dark_web_forums/Forums/Cardingleaks/parser.py

__author__ = 'DarkWeb'

# 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 cardingleaks_description_parser(soup: Tag):

    # Fields to be parsed

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

    li = soup.find("h1", {"class": "p-title-value"})
    topic = cleanString(li.text.strip())

    post_list: ResultSet[Tag] = soup.find("div", {"class": "block-body js-replyNewMessageContainer"}).find_all("article", {"data-author": True})

    for ipost in post_list:
        username = ipost.get('data-author')
        user.append(username)
        
        user_status = ipost.find("h5", {"class": "userTitle message-userTitle"}).text
        status.append(cleanString(user_status.strip()))
        
        user_statistics: ResultSet[Tag] = ipost.find("div", {"class": "message-userExtras"}).find_all("dl", {"class": "pairs pairs--justified"})
        
        user_reputation = "-1"
        
        for stat in user_statistics:
            data_type = stat.find("span").get("data-original-title")
            if data_type == "Points":
                user_reputation = stat.find("dd").text
                break
            
        reputation.append(cleanString(user_reputation.strip()))
        
        interest.append("-1")
        
        sign.append("-1")
        
        user_post = ipost.find("div", {"class": "message-content js-messageContent"}).text
        post.append(cleanString(user_post.strip()))
        
        feedback.append("-1")
        
        datetime_text = ipost.find("ul", {"class": "message-attribution-main listInline"}).find("time").get("datetime")
        datetime_obj = datetime.strptime(datetime_text, "%Y-%m-%dT%H:%M:%S%z")
        addDate.append(datetime_obj)

        img = ipost.find('div', {"class": "message-avatar"}).find('img')
        img = img.get('src').split('base64,')[-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)

    # Sending the results

    return row

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

def cardingleaks_listing_parser(soup: Tag):

    nm = 0              # *this variable should receive the number of topics
    forum = "Cardingleaks"   # 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 = []           # 6 this 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)

    # Finding the board (should be just one)

    li = soup.find("h1", {"class": "p-title-value"})
    board = cleanString(li.text.strip())

    thread_list: ResultSet[Tag] = soup.find("div", {"class": "structItemContainer-group js-threadList"}).find_all("div", {"data-author": True})
    
    nm = len(thread_list)
    
    for thread in thread_list:
        thread_author = thread.get("data-author")
        author.append(thread_author)
        
        thread_topic = thread.find("div", {"class": "structItem-title"}).text
        topic.append(cleanString(thread_topic.strip()))
        
        thread_view = thread.find("dl", {"class": "pairs pairs--justified structItem-minor"}).find("dd").text
        # Context text view count (i.e., 8.8K) to numerical (i.e., 8800)
        if thread_view.find("K") > 0:
            thread_view = str(int(float(thread_view.replace("K", "")) * 1000))
        views.append(thread_view)
        
        thread_posts = thread.find("dl", {"class": "pairs pairs--justified"}).find("dd").text
        posts.append(cleanString(thread_posts.strip()))
        
        thread_href = thread.find("div", {"class": "structItem-title"}).find("a").get("href")
        href.append(thread_href)
        
        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)

    return organizeTopics(forum, nm, board, author, topic, views, posts, href, addDate)


def cardingleaks_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').get('href')

        href.append(link)

    return href