chg: Refactor code organisation

bin/async_scrape.py

@@ -5,9 +5,8 @@ from pathlib import Path
 import logging
 
 from lookyloo.abstractmanager import AbstractManager
-from lookyloo.helpers import get_homedir, get_socket_path
-from lookyloo import scrape
-from redis import Redis
+from lookyloo.helpers import get_homedir
+from lookyloo.lookyloo import Lookyloo
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO, datefmt='%I:%M:%S')
@@ -19,15 +18,10 @@ class AsyncScraper(AbstractManager):
         super().__init__(loglevel)
         if not storage_directory:
             self.storage_directory = get_homedir() / 'scraped'
-        self.redis = Redis(unix_socket_path=get_socket_path('cache'), decode_responses=True)
+        self.lookyloo = Lookyloo(loglevel=loglevel)
 
     def _to_run_forever(self):
-        uuid = self.redis.spop('to_scrape')
-        if not uuid:
-            return
-        to_scrape = self.redis.hgetall(uuid)
-        to_scrape['perma_uuid'] = uuid
-        scrape(**to_scrape)
+        self.lookyloo.process_scrape_queue()
 
 
 if __name__ == '__main__':

bin/start.py

@@ -10,4 +10,5 @@ if __name__ == '__main__':
     get_homedir()
     p = Popen(['run_backend.py', '--start'])
     p.wait()
+    Popen(['async_scrape.py'])
     Popen(['start_website.py'])

bin/start_website.py

@@ -5,10 +5,10 @@ from subprocess import Popen
 from lookyloo.helpers import get_homedir
 
 if __name__ == '__main__':
-    website_dir = get_homedir()
+    website_dir = get_homedir() / 'website'
     Popen([str(website_dir / '3rdparty.sh')], cwd=website_dir)
     try:
-        Popen(['gunicorn', '--worker-class', 'eventlet', '-w', '10', '-b', '0.0.0.0:5100', 'lookyloo:app'],
+        Popen(['gunicorn', '--worker-class', 'eventlet', '-w', '10', '-b', '0.0.0.0:5100', 'web:app'],
               cwd=website_dir).communicate()
     except KeyboardInterrupt:
         print('Stopping gunicorn.')

lookyloo.ini

@@ -1,13 +0,0 @@
-[uwsgi]
-module = wsgi:app
-
-master = true
-processes = 5
-
-socket = lookyloo.sock
-chmod-socket = 660
-vacuum = true
-
-die-on-term = true
-
-wsgi-disable-file-wrapper = true

lookyloo/__init__.py

@@ -1,264 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-
-import json
-
-from har2tree import CrawledTree
-from scrapysplashwrapper import crawl
-
-from flask import Flask, render_template, request, session, send_file, redirect, url_for, Response
-from flask_bootstrap import Bootstrap
-
-from datetime import datetime
-
-import pickle
-import tempfile
-import pathlib
-import time
-
-from zipfile import ZipFile, ZIP_DEFLATED
-from io import BytesIO
-import base64
-import os
-from uuid import uuid4
-
-from pysanejs import SaneJS
-
-from .helpers import get_homedir, get_socket_path
-from redis import Redis
-
-app = Flask(__name__)
-
-
-secret_file_path = get_homedir() / 'secret_key'
-
-if not secret_file_path.exists() or secret_file_path.stat().st_size < 64:
-    with secret_file_path.open('wb') as f:
-        f.write(os.urandom(64))
-
-with secret_file_path.open('rb') as f:
-    app.config['SECRET_KEY'] = f.read()
-
-Bootstrap(app)
-app.config['BOOTSTRAP_SERVE_LOCAL'] = True
-app.config['SESSION_COOKIE_NAME'] = 'lookyloo'
-app.debug = False
-
-HAR_DIR = get_homedir() / 'scraped'
-HAR_DIR.mkdir(parents=True, exist_ok=True)
-
-SPLASH = 'http://127.0.0.1:8050'
-SANE_JS = 'http://127.0.0.1:5007'
-
-if SANE_JS:
-    try:
-        sanejs = SaneJS(SANE_JS)
-        if sanejs.is_up:
-            has_sane_js = True
-        else:
-            has_sane_js = False
-    except Exception:
-        has_sane_js = False
-
-r = Redis(unix_socket_path=get_socket_path('cache'), decode_responses=True)
-
-
-def get_report_dirs():
-    # Cleanup HAR_DIR of failed runs.
-    for report_dir in HAR_DIR.iterdir():
-        if report_dir.is_dir() and not report_dir.iterdir():
-            report_dir.rmdir()
-        if not (report_dir / 'uuid').exists():
-            # Create uuid if missing
-            with (report_dir / 'uuid').open('w') as f:
-                f.write(str(uuid4()))
-    return sorted(HAR_DIR.iterdir(), reverse=True)
-
-
-def get_lookup_dirs():
-    # Build lookup table trees
-    lookup_dirs = {}
-    for report_dir in get_report_dirs():
-        with (report_dir / 'uuid').open() as f:
-            lookup_dirs[f.read().strip()] = report_dir
-    return lookup_dirs
-
-
-def cleanup_old_tmpfiles():
-    for tmpfile in pathlib.Path(tempfile.gettempdir()).glob('lookyloo*'):
-        if time.time() - tmpfile.stat().st_atime > 36000:
-            tmpfile.unlink()
-
-
-def load_image(report_dir):
-    with open(list(report_dir.glob('*.png'))[0], 'rb') as f:
-        return BytesIO(f.read())
-
-
-def load_tree(report_dir):
-    session.clear()
-    har_files = sorted(report_dir.glob('*.har'))
-    ct = CrawledTree(har_files)
-    ct.find_parents()
-    ct.join_trees()
-    temp = tempfile.NamedTemporaryFile(prefix='lookyloo', delete=False)
-    pickle.dump(ct, temp)
-    temp.close()
-    session["tree"] = temp.name
-    return ct.to_json(), ct.start_time.isoformat(), ct.user_agent, ct.root_url
-
-
-def sane_js_query(sha512: str):
-    if has_sane_js:
-        return sanejs.sha512(sha512)
-    return {'response': []}
-
-
-def scrape(url, depth: int=1, user_agent: str=None, perma_uuid: str=None):
-    if not url.startswith('http'):
-        url = f'http://{url}'
-    items = crawl(SPLASH, url, depth, user_agent=user_agent, log_enabled=True, log_level='INFO')
-    if not items:
-        # broken
-        pass
-    if not perma_uuid:
-        perma_uuid = str(uuid4())
-    width = len(str(len(items)))
-    dirpath = HAR_DIR / datetime.now().isoformat()
-    dirpath.mkdir()
-    for i, item in enumerate(items):
-        harfile = item['har']
-        png = base64.b64decode(item['png'])
-        child_frames = item['childFrames']
-        html = item['html']
-        with (dirpath / '{0:0{width}}.har'.format(i, width=width)).open('w') as f:
-            json.dump(harfile, f)
-        with (dirpath / '{0:0{width}}.png'.format(i, width=width)).open('wb') as f:
-            f.write(png)
-        with (dirpath / '{0:0{width}}.html'.format(i, width=width)).open('w') as f:
-            f.write(html)
-        with (dirpath / '{0:0{width}}.frames.json'.format(i, width=width)).open('w') as f:
-            json.dump(child_frames, f)
-        with (dirpath / 'uuid').open('w') as f:
-            f.write(perma_uuid)
-    return perma_uuid
-
-
-@app.route('/submit', methods=['POST', 'GET'])
-def submit():
-    to_query = request.get_json(force=True)
-    perma_uuid = str(uuid4())
-    p = r.pipeline()
-    p.hmset(perma_uuid, to_query)
-    p.sadd('to_scrape', perma_uuid)
-    p.execute()
-    return Response(perma_uuid, mimetype='text/text')
-
-
-@app.route('/scrape', methods=['GET', 'POST'])
-def scrape_web():
-    if request.form.get('url'):
-        perma_uuid = scrape(request.form.get('url'), request.form.get('depth'))
-        return redirect(url_for('tree', tree_uuid=perma_uuid))
-    return render_template('scrape.html')
-
-
-@app.route('/tree/hostname/<node_uuid>/text', methods=['GET'])
-def hostnode_details_text(node_uuid):
-    with open(session["tree"], 'rb') as f:
-        ct = pickle.load(f)
-    hostnode = ct.root_hartree.get_host_node_by_uuid(node_uuid)
-    urls = []
-    for url in hostnode.urls:
-        urls.append(url.name)
-    content = '''# URLs
-
-{}
-'''.format('\n'.join(urls))
-    to_return = BytesIO(content.encode())
-    to_return.seek(0)
-    return send_file(to_return, mimetype='text/markdown',
-                     as_attachment=True, attachment_filename='file.md')
-
-
-@app.route('/tree/hostname/<node_uuid>', methods=['GET'])
-def hostnode_details(node_uuid):
-    with open(session["tree"], 'rb') as f:
-        ct = pickle.load(f)
-    hostnode = ct.root_hartree.get_host_node_by_uuid(node_uuid)
-    urls = []
-    for url in hostnode.urls:
-        if hasattr(url, 'body_hash'):
-            sane_js_r = sane_js_query(url.body_hash)
-            if sane_js_r.get('response'):
-                url.add_feature('sane_js_details', sane_js_r['response'])
-                print('######## SANEJS ##### ', url.sane_js_details)
-        urls.append(url.to_json())
-    return json.dumps(urls)
-
-
-@app.route('/tree/url/<node_uuid>', methods=['GET'])
-def urlnode_details(node_uuid):
-    with open(session["tree"], 'rb') as f:
-        ct = pickle.load(f)
-    urlnode = ct.root_hartree.get_url_node_by_uuid(node_uuid)
-    to_return = BytesIO()
-    got_content = False
-    if hasattr(urlnode, 'body'):
-        body_content = urlnode.body.getvalue()
-        if body_content:
-            got_content = True
-            with ZipFile(to_return, 'w', ZIP_DEFLATED) as zfile:
-                zfile.writestr(urlnode.filename, urlnode.body.getvalue())
-    if not got_content:
-        with ZipFile(to_return, 'w', ZIP_DEFLATED) as zfile:
-            zfile.writestr('file.txt', b'Response body empty')
-    to_return.seek(0)
-    return send_file(to_return, mimetype='application/zip',
-                     as_attachment=True, attachment_filename='file.zip')
-
-
-@app.route('/tree/<string:tree_uuid>/image', methods=['GET'])
-def image(tree_uuid):
-    lookup_dirs = get_lookup_dirs()
-    report_dir = lookup_dirs[tree_uuid]
-    to_return = load_image(report_dir)
-    return send_file(to_return, mimetype='image/png',
-                     as_attachment=True, attachment_filename='image.png')
-
-
-@app.route('/tree/<string:tree_uuid>', methods=['GET'])
-def tree(tree_uuid):
-    lookup_dirs = get_lookup_dirs()
-    report_dir = lookup_dirs[tree_uuid]
-    tree_json, start_time, user_agent, root_url = load_tree(report_dir)
-    return render_template('tree.html', tree_json=tree_json, start_time=start_time,
-                           user_agent=user_agent, root_url=root_url, tree_uuid=tree_uuid)
-
-
-@app.route('/', methods=['GET'])
-def index():
-    if request.method == 'HEAD':
-        # Just returns ack if the webserver is running
-        return 'Ack'
-    cleanup_old_tmpfiles()
-    session.clear()
-    titles = []
-    if not HAR_DIR.exists():
-        HAR_DIR.mkdir(parents=True)
-    for report_dir in get_report_dirs():
-        har_files = sorted(report_dir.glob('*.har'))
-        if not har_files:
-            continue
-        with har_files[0].open() as f:
-            j = json.load(f)
-            title = j['log']['pages'][0]['title']
-        with (report_dir / 'uuid').open() as f:
-            uuid = f.read().strip()
-        titles.append((uuid, title))
-
-    return render_template('index.html', titles=titles)
-
-
-if __name__ == '__main__':
-    app.run(port=5001, threaded=True)

lookyloo/lookyloo.py

@@ -0,0 +1,137 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import json
+
+from scrapysplashwrapper import crawl
+from har2tree import CrawledTree
+import pickle
+
+from datetime import datetime
+
+import tempfile
+import pathlib
+import time
+
+from io import BytesIO
+import base64
+from uuid import uuid4
+
+from pysanejs import SaneJS
+
+from pathlib import Path
+from .helpers import get_homedir, get_socket_path
+from redis import Redis
+
+import logging
+
+
+class Lookyloo():
+
+    def __init__(self, splash_url: str='http://127.0.0.1:8050', loglevel: int=logging.DEBUG):
+        self.__init_logger(loglevel)
+        self.redis = Redis(unix_socket_path=get_socket_path('cache'), decode_responses=True)
+        self.scrape_dir = get_homedir() / 'scraped'
+        self.splash_url = splash_url
+        if not self.scrape_dir.exists():
+            self.scrape_dir.mkdir(parents=True, exist_ok=True)
+
+        # Try to reach sanejs
+        self.sanejs = SaneJS()
+        if not self.sanejs.is_up:
+            self.sanejs = None
+
+    def __init_logger(self, loglevel) -> None:
+        self.logger = logging.getLogger(f'{self.__class__.__name__}')
+        self.logger.setLevel(loglevel)
+
+    @property
+    def report_dirs(self):
+        for report_dir in self.scrape_dir.iterdir():
+            if report_dir.is_dir() and not report_dir.iterdir():
+                # Cleanup self.scrape_dir of failed runs.
+                report_dir.rmdir()
+            if not (report_dir / 'uuid').exists():
+                # Create uuid if missing
+                with (report_dir / 'uuid').open('w') as f:
+                    f.write(str(uuid4()))
+        return sorted(self.scrape_dir.iterdir(), reverse=True)
+
+    @property
+    def lookup_dirs(self):
+        # Build lookup table trees
+        lookup_dirs = {}
+        for report_dir in self.report_dirs:
+            with (report_dir / 'uuid').open() as f:
+                lookup_dirs[f.read().strip()] = report_dir
+        return lookup_dirs
+
+    def enqueue_scrape(self, query: dict):
+        perma_uuid = str(uuid4())
+        p = self.redis.pipeline()
+        p.hmset(perma_uuid, query)
+        p.sadd('to_scrape', perma_uuid)
+        p.execute()
+        return perma_uuid
+
+    def process_scrape_queue(self):
+        uuid = self.redis.spop('to_scrape')
+        if not uuid:
+            return
+        to_scrape = self.redis.hgetall(uuid)
+        self.redis.delete(uuid)
+        to_scrape['perma_uuid'] = uuid
+        self.scrape(**to_scrape)
+
+    def load_tree(self, report_dir: Path):
+        har_files = sorted(report_dir.glob('*.har'))
+        ct = CrawledTree(har_files)
+        ct.find_parents()
+        ct.join_trees()
+        temp = tempfile.NamedTemporaryFile(prefix='lookyloo', delete=False)
+        pickle.dump(ct, temp)
+        temp.close()
+        return temp.name, ct.to_json(), ct.start_time.isoformat(), ct.user_agent, ct.root_url
+
+    def cleanup_old_tmpfiles(self):
+        for tmpfile in pathlib.Path(tempfile.gettempdir()).glob('lookyloo*'):
+            if time.time() - tmpfile.stat().st_atime > 36000:
+                tmpfile.unlink()
+
+    def load_image(self, report_dir):
+        with open(list(report_dir.glob('*.png'))[0], 'rb') as f:
+            return BytesIO(f.read())
+
+    def sane_js_query(self, sha512: str):
+        if self.sanejs:
+            return self.sanejs.sha512(sha512)
+        return {'response': []}
+
+    def scrape(self, url, depth: int=1, user_agent: str=None, perma_uuid: str=None):
+        if not url.startswith('http'):
+            url = f'http://{url}'
+        items = crawl(self.splash_url, url, depth, user_agent=user_agent, log_enabled=True, log_level='INFO')
+        if not items:
+            # broken
+            pass
+        if not perma_uuid:
+            perma_uuid = str(uuid4())
+        width = len(str(len(items)))
+        dirpath = self.scrape_dir / datetime.now().isoformat()
+        dirpath.mkdir()
+        for i, item in enumerate(items):
+            harfile = item['har']
+            png = base64.b64decode(item['png'])
+            child_frames = item['childFrames']
+            html = item['html']
+            with (dirpath / '{0:0{width}}.har'.format(i, width=width)).open('w') as f:
+                json.dump(harfile, f)
+            with (dirpath / '{0:0{width}}.png'.format(i, width=width)).open('wb') as f:
+                f.write(png)
+            with (dirpath / '{0:0{width}}.html'.format(i, width=width)).open('w') as f:
+                f.write(html)
+            with (dirpath / '{0:0{width}}.frames.json'.format(i, width=width)).open('w') as f:
+                json.dump(child_frames, f)
+            with (dirpath / 'uuid').open('w') as f:
+                f.write(perma_uuid)
+        return perma_uuid

website/3rdparty.sh

@@ -5,8 +5,8 @@ set -x
 
 mkdir -p web/static/
 
-wget https://d3js.org/d3.v5.js -O lookyloo/static/d3.v5.js
+wget https://d3js.org/d3.v5.js -O web/static/d3.v5.js
 
 FileSaver="5733e40e5af936eb3f48554cf6a8a7075d71d18a"
 
-wget https://cdn.rawgit.com/eligrey/FileSaver.js/${FileSaver}/FileSaver.js -O lookyloo/static/FileSaver.js
+wget https://cdn.rawgit.com/eligrey/FileSaver.js/${FileSaver}/FileSaver.js -O web/static/FileSaver.js

website/web/__init__.py

@@ -0,0 +1,151 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import json
+import pickle
+from zipfile import ZipFile, ZIP_DEFLATED
+from io import BytesIO
+import os
+
+from flask import Flask, render_template, request, session, send_file, redirect, url_for, Response
+from flask_bootstrap import Bootstrap
+
+from lookyloo.helpers import get_homedir
+from lookyloo.lookyloo import Lookyloo
+
+
+app = Flask(__name__)
+
+secret_file_path = get_homedir() / 'secret_key'
+
+if not secret_file_path.exists() or secret_file_path.stat().st_size < 64:
+    with secret_file_path.open('wb') as f:
+        f.write(os.urandom(64))
+
+with secret_file_path.open('rb') as f:
+    app.config['SECRET_KEY'] = f.read()
+
+Bootstrap(app)
+app.config['BOOTSTRAP_SERVE_LOCAL'] = True
+app.config['SESSION_COOKIE_NAME'] = 'lookyloo'
+app.debug = False
+
+lookyloo = Lookyloo()
+
+
+# keep
+def load_tree(report_dir):
+    session.clear()
+    temp_file_name, tree_json, tree_time, tree_ua, tree_root_url = lookyloo.load_tree(report_dir)
+    session["tree"] = temp_file_name
+    return tree_json, tree_time, tree_ua, tree_root_url
+
+
+@app.route('/submit', methods=['POST', 'GET'])
+def submit():
+    to_query = request.get_json(force=True)
+    perma_uuid = lookyloo.enqueue_scrape(to_query)
+    return Response(perma_uuid, mimetype='text/text')
+
+
+@app.route('/scrape', methods=['GET', 'POST'])
+def scrape_web():
+    if request.form.get('url'):
+        perma_uuid = lookyloo.scrape(request.form.get('url'), request.form.get('depth'))
+        return redirect(url_for('tree', tree_uuid=perma_uuid))
+    return render_template('scrape.html')
+
+
+@app.route('/tree/hostname/<node_uuid>/text', methods=['GET'])
+def hostnode_details_text(node_uuid):
+    with open(session["tree"], 'rb') as f:
+        ct = pickle.load(f)
+    hostnode = ct.root_hartree.get_host_node_by_uuid(node_uuid)
+    urls = []
+    for url in hostnode.urls:
+        urls.append(url.name)
+    content = '''# URLs
+
+{}
+'''.format('\n'.join(urls))
+    to_return = BytesIO(content.encode())
+    to_return.seek(0)
+    return send_file(to_return, mimetype='text/markdown',
+                     as_attachment=True, attachment_filename='file.md')
+
+
+@app.route('/tree/hostname/<node_uuid>', methods=['GET'])
+def hostnode_details(node_uuid):
+    with open(session["tree"], 'rb') as f:
+        ct = pickle.load(f)
+    hostnode = ct.root_hartree.get_host_node_by_uuid(node_uuid)
+    urls = []
+    for url in hostnode.urls:
+        if hasattr(url, 'body_hash'):
+            sane_js_r = lookyloo.sane_js_query(url.body_hash)
+            if sane_js_r.get('response'):
+                url.add_feature('sane_js_details', sane_js_r['response'])
+                print('######## SANEJS ##### ', url.sane_js_details)
+        urls.append(url.to_json())
+    return json.dumps(urls)
+
+
+@app.route('/tree/url/<node_uuid>', methods=['GET'])
+def urlnode_details(node_uuid):
+    with open(session["tree"], 'rb') as f:
+        ct = pickle.load(f)
+    urlnode = ct.root_hartree.get_url_node_by_uuid(node_uuid)
+    to_return = BytesIO()
+    got_content = False
+    if hasattr(urlnode, 'body'):
+        body_content = urlnode.body.getvalue()
+        if body_content:
+            got_content = True
+            with ZipFile(to_return, 'w', ZIP_DEFLATED) as zfile:
+                zfile.writestr(urlnode.filename, urlnode.body.getvalue())
+    if not got_content:
+        with ZipFile(to_return, 'w', ZIP_DEFLATED) as zfile:
+            zfile.writestr('file.txt', b'Response body empty')
+    to_return.seek(0)
+    return send_file(to_return, mimetype='application/zip',
+                     as_attachment=True, attachment_filename='file.zip')
+
+
+@app.route('/tree/<string:tree_uuid>/image', methods=['GET'])
+def image(tree_uuid):
+    lookup_dirs = lookyloo.lookup_dirs
+    report_dir = lookup_dirs[tree_uuid]
+    to_return = lookyloo.load_image(report_dir)
+    return send_file(to_return, mimetype='image/png',
+                     as_attachment=True, attachment_filename='image.png')
+
+
+@app.route('/tree/<string:tree_uuid>', methods=['GET'])
+def tree(tree_uuid):
+    lookup_dirs = lookyloo.lookup_dirs
+    report_dir = lookup_dirs[tree_uuid]
+    tree_json, start_time, user_agent, root_url = load_tree(report_dir)
+    return render_template('tree.html', tree_json=tree_json, start_time=start_time,
+                           user_agent=user_agent, root_url=root_url, tree_uuid=tree_uuid)
+
+
+@app.route('/', methods=['GET'])
+def index():
+    if request.method == 'HEAD':
+        # Just returns ack if the webserver is running
+        return 'Ack'
+    lookyloo.cleanup_old_tmpfiles()
+    session.clear()
+    titles = []
+    for report_dir in lookyloo.report_dirs:
+        har_files = sorted(report_dir.glob('*.har'))
+        if not har_files:
+            continue
+        with har_files[0].open() as f:
+            j = json.load(f)
+            title = j['log']['pages'][0]['title']
+        with (report_dir / 'uuid').open() as f:
+            uuid = f.read().strip()
+        titles.append((uuid, title))
+
+    return render_template('index.html', titles=titles)

wsgi.py

@@ -1,7 +0,0 @@
-#!/usr/bin/env python
-# -*- coding: utf-8 -*-
-
-from lookyloo import app
-
-if __name__ == "__main__":
-    app.run()