#!/usr/bin/env python3
import datetime
import os
from urllib.parse import urlparse, urlunparse, urljoin

import bs4
from urllib.request import Request, urlopen
from os.path import exists
from shutil import move
from hashlib import md5
import configparser


def get_page_hash(text: str):
    text = text.replace(' ', '').replace('\r', '').replace('\n', '')

    return md5(text.encode('utf-8')).hexdigest()


def spider(target, exclude):
    parsed_target = urlparse(target)
    return spider_rec(dict(), dict(), target, parsed_target, exclude)


def spider_rec(links, checksums, current_href, base_parse, exclude):
    target_url = urlunparse(base_parse)
    parse_result = urlparse(urljoin(target_url, current_href))
    req = Request(urlunparse(parse_result))

    postfix = parse_result.path
    if parse_result.query:
        postfix += "?" + parse_result.query

    if len(postfix) == 0:
        postfix = "/"

    if parse_result.hostname == base_parse.hostname:
        html_page = urlopen(req)

        soup = bs4.BeautifulSoup(html_page, "lxml")

        checksums[postfix] = get_page_hash(soup.getText())
        links[postfix] = 1

        for link in soup.findAll('a'):
            href = link.get('href')
            href = href.replace(" ", "%20")

            if "mailto:" not in href:
                if not urlparse(href).hostname:
                    href_parse = urlparse(urljoin(target_url, href))
                    href = href_parse.path

                    if href_parse.query:
                        href += "?" + href_parse.query

                if href not in links.keys():
                    found = False
                    for d in exclude:
                        if d in href:
                            found = True
                            break

                    if found:
                        continue

                    spider_rec(links, checksums, href, base_parse, exclude)
                else:
                    links[href] += 1
    return links, checksums


def cmp(p1, p2):
    with open(p1, 'r') as f1:
        with open(p2, 'r') as f2:
            l1 = f1.readlines()
            l2 = f2.readlines()
            if len(l1) == len(l2):
                for i in range(len(l1)):
                    if l1[i] != l2[i]:
                        return False
            else:
                return False

    return True


def main():
    print("Reading conf...")

    config = configparser.ConfigParser()
    config.read('crawl.conf')
    config = config['Config']

    target = config['site']
    path = config['target']
    checksums_path = config['checksums']

    ignores = config['ignore'].split(', ')

    checksums = dict()
    try:
        with open(checksums_path, 'r') as checksums_file:
            for line in checksums_file.readlines():
                thirds = line.split()
                checksums[thirds[0]] = thirds[1:]
    except FileNotFoundError:
        print("No checksums file found at path, new file will be created.")

    print("Crawling site...")
    links, new_checksums = spider(target, ignores)
    date = datetime.datetime.utcnow()

    existed = exists(path)
    oldpath = path
    if existed:
        print("Sitemap already exists, creating temp...")
        path = "newmap.xml"

    print("Writing to target file...")
    out = open(path, 'w')
    out.write("<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n")
    out.write("<!--\n")
    out.write("\tSitemap generator by Ben Morgan - www.benrmorgan.com\n")
    out.write("-->\n")
    out.write("<urlset xmlns=\"http://www.sitemaps.org/schemas/sitemap/0.9\">\n")

    sorted_links = dict(sorted(links.items(), key=lambda item: item[1], reverse=True))

    ordered = []
    level = 0
    old_num = sorted_links[list(sorted_links.keys())[0]]
    for l in sorted_links.keys():
        if sorted_links[l] != old_num:
            level += 1
            old_num = sorted_links[l]

        ordered.append((l, str(float(str(round(pow(0.8, level), 2))))))

    checksums_out = open(checksums_path, 'w')

    different_count = 0
    for l in ordered:
        lastmod = date.strftime("%Y-%m-%dT%H:%M:%S+00:00")
        if l[0] in checksums.keys() and checksums[l[0]][0] == new_checksums[l[0]]:
            lastmod = checksums[l[0]][1]
            different_count += 1
        
        checksums_out.write(f"{l[0]} {new_checksums[l[0]]} {lastmod}\n")
        
        if l[0] == '/':
            l = target + l[0]

        out.write("\t<url>\n")
        out.write("\t\t<loc>" + l[0] + "</loc>\n")

        out.write("\t\t<lastmod>" + lastmod + "</lastmod>\n")
        out.write("\t\t<priority>" + str(l[1]) + "</priority>\n")
        out.write("\t</url>\n")

    out.write("</urlset>\n")
    out.close()

    checksums_out.close()

    if existed and not cmp(oldpath, path):
        print("Creating old sitemap backup...")
        move(oldpath, oldpath + "-old")
        print("Overwriting old sitemap with new one...")
        move(path, oldpath)
    elif existed:
        print("Sitemaps are the same, removing temp...")
        os.remove(path)

    print("Done.")
    print(f"Crawled {len(links.keys())} pages.")
    print(f"Found {different_count} modified pages.")


main()