# -*- coding: utf-8 -*-

# parser.py
# Version : 0.

# This file is part of Qarte+7
#    
# Author : Vincent Vande Vyvre <vincent.vandevyvre@swing.be>
# Copyright: 2009-2010 Adrien Beudin beudbeud@gmail.com
# Copyright: 2011 Vincent Vande Vyvre
# Licence: GPL3
# Home page : https://code.launchpad.net/~vincent-vandevyvre/+junk/Qarte+7
#

import os, re
import time
import urllib2, xml.dom.minidom
import sys
import BeautifulSoup as BS

def unescape_html(text):
    return BS.BeautifulStoneSoup(text, convertEntities=BS.BeautifulStoneSoup.HTML_ENTITIES).contents[0]

time_re = re.compile("^\d\d[h:]\d\d$")
monthes = [u"ja", u"f", u"mar", u"av", u"mai", u"juin", u"juil", 
               u"ao", u"se", u"oc", u"no", u"d", 
               u"Ja", u"F", u"Mä", u"Ap", u"Mai", u"Jun", u"Jul", 
               u"Au", u"Se", u"Ok", u"No", u"De"]

def parse_date( date_str ):
    date_array = date_str.split(",")
    if time_re.search(date_array[-1].strip()) is None:
        return ""
    time_ = date_array[-1].strip()
    if date_array[0].strip() in (u"Aujourd'hui", u"Heute"):
        date_ = time.strftime("%Y %m %d")
    elif date_array[0].strip() in (u"Hier", u"Gestern"):
        date_ = time.strftime("%Y %m %d", time.localtime(time.time() - (24*60*60)))
    else:
        array = date_array[1].split()
        day = array[0].strip(".")
        month = array[1]
        for idx, arr in enumerate(monthes):
            if month.startswith(arr):
                if idx > 11:
                    idx -= 12
                month = "%02d" % (idx+1)
                break
        year = array[2]
        date_ = "%s %s %s" % (year, month, day)
    return date_ + ", " + time_

class Parser(object):

    # Constantes
    ARTE_WEB_ROOT = 'http://videos.arte.tv'
    INDEX_TAG = 'index'
    TITLE_TAG = 'bigTitle'
    DATE_TAG = 'startDate'
    URL_TAG = 'targetURL'
    MMS_TAG = 'mmsURL'
    RESUME_TAG = 'resume'
    IMAGE_TAG = 'previewPictureURL'

    # Contenu du catalogue
    videos = []
    # Expressions régulières utilisée pour la recherche dans les pages web
    xmlRE = re.compile('xmlURL", "(.*\.xml)"')
    wmvRE = re.compile('availableFormats.*=.*"(.*HQ.*wmv.*)"')
    mmsRE = re.compile('"(mms.*)"')
    resumeRE = re.compile('<p class="text">([^<]*)<')

    def __init__(self, l):
        lang = "/%s/" % l
        max_video_displayed = 200
        self.error = False
        # Stage 1
        base_page_url = self.ARTE_WEB_ROOT + lang + "videos/" 
        #we first load the page in order to get the page url 
        #with the correct index
        html_content = urllib2.urlopen(base_page_url).read()
        count, soup = 0, False
        while 1:
            try:
                soup = BS.BeautifulSoup(html_content)
            except Exception, why:
                self.error = "Error in stage 1 : BeautifulSoup can't read the "\
                                "page\n %s" % why
                # Bad formatted line, remove it
                count += 1
                print "%s error(s) : %s" % (count, why)
                if count == 30:
                    # 30 bad formatted lines!, Abort
                    soup = False
                    break
                if "malformed" in str(why):
                    l = str(why).split(",")
                    nl = l[-2].split(" ")[-1]
                    htl = html_content.split("\n")
                    html_content = html_content.replace(htl[int(nl) - 1], "")
                else:
                    soup = False
                    break
            else:
                break
           
        if soup:
            # Stage 2
            #print "Stage 2"
            self.error = False
            try:
                found_url = 0
                for j in soup.findAll('script'): 
                    #we will look for the script in the page that has the url 
                    #with the correct index
                    for text in j:
                        if "videowallSettings" in text: 
                            #when the script is found, we will collect the url
                            for word in text.split():
                                if "asThumbnail" in word: 
                                    #there are 4 different urls, we want the one 
                                    #that displays thumbnails
                                    base_page_url = self.ARTE_WEB_ROOT + \
                                            word.replace('"','')  + "?hash=" + \
                                            lang.replace('/','') + "/thumb///1/"\
                                            + str(max_video_displayed) + "/"
                                    found_url = 1
                                    break
                        if found_url:
                            break
                    if found_url:
                        break
                
            except Exception, why:
                self.error = "Error stage 2, URL not found : %s" % why
            else:
                # Stage 3
                try:                               
                    html_content = urllib2.urlopen(base_page_url).read() 
                    soup = BS.BeautifulSoup(html_content)
                    for i in soup.findAll('div', {"class":"video"}):
                        try:
                            video = dict()
                            for h in i.findAll('h2'):
                                for a in h.findAll('a'):
                                    video['targetURL'] = self.ARTE_WEB_ROOT + a['href']
                                    video['targetURL'] = video['targetURL'].replace("/fr/", lang)
                                    video['bigTitle'] = unescape_html( a.string)
                            for p in i.findAll('p'):
                                if 'class' in p:
                                    if p['class'] == 'teaserText':
                                        video['summary'] = p.string
                                else:
                                    if p.string != "" and not p.string.endswith("vues") \
                                                        and not p.string.endswith("Aufrufe"):
                                        video['startDate'] = parse_date( p.string )
                            #get thumbnail image
                            for t in i.findAll( 'img', {"class":"thumbnail"}):
                                video['previewPictureURL'] = self.ARTE_WEB_ROOT + t['src']
                                video['previewPictureURL'] = video['previewPictureURL']\
                                                        .replace("/fr/", lang)
                            self.videos.append(video)
                        except:
                            continue
                    
                except Exception, why:
                    self.error = "Error in stage 3 : %s" % why
            
        """
        with open('datas.txt', 'w') as dat:
            for idx, video in enumerate(self.videos):
                print type(video['bigTitle'])
                txt = "%s   %s  %s\n" % (idx, video['bigTitle'], video['startDate'])
                txt = txt + u"\tThumbnail:  %s\n" % video['previewPictureURL']
                txt = txt + u"\tVideo:      %s\n\n" % video['targetURL']
                dat.write(txt.encode('utf-8', 'ignore'))
        """