#!/usr/bin/python3
from bs4 import BeautifulSoup
import configparser
from datetime import datetime
from dateutil.parser import parse
import feedparser
import hashlib
import json
import os
import requests
import re
import sys
import time
import logging
from logging.config import fileConfig

appconfig = configparser.ConfigParser()
appconfig.read('blogs-i-read_v2.ini')
blogs_to_read  = appconfig['blogsiread']['blogfile']
cronlinks_file = appconfig['blogsiread']['cronlinksfile']
feed_timeout   = float(appconfig['blogsiread']['feedtimeout'])

fileConfig('logging_config.ini')
logger = logging.getLogger("blogs-i-read_v2")
if os.environ.get('LOGLEVEL'):
    logger.setLevel(level=os.environ.get('LOGLEVEL', 'WARNING').upper())

with open(blogs_to_read, 'r') as blogfile:
    blogs = json.load(blogfile)

met_offset = 3600
md5_sums = {}
try:
    with open('md5_sums.json', 'r') as file:
        md5_sums = json.load(file)
except:
    logger.debug('could not open md5_sums.json')

# Dictionary to store the results
results = {}
        
def get_feed_content(url):
    count = 0
    feed = ''
    while count <= 3:
        count += 1
        #if True:
        try:
            if count > 1:
                logger.debug(f"attempt {count} to read from {url}")
            response = requests.get(url,
                headers={'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1)'},
                                    timeout=feed_timeout )
            response.encoding = 'utf-8'
            feed = feedparser.parse(response.text)
            break
        except:
            if count == 3:
                break
    return feed

def examine_feed(url):
    (md5, post_title, post_url, last_update) = get_default_values(url)
    feed = get_feed_content(url)
    try:
        post_title = feed.entries[0].title
        post_url = feed.entries[0].link
        if 'theonlinephotographer' in post_url:
            try:
                post_url = feed.entries[0].feedburner_origlink
            except:
                pass
        logger.debug(post_url)
        old_md5 = hashlib.md5( post_title.encode('utf-8')
                               + feed.entries[0].updated.encode('utf-8') ).hexdigest()
        md5 = 'v2_' + hashlib.md5( post_title.encode('utf-8')
                                   + post_url.encode('utf-8') ).hexdigest()
        # make it dependant on change
        if url in md5_sums:
            logger.debug('existent feed')
            if md5_sums[url]['md5'] not in [ md5, old_md5 ]:
                logger.debug(f'hashes NOT equal')
                last_update = int(time.mktime(datetime.utcnow().timetuple())) + met_offset
            else:
                logger.debug('newhash equal to old or new saved hashes')
                last_update = md5_sums[url]['timestamp']
        else:
            logger.debug('new feed')
    except:
        logger.info(f'error when parsing feed {url}')
        try:
        #if True:
            post_title = md5_sums[url]['current_title']
            post_url = md5_sums[url]['post_url']
            md5 = md5_sums[url]['md5']
            last_update = md5_sums[url]['timestamp']
        except:
            pass
    logger.debug(f"last_update: {last_update}")
    return md5, post_title, post_url, last_update

def examine_photoplacegallery(soup, url, md5):
    (post_title, post_url, last_update) = ['', '', 0]
    # logger.debug('examine_photoplacegallery')
    prothost    = re.search(r'^http[s]*:\/\/[\w\.]*', url).group()
    firstah3    = soup.find_all('a','h3')[0]
    post_title  = firstah3.string
    post_url    = prothost + firstah3.get('href')
    if url in md5_sums:
        logger.debug(f'found {url} in md5_sums')
        if md5_sums[url]['md5'] != md5:
            logger.debug('md5 not equal')
            md5_sums[url]['md5'] = md5
            last_update = time.time()
        else:
            logger.debug('md5 equal')
            md5 = md5_sums[url]['md5']
            last_update = md5_sums[url]['timestamp']
    else:
        last_update = time.time()
    return md5, post_title, post_url, last_update

def examine_lfionline(soup, url, md5):
    (post_title, post_url, last_update) = ['', '', time.time()]
    logger.debug('examine_lfionline')
    all_cards = soup.find_all(name="div", class_="card")
    for card in all_cards:
        if not card.find_all('img', src=lambda x: x.endswith('.svg')):
            post_url    = card.find('a')['href']
            post_title  = card.find(name="h3").text
            break
    if url in md5_sums:
        logger.debug(f'found {url} in md5_sums')
        if md5_sums[url]['md5'] != md5:
            logger.debug('md5 not equal')
            md5_sums[url]['md5'] = md5
        else:
            logger.debug('md5 equal')
            md5 = md5_sums[url]['md5']
            last_update = md5_sums[url]['timestamp']
    logger.debug(f"{post_title} {post_url} {last_update}")
    return md5, post_title, post_url, last_update

def examine_generic_website(soup, url, md5):
    (post_title, post_url, last_update) = ['', '', 0]
    prothost    = re.search(r'^http[s]*:\/\/[\w\.]*', url).group()
    logger.debug(url)
    if url in md5_sums:
        # logger.debug(f"found {url} in md5_sums: \n\t{md5_sums[url]['md5']} vs \n\t{md5}")
        if md5_sums[url]['md5'] != md5:
            logger.debug('md5 not equal')
            md5_sums[url]['md5'] = md5
            last_update = int(time.mktime(datetime.utcnow().timetuple())) + met_offset
        else:
            logger.debug('md5 equal')
            # logger.debug(md5_sums[url]['timestamp'])
            if md5_sums[url]['timestamp'] > 0:
                last_update = md5_sums[url]['timestamp']
            else:
                last_update = int(time.mktime(datetime.utcnow().timetuple())) + met_offset
    else:
        last_update = int(time.mktime(datetime.utcnow().timetuple())) + met_offset
    logger.debug(last_update)
    return md5, post_title, post_url, last_update

def get_default_values(url):
    # initialize variables, suitable for new urls
    (md5, post_title, post_url, last_update) = \
        ['', '', '', int(time.mktime(datetime.utcnow().timetuple())) + met_offset]
    if url in md5_sums:
        # get stored values if they exist
        try:
            md5 = md5_sums[url]['md5']
            last_update = md5_sums[url]['timestamp']
            post_title = md5_sums[url]['current_title']
            post_url = md5_sums[url]['post_url']
        except:
            pass
    logger.debug(f"last_update: {last_update}")
    return(md5, post_title, post_url, last_update)

def examine_url(url):
    (md5, post_title, post_url, last_update) = get_default_values(url)
    logger.debug(url)
    try:
        loaded_cookies = md5_sums[url]['cookies']
    except:
        loaded_cookies = {}
    #if True:
    try:
        response = requests.get(url, cookies=loaded_cookies, timeout=feed_timeout)
        saved_cookies = requests.utils.dict_from_cookiejar(response.cookies)
        cookies_json = json.dumps(saved_cookies, indent=4)
        md5_sums[url]['cookies'] = saved_cookies
        soup = BeautifulSoup(response.text, 'html.parser')
        all_text = "".join(soup.body.get_text())
        md5 = hashlib.sha256(all_text.encode('utf-8')).hexdigest()
        body = soup.find('body')
        if 'lfi-online.de' in url:
            (md5, post_title, post_url, last_update) = examine_lfionline(soup, url, md5)
        elif "photoplacegallery.com" in url:
            (md5, post_title, post_url, last_update) = examine_photoplacegallery(soup, url, md5)
        else:
            (md5, post_title, post_url, last_update) = examine_generic_website(soup, url, md5)
    except:
        logger.warning(f'Error in {url}')
    return md5, post_title, post_url, last_update

def needs_update(url):
    if len(sys.argv) > 1:
        return True
    if url not in md5_sums:
        return True
    last_update = md5_sums[url]['timestamp']
    epoch = time.mktime(datetime.utcnow().timetuple()) + met_offset 
    logger.debug(f"{last_update} - {epoch} : {((epoch - last_update)/3600):.1f} hours old")
    minute = datetime.utcfromtimestamp(epoch).minute
    quarter = 0
    if 15 <= minute < 30:
        quarter = 1
    elif 30 <= minute < 45:
        quarter = 2
    else:
        quarter = 3
        
    diff = epoch - last_update
    if diff > 3600*24*7:
        if quarter == 1:
            return True
        else:
            return False
    elif diff > 3600*18:
        return True
    elif diff > 3600*12:
        if quarter % 2 == 1:
            return True
        else:
            return False
    elif diff > 3600*6:
        if quarter == 1:
            return True
        else:
            return False
    else:
        return False

# Function to get the title, MD5 hash of the HTML content, and the time of the last change for a given URL
def get_url_info( blog ):
    if 'feed' in blog.keys():
        url = blog['feed']
    else:
        url = blog['url']
    if needs_update(url):
        logger.debug(f"{url} needs update")
        if 'feed' in blog.keys():
            (md5, post_title, post_url, last_update) = examine_feed(blog['feed'])
        else:
            (md5, post_title, post_url, last_update) = examine_url(blog['url']) 
    else:
        logger.debug(f"{url} needs NO update")
        md5 = md5_sums[url]['md5']
        post_title = md5_sums[url]['current_title']
        post_url = md5_sums[url]['post_url']
        last_update = md5_sums[url]['timestamp']

    if url not in md5_sums.keys():
        md5_sums[url] = {}
    md5_sums[url]['post_url'] = post_url
    md5_sums[url]['current_title'] = post_title
    md5_sums[url]['md5'] = md5
    md5_sums[url]['timestamp'] = last_update
    
    results[url] = { 'blog_url': blog['url'],
                     'blog_title': blog['title'],
                     'current_title': post_title,
                     'post_url': post_url,
                     'md5': md5,
                     'last_update': last_update }
    
# ------------------------------------------------------------- main ---
filter = False
if len(sys.argv) > 1:
    filter = sys.argv[1]
# Loop through the list of URLs and call the function for each URL
for b in blogs:
    if filter:
        if filter in b['url']:
            get_url_info(b)
    else:
        get_url_info(b)

# # save results for development
# with open('results.json', 'w') as file:
#     json.dump(results, file, indent=4)

sorted_data = dict(sorted(results.items(), key=lambda x: x[1]["last_update"], reverse=True))

utc_time = datetime.utcnow()
epoch_time = int(time.mktime(utc_time.timetuple())) + met_offset 

time_separator_flag = 0

with open(cronlinks_file, "w") as cronlinks:
    cronlinks.write("\n<li style='font-weight: bold;'>Hot from the Blogosphere</li>\n\t<ul>\n")
    for r in sorted_data:
        if not sorted_data[r]['current_title']:
            sorted_data[r]['current_title'] = ''
        
        lupd = sorted_data[r]['last_update']
        if epoch_time - lupd > 10*3188967: 
            if time_separator_flag < 4:
                cronlinks.write("</ul>\n<li style='font-weight: bold;'>From medieval ages</li>\n\t<ul>\n")
            time_separator_flag = 4
        elif epoch_time - lupd > 3188967: 
            if time_separator_flag < 3:
                cronlinks.write("</ul>\n<li style='font-weight: bold;'>A month and older</li>\n\t<ul>\n")
            time_separator_flag = 3
        elif epoch_time - int(lupd) > 815000:
            if time_separator_flag < 2:
                cronlinks.write("</ul>\n<li style='font-weight: bold;'>A week and older</li>\n\t<ul>\n")
            time_separator_flag = 2
        elif epoch_time - lupd > 150000:
            if time_separator_flag < 1:
                cronlinks.write("</ul>\n<li style='font-weight: bold;'>A day and older</li>\n\t<ul>\n")
            time_separator_flag = 1
        sdr = sorted_data[r]
        cronlinks.write(f"\t<li><a href='{sdr['blog_url']}' target='_blank' title='{datetime.fromtimestamp(sdr['last_update'])}'>{sdr['blog_title']}</a>" +
                        "&nbsp;//&nbsp;" +
                        f"<a href='{sdr['post_url']}' target='_blank'>{sdr['current_title']}</a></li>\n")
        # \t<!-- {datetime.datetime.fromtimestamp(lupd)} // {epoch_time} - {lupd} = {epoch_time - lupd} :: {time_separator_flag} -->\n"

# save hashes and timestamps
with open('md5_sums.json', 'w') as file:
    json.dump(md5_sums, file, indent=4)

# local variables:
# compile-command: "python3 blogs-i-read_v2.py"
# end: