ail-framework/bin/crawlers/Crawler.py

#!/usr/bin/env python3
# -*-coding:UTF-8 -*

import os
import logging.config
import sys
import time

from pyail import PyAIL
from requests.exceptions import ConnectionError

sys.path.append(os.environ['AIL_BIN'])
##################################
# Import Project packages
##################################
from modules.abstract_module import AbstractModule
from lib import ail_logger
from lib import crawlers
from lib.ConfigLoader import ConfigLoader
from lib.objects import CookiesNames
from lib.objects import Etags
from lib.objects.Domains import Domain
from lib.objects import Favicons
from lib.objects.Items import Item
from lib.objects import Screenshots
from lib.objects import Titles
from trackers.Tracker_Yara import Tracker_Yara

logging.config.dictConfig(ail_logger.get_config(name='crawlers'))

class Crawler(AbstractModule):

    def __init__(self):
        super(Crawler, self, ).__init__()

        self.logger = logging.getLogger(f'{self.__class__.__name__}')

        # Waiting time in seconds between to message processed
        self.pending_seconds = 1

        self.tracker_yara = Tracker_Yara(queue=False)

        config_loader = ConfigLoader()

        self.default_har = config_loader.get_config_boolean('Crawler', 'default_har')
        self.default_screenshot = config_loader.get_config_boolean('Crawler', 'default_screenshot')
        self.default_depth_limit = config_loader.get_config_int('Crawler', 'default_depth_limit')

        ail_url_to_push_discovery = config_loader.get_config_str('Crawler', 'ail_url_to_push_onion_discovery')
        ail_key_to_push_discovery = config_loader.get_config_str('Crawler', 'ail_key_to_push_onion_discovery')
        if ail_url_to_push_discovery and ail_key_to_push_discovery:
            ail = PyAIL(ail_url_to_push_discovery, ail_key_to_push_discovery, ssl=False)
            if ail.ping_ail():
                self.ail_to_push_discovery = ail
        else:
            self.ail_to_push_discovery = None

        # TODO: LIMIT MAX NUMBERS OF CRAWLED PAGES

        # update hardcoded blacklist
        crawlers.load_blacklist()
        # update captures cache
        crawlers.reload_crawler_captures()

        self.crawler_scheduler = crawlers.CrawlerScheduler()

        # LACUS
        self.lacus = crawlers.get_lacus()
        self.is_lacus_up = crawlers.is_lacus_connected(delta_check=0)

        # Capture
        self.har = None
        self.screenshot = None
        self.root_item = None
        self.date = None
        self.items_dir = None
        self.original_domain = None
        self.domain = None

        # TODO Replace with warning list ???
        self.placeholder_screenshots = {'07244254f73e822bd4a95d916d8b27f2246b02c428adc29082d09550c6ed6e1a'   # blank
                                        '27e14ace10b0f96acd2bd919aaa98a964597532c35b6409dff6cc8eec8214748',  # not found
                                        '3e66bf4cc250a68c10f8a30643d73e50e68bf1d4a38d4adc5bfc4659ca2974c0'}  # 404

        # Send module state to logs
        self.logger.info('Crawler initialized')

    def refresh_lacus_status(self):
        try:
            lacus_up = self.is_lacus_up
            self.is_lacus_up = crawlers.get_lacus().is_up
            # refresh lacus
            if not lacus_up and self.is_lacus_up:
                self.lacus = crawlers.get_lacus()
        except:
            self.is_lacus_up = False
        if not self.is_lacus_up:
            print("Can't reach lacus server", int(time.time()))
            time.sleep(30)

    def print_crawler_start_info(self, url, domain_url):
        print()
        print()
        print('\033[92m------------------START CRAWLER------------------\033[0m')
        print(f'crawler type:     {self.domain}')
        print('\033[92m-------------------------------------------------\033[0m')
        print(f'url:         {url}')
        print(f'domain:      {self.domain}')
        print(f'domain_url:  {domain_url}')
        print()

    def get_message(self):
        # Crawler Scheduler
        self.crawler_scheduler.update_queue()
        self.crawler_scheduler.process_queue()

        self.refresh_lacus_status()  # TODO LOG ERROR
        if not self.is_lacus_up:
            return None

        # Check if a new Capture can be Launched
        if crawlers.get_nb_crawler_captures() < crawlers.get_crawler_max_captures():
            task_row = crawlers.add_task_to_lacus_queue()
            if task_row:
                task, priority = task_row
                task.start()
                task_uuid = task.uuid
                try:
                    self.enqueue_capture(task_uuid, priority)
                except ConnectionError:
                    print(task_row)
                    task = crawlers.CrawlerTask(task_uuid)
                    task.add_to_db_crawler_queue(priority)
                    self.refresh_lacus_status()
                    return None

        # Get CrawlerCapture Object
        capture = crawlers.get_crawler_capture()
        if capture:
            try:
                status = self.lacus.get_capture_status(capture.uuid)
                if status == crawlers.CaptureStatus.DONE:
                    return capture
                elif status == crawlers.CaptureStatus.UNKNOWN:
                    capture_start = capture.get_start_time(r_str=False)
                    if capture_start == 0:
                        task = capture.get_task()
                        task.delete()
                        capture.delete()
                        self.logger.warning(f'capture UNKNOWN ERROR STATE, {task.uuid} Removed from queue')
                        return None
                    if int(time.time()) - capture_start > 600:  # TODO ADD in new crawler config
                        task = capture.get_task()
                        task.reset()
                        capture.delete()
                        self.logger.warning(f'capture UNKNOWN Timeout, {task.uuid} Send back in queue')
                    else:
                        capture.update(status)
                else:
                    capture.update(status)
                    print(capture.uuid, crawlers.CaptureStatus(status).name, int(time.time()))

            except ConnectionError:
                print(capture.uuid)
                capture.update(-1)
                self.refresh_lacus_status()

        time.sleep(self.pending_seconds)

    def enqueue_capture(self, task_uuid, priority):
        task = crawlers.CrawlerTask(task_uuid)
        # print(task)
        # task = {
        #         'uuid': task_uuid,
        #         'url': 'https://foo.be',
        #         'domain': 'foo.be',
        #         'depth': 1,
        #         'har': True,
        #         'screenshot': True,
        #         'user_agent': crawlers.get_default_user_agent(),
        #         'cookiejar': [],
        #         'header': '',
        #         'proxy': 'force_tor',
        #         'parent': 'manual',
        # }

        url = task.get_url()
        force = priority != 0
        # TODO timeout

        # TODO HEADER
        # capture_uuid = self.lacus.enqueue(url='https://cpg.circl.lu:7000',
        #                                   force=force,
        #                                   general_timeout_in_sec=120)

        # with_favicon = True,
        capture_uuid = self.lacus.enqueue(url=url,
                                          depth=task.get_depth(),
                                          user_agent=task.get_user_agent(),
                                          proxy=task.get_proxy(),
                                          cookies=task.get_cookies(),
                                          with_favicon=True,
                                          force=force,
                                          general_timeout_in_sec=90)  # TODO increase timeout if onion ????

        crawlers.create_capture(capture_uuid, task_uuid)
        print(task.uuid, capture_uuid, 'launched')

        if self.ail_to_push_discovery:

            if task.get_depth() == 1 and priority < 10 and task.get_domain().endswith('.onion'):
                har = task.get_har()
                screenshot = task.get_screenshot()
                # parent_id = task.get_parent()
                # if parent_id != 'manual' and parent_id != 'auto':
                #     parent = parent_id[19:-36]
                # else:
                #     parent = 'AIL_capture'

                if not url:
                    raise Exception(f'Error: url is None, {task.uuid}, {capture_uuid}, {url}')

                self.ail_to_push_discovery.add_crawler_capture(task_uuid, capture_uuid, url, har=har,  # parent=parent,
                                                               screenshot=screenshot, depth_limit=1, proxy='force_tor')
                print(task.uuid, capture_uuid, url, 'Added to ail_to_push_discovery')
        return capture_uuid

    # CRAWL DOMAIN
    def compute(self, capture):
        print('saving capture', capture.uuid)

        task = capture.get_task()
        domain = task.get_domain()
        print(domain)
        if not domain:
            if self.debug:
                raise Exception(f'Error: domain {domain} - task {task.uuid} - capture {capture.uuid}')
            else:
                self.logger.critical(f'Error: domain {domain} - task {task.uuid} - capture {capture.uuid}')
                print(f'Error: domain {domain}')
            return None

        self.domain = Domain(domain)
        self.original_domain = Domain(domain)

        epoch = int(time.time())
        parent_id = task.get_parent()

        entries = self.lacus.get_capture(capture.uuid)

        print(entries.get('status'))
        self.har = task.get_har()
        self.screenshot = task.get_screenshot()
        # DEBUG
        # self.har = True
        # self.screenshot = True
        self.date = crawlers.get_current_date(separator=True)
        self.items_dir = crawlers.get_date_crawled_items_source(self.date)
        self.root_item = None

        # Save Capture
        self.save_capture_response(parent_id, entries)

        self.domain.update_daterange(self.date.replace('/', ''))
        # Origin + History + tags
        if self.root_item:
            self.domain.set_last_origin(parent_id)
            self.domain.update_vanity_cluster()
            # Tags
            for tag in task.get_tags():
                self.domain.add_tag(tag)
        self.domain.add_history(epoch, root_item=self.root_item)

        if self.domain != self.original_domain:
            self.original_domain.update_daterange(self.date.replace('/', ''))
            if self.root_item:
                self.original_domain.set_last_origin(parent_id)
                # Tags
                for tag in task.get_tags():
                    self.domain.add_tag(tag)
            self.original_domain.add_history(epoch, root_item=self.root_item)
            # crawlers.update_last_crawled_domain(self.original_domain.get_domain_type(), self.original_domain.id, epoch)

        crawlers.update_last_crawled_domain(self.domain.get_domain_type(), self.domain.id, epoch)
        print('capture:', capture.uuid, 'completed')
        print('task:   ', task.uuid, 'completed')
        print()
        task.remove()

    def save_capture_response(self, parent_id, entries):
        print(entries.keys())
        if 'error' in entries:
            # TODO IMPROVE ERROR MESSAGE
            self.logger.warning(str(entries['error']))
            print(entries.get('error'))
            if entries.get('html'):
                print('retrieved content')
                # print(entries.get('html'))

        if 'last_redirected_url' in entries and entries.get('last_redirected_url'):
            last_url = entries['last_redirected_url']
            unpacked_last_url = crawlers.unpack_url(last_url)
            current_domain = unpacked_last_url['domain']
            # REDIRECTION TODO CHECK IF TYPE CHANGE
            if current_domain != self.domain.id and not self.root_item:
                self.logger.warning(f'External redirection {self.domain.id} -> {current_domain}')
                print(f'External redirection {self.domain.id} -> {current_domain}')
                if not self.root_item:
                    self.domain = Domain(current_domain)
        # TODO LAST URL
        # FIXME
        else:
            last_url = f'http://{self.domain.id}'

        if 'html' in entries and entries.get('html'):
            item_id = crawlers.create_item_id(self.items_dir, self.domain.id)
            item = Item(item_id)
            print(item.id)

            gzip64encoded = crawlers.get_gzipped_b64_item(item.id, entries['html'])
            # send item to Global
            relay_message = f'crawler {gzip64encoded}'
            self.add_message_to_queue(obj=item, message=relay_message, queue='Importers')

            # Tag # TODO replace me with metadata to tags
            msg = f'infoleak:submission="crawler"'  # TODO FIXME
            self.add_message_to_queue(obj=item, message=msg, queue='Tags')

            # TODO replace me with metadata to add
            crawlers.create_item_metadata(item_id, last_url, parent_id)
            if self.root_item is None:
                self.root_item = item_id
            parent_id = item_id

            title_content = crawlers.extract_title_from_html(entries['html'])
            if title_content:
                title = Titles.create_title(title_content)
                title.add(item.get_date(), item)
                # Tracker
                self.tracker_yara.compute_manual(title)
                if not title.is_tags_safe():
                    unsafe_tag = 'dark-web:topic="pornography-child-exploitation"'
                    self.domain.add_tag(unsafe_tag)
                    item.add_tag(unsafe_tag)

            # SCREENSHOT
            if self.screenshot:
                if 'png' in entries and entries.get('png'):
                    screenshot = Screenshots.create_screenshot(entries['png'], b64=False)
                    if screenshot:
                        if not screenshot.is_tags_safe():
                            unsafe_tag = 'dark-web:topic="pornography-child-exploitation"'
                            self.domain.add_tag(unsafe_tag)
                            item.add_tag(unsafe_tag)
                        # Remove Placeholder pages # TODO Replace with warning list ???
                        if screenshot.id not in self.placeholder_screenshots:
                            # Create Correlations
                            screenshot.add_correlation('item', '', item_id)
                            screenshot.add_correlation('domain', '', self.domain.id)
            # HAR
            if self.har:
                if 'har' in entries and entries.get('har'):
                    har_id = crawlers.create_har_id(self.date, item_id)
                    crawlers.save_har(har_id, entries['har'])
                    for cookie_name in crawlers.extract_cookies_names_from_har(entries['har']):
                        print(cookie_name)
                        cookie = CookiesNames.create(cookie_name)
                        cookie.add(self.date.replace('/', ''), self.domain)
                    for etag_content in crawlers.extract_etag_from_har(entries['har']):
                        print(etag_content)
                        etag = Etags.create(etag_content)
                        etag.add(self.date.replace('/', ''), self.domain)
                    crawlers.extract_hhhash(entries['har'], self.domain.id, self.date.replace('/', ''))

            # FAVICON
            if entries.get('potential_favicons'):
                for favicon in entries['potential_favicons']:
                    fav = Favicons.create(favicon)
                    fav.add(item.get_date(), item)

        # Next Children
        entries_children = entries.get('children')
        if entries_children:
            for children in entries_children:
                self.save_capture_response(parent_id, children)


if __name__ == '__main__':
    module = Crawler()
    module.debug = True
    module.run()