_metrics(df, url_groups, vp_path)

    df = pd.read_csv(f"{vp_path}/metrics_grps.csv")

    vp_results.append(df)

evals = utils.evaluate_results(vp_results)
pprint(evals)
evals.to_csv(f"{data_path}/evaluation.csv")