dark_web_forums/Forums/Procrax/parser.py

__author__ = 'Helium'

# Here, we are importing the auxiliary functions to clean or convert data
from Forums.Utilities.utilities import *
from datetime import date
from datetime import timedelta
import re

# Here, we are importing BeautifulSoup to search through the HTML tree
from bs4 import BeautifulSoup, ResultSet, Tag

# This is the method to parse the Description Pages (one page to each topic in the Listing Pages)


def procrax_description_parser(soup: Tag):

    # Fields to be parsed

    topic = "-1"           # 0 topic name
    user = []              # 1 all users of each post
    addDate = []           # 2 all dated of each post
    feedback = []          # 3 all feedbacks of each vendor (this was found in just one Forum and with a number format)
    status = []            # 4 all user's authority in each post such as (adm, member, dangerous)
    reputation = []        # 5 all user's karma in each post (usually found as a number)
    sign = []              # 6 all user's signature in each post (usually a standard message after the content of the post)
    post = []              # 7 all messages of each post
    interest = []          # 8 all user's interest in each post
    image_user = []        # 9 all user avatars of each post
    image_post = []        # 10 all first images of each post

    # Finding the topic (should be just one coming from the Listing Page)

    li = soup.find("h1", {"class": "p-title-value"})
    topic = li.text
    
    thread: ResultSet[Tag] = soup.find("div", {"class": "block-body js-replyNewMessageContainer"}).find_all("article", {"data-author": True})

    for ipost in thread:
        username = ipost.find("h4", {"class": "message-name"}).text
        user.append(cleanString(username.strip()))
        
        date_posted = ipost.find("ul", {"class": "message-attribution-main listInline"}).find("time").get("datetime")
        datetime_obj = datetime.strptime(date_posted, "%Y-%m-%dT%H:%M:%S%z")
        addDate.append(datetime_obj)

        feedback.append("-1")
        
        user_status = ipost.find("h5", {"class": "userTitle message-userTitle"}).text
        status.append(cleanString(user_status.strip()))
        
        user_lvl = ipost.find("div", {"class": "afAwardLevel"})
        if user_lvl is not None:
            user_lvl = user_lvl.text
            reputation.append(cleanString(user_lvl.strip()))
        else:
            reputation.append('-1')
        
        sign.append("-1")
        
        user_post = ipost.find("article", {"class": "message-body js-selectToQuote"}).text
        post.append(cleanString(user_post.strip()))
        
        interest.append("-1")
        
        bbWrapper = ipost.find('div', {"class": "bbWrapper"})
        if bbWrapper is not None:
            img = bbWrapper.find('img')
            if img is not None:
               img = img.get('src').split('base64,')[-1]
            else:
               img = "-1"
        else:
            img = "-1"
        image_post.append(img)

        avatar = ipost.find("a", {"class": "avatar avatar--m"})
        if avatar is not None:
           img = avatar.find('img')
           if img is not None:
              img = img.get('src').split('base64,')[-1]
           else:
              img = "-1"
        else:
            img = "-1"
        image_user.append(img)

    # Populate the final variable (this should be a list with all fields scraped)

    row = (topic, user, status, reputation, interest, sign, post, feedback, addDate, image_user, image_post)

    # Sending the results

    return row

# This is the method to parse the Listing Pages (one page with many posts)

def procrax_listing_parser(soup: Tag):

    nm = 0                  # this variable should receive the number of topics
    forum: str = "Procrax"  # 0 *forum name
    board = "-1"            # 1 board name (the previous level of the topic in the Forum categorization tree.
                            # For instance: Security/Malware/Tools to hack Facebook. The board here should be Malware)

    author = []             # 2 all authors of each topic
    topic = []              # 3 all topics
    views = []              # 4 number of views of each topic
    posts = []              # 5 number of posts of each topic
    href = []               # 6this variable should receive all cleaned urls (we will use this to do the marge between
                            # Listing and Description pages)
    addDate = []            # 7 when the topic was created (difficult to find)
    image_author = []       # 8 all author avatars used in each topic

    # Finding the board (should be just one)
    li = soup.find("h1", {"class": "p-title-value"})
    board = cleanString(li.text.strip())

    threads_list = soup.find("div", {"class": "structItemContainer-group js-threadList"}).find_all("div", {"data-author": True})

    sticky = soup.find("div", {"class": "structItemContainer-group structItemContainer-group--sticky"})
    if sticky is not None:
        threads_list = sticky.find_all("div", {"data-author": True}) + threads_list
    
    nm = len(threads_list)
    
    for thread in threads_list:
        thread_title = thread.find("div", {"class": "structItem-title"}).text
        topic.append(cleanString(thread_title.strip()))

        author_icon = thread.find('a', {"class": "avatar avatar--s"})
        if author_icon != None:
            author_icon = author_icon.find('img')
            if author_icon != None:
               author_icon = author_icon.get('src')
               author_icon = author_icon.split('base64,')[-1]
            else:
               author_icon = "-1"
        else:
            author_icon = "-1"
        image_author.append(author_icon)

        thread_author = thread.get("data-author")
        author.append(cleanString(thread_author))
        
        thread_views = thread.find("dl", {"class": "pairs pairs--justified structItem-minor"}).find('dd').text
        thread_views = thread_views.lower().replace("k", "000")
        thread_views = thread_views.lower().replace("m", "000000")
        views.append(thread_views.strip())
        
        thread_replies = thread.find("dl", {"class": "pairs pairs--justified"}).find('dd').text
        # All threads contain one topic post and reply posts
        thread_total_posts = thread_replies.lower().replace("k", "000")
        posts.append(thread_total_posts.strip())
        
        thread_date = thread.find("li", {"class": "structItem-startDate"}).find("time").get("datetime")
        datetime_obj = datetime.strptime(thread_date, "%Y-%m-%dT%H:%M:%S%z")
        addDate.append(datetime_obj)
        
        thread_link: str = thread.find("div", {"class": "structItem-title"}).find('a', {'class': ''}).get('href')
        href.append(thread_link)
    
    
    return organizeTopics(
        forum=forum,
        nm=nm,
        board=board,
        author=author,
        topic=topic,
        views=views,
        posts=posts,
        addDate=addDate,
        href=href,
        image_author=image_author
    )


def procrax_links_parser(soup):

    # Returning all links that should be visited by the Crawler

    href = []

    listing = soup.find_all('div', {"class": "structItem-title"})

    for a in listing:
        link = a.find('a', {'class': ''}).get('href')

        href.append(link)

    return href