#!/usr/bin/env python3
# encoding: utf-8

##
# This script checks all media files in the repository for whether they've been modified or added without updating the file tracking image, sound, and music copyright
##

import argparse
import contextlib
import csv
import hashlib
from operator import itemgetter
import os
from pathlib import Path
from subprocess import check_output
import sys

##
# csv file layout:
# [0] = current git commit date
# [1] = file path, relative to the repository root
# [2] = license name(s)
# [3] = authorship information
# [4] = notes
# [5] = new git commit date, if different from the value in [0]
# [6] = current md5 hash
##

def do_git(file):
    return str(check_output(["git", "log", "-1", "--format=%ad", "--date=format:%Y/%m/%d", file]), 'UTF-8').rstrip('\n')

def do_hash(file):
    md5 = hashlib.md5()
    with open(file, 'rb') as f:
        while True:
            data = f.read(65536)
            if not data:
                break
            md5.update(data)
    return str(md5.hexdigest())

##
# program logic start
##

args = argparse.ArgumentParser()
args.add_argument("--repo", default=".", help="The directory of the Wesnoth repository to run this script against.")
args.add_argument("--output", default="output.csv", help="The file to write the results of this script to.")
args.add_argument("--input", default="copyrights.csv", help="The file to read the existing copyright data from.")
options = args.parse_args()
os.chdir(options.repo)

with contextlib.suppress(FileNotFoundError):
    os.remove(options.output)

csv_data = {}
added = []
changed = []
unchanged = []
removed = []
# Sanity-check that the input is in the same order as the output would be
csvfile_needs_sorting = False

with open(options.input) as csvfile:
    reader = csv.reader(csvfile)
    previous_file = ""
    for row in reader:
        if row[0] == "Date":
            continue

        file = row[1]
        if file < previous_file:
            csvfile_needs_sorting = True
        previous_file = file

        if not os.path.exists(file):
            removed.append(file)
            continue

        csv_data[file] = row

for root, _, files in os.walk(options.repo):
    for filename in files:
        filetype = Path(filename).suffix
        if filetype == ".png" or filetype == ".jpg" or filetype == ".webp" or filetype == ".wav" or filetype == ".ogg":
            file = os.path.normpath(os.path.join(root, filename))
            hash = do_hash(file)

            if not file in csv_data:
                added.append(["", file, "", "", "", do_git(file), hash])
            elif csv_data[file][6] != hash:
                csv_data[file][5] = do_git(file)
                csv_data[file][6] = hash
                changed.append(csv_data[file])
            else:
                unchanged.append(csv_data[file])

added.sort(key=itemgetter(1))
changed.sort(key=itemgetter(1))
unchanged.sort(key=itemgetter(1))

final_output = added + changed + unchanged

if options.output != "":
    with open(options.output, 'w') as f:
        f.write("Date,File,License,Author - Real Name(other name);Real Name(other name);etc,Notes,Needs Update,MD5\n")
        for row in final_output:
            f.write(",".join(row)+"\n")
else:
    for row in final_output:
        print(",".join(row))

any_check_failed = False
if len(removed) > 0:
    any_check_failed = True
    print("There are "+str(len(removed))+" removed images")

if len(added) > 0 or len(changed) > 0:
    any_check_failed = True
    print("There are "+str(len(added))+" new images")
    print("There are "+str(len(changed))+" changed images")

if csvfile_needs_sorting:
    any_check_failed = True
    print("The input file isn’t sorted by filename")
    print("    Changed or newly added lines are put at the top for easy editing,\n    but you should run the tool again after editing to sort the file.")

if any_check_failed:
    sys.exit(1)