chg: use template

bin/archiver.py

@@ -10,9 +10,8 @@ from typing import Dict, List
 
 from redis import Redis
 
-from lookyloo.abstractmanager import AbstractManager
-from lookyloo.helpers import (get_captures_dir, get_config, get_homedir,
-                              get_socket_path)
+from lookyloo.default import AbstractManager, get_config, get_homedir, get_socket_path
+from lookyloo.helpers import get_captures_dir
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO)

bin/async_capture.py

@@ -16,10 +16,8 @@ from defang import refang  # type: ignore
 from redis import Redis
 from scrapysplashwrapper import crawl
 
-from lookyloo.abstractmanager import AbstractManager
-from lookyloo.helpers import (get_captures_dir, get_config, get_socket_path,
-                              get_splash_url, load_cookies, safe_create_dir,
-                              splash_status)
+from lookyloo.default import AbstractManager, get_config, get_socket_path, safe_create_dir
+from lookyloo.helpers import get_captures_dir, get_splash_url, load_cookies, splash_status
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO)

bin/background_indexer.py

@@ -5,7 +5,7 @@ import logging
 import os
 from datetime import datetime, timedelta
 
-from lookyloo.abstractmanager import AbstractManager
+from lookyloo.default import AbstractManager
 from lookyloo.exceptions import MissingUUID, NoValidHarFile
 from lookyloo.lookyloo import Lookyloo
 

bin/background_processing.py

@@ -10,9 +10,7 @@ from typing import Any, Dict
 from redis import Redis
 from werkzeug.useragents import UserAgent
 
-from lookyloo.abstractmanager import AbstractManager
-from lookyloo.helpers import (get_config, get_homedir, get_socket_path,
-                              safe_create_dir)
+from lookyloo.default import AbstractManager, get_config, get_homedir, get_socket_path, safe_create_dir
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO)

bin/run_backend.py

@@ -11,7 +11,7 @@ from typing import List, Optional, Union
 from redis import Redis
 from redis.exceptions import ConnectionError
 
-from lookyloo.helpers import get_homedir, get_socket_path
+from lookyloo.default import get_homedir, get_socket_path
 
 
 def check_running(name: str) -> bool:

bin/shutdown.py

@@ -3,7 +3,7 @@
 
 import time
 
-from lookyloo.abstractmanager import AbstractManager
+from lookyloo.default import AbstractManager
 
 
 def main():

bin/start.py

@@ -3,7 +3,7 @@
 
 from subprocess import Popen, run
 
-from lookyloo.helpers import get_config, get_homedir
+from lookyloo.default import get_config, get_homedir
 
 
 def main():

bin/start_website.py

@@ -4,8 +4,7 @@
 import logging
 from subprocess import Popen
 
-from lookyloo.abstractmanager import AbstractManager
-from lookyloo.helpers import get_config, get_homedir
+from lookyloo.default import get_config, get_homedir, AbstractManager
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO)

bin/stop.py

@@ -5,7 +5,7 @@ from subprocess import Popen
 
 from redis import Redis
 
-from lookyloo.helpers import get_homedir, get_socket_path
+from lookyloo.default import get_homedir, get_socket_path
 
 
 def main():

bin/update.py

@@ -10,7 +10,7 @@ import subprocess
 import sys
 from pathlib import Path
 
-from lookyloo.helpers import get_homedir
+from lookyloo.default import get_homedir
 
 logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s:%(message)s',
                     level=logging.INFO)

config/generic.json.sample

@@ -6,6 +6,7 @@
   "public_domain": "lookyloo.myorg.local",
   "website_listen_ip": "0.0.0.0",
   "website_listen_port": 5100,
+  "systemd_service_name": "lookyloo",
   "splash_url": "http://127.0.0.1:8050",
   "default_user_agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.138 Safari/537.36",
   "users": {},
@@ -52,6 +53,7 @@
     "public_domain": "Domain where the instance can be reached. Used for permalinks (e-mail, MISP export).",
     "website_listen_ip": "IP Flask will listen on. Defaults to 0.0.0.0, meaning all interfaces.",
     "website_listen_port": "Port Flask will listen on.",
+    "systemd_service_name": "(Optional) Name of the systemd service if your project has one.",
     "splash_url": "URL to connect to splash",
     "default_user_agent": "Ultimate fallback if the capture form, or the asynchronous submission, doesn't provide a user agent.",
     "users": "It is some kind of an admin accounts. Format: {username: password}",

lookyloo/capturecache.py

@@ -20,9 +20,8 @@ from redis import Redis
 
 from .context import Context
 from .indexing import Indexing
-from .exceptions import (LookylooException, MissingCaptureDirectory, NoValidHarFile,
-                         MissingUUID, TreeNeedsRebuild)
-from .helpers import try_make_file, get_config
+from .default import LookylooException, try_make_file, get_config
+from .exceptions import MissingCaptureDirectory, NoValidHarFile, MissingUUID, TreeNeedsRebuild
 
 
 class CaptureCache():

lookyloo/context.py

@@ -10,8 +10,8 @@ from urllib.parse import urlsplit
 from har2tree import CrawledTree, HostNode, URLNode
 from redis import Redis
 
-from .helpers import (get_config, get_homedir, get_resources_hashes,
-                      get_socket_path, load_known_content, serialize_to_json)
+from .default import get_config, get_homedir, get_socket_path
+from .helpers import get_resources_hashes, load_known_content, serialize_to_json
 from .modules import SaneJavaScript
 
 

lookyloo/default/__init__.py

@@ -0,0 +1,14 @@
+env_global_name: str = 'LOOKYLOO_HOME'
+
+from .exceptions import LookylooException  # noqa
+
+# NOTE: the imports below are there to avoid too long paths when importing the
+# classes/methods in the rest of the project while keeping all that in a subdirectory
+# and allow to update them easily.
+# You should not have to change anything in this file below this line.
+
+from .abstractmanager import AbstractManager  # noqa
+
+from .exceptions import MissingEnv, CreateDirectoryException, ConfigError  # noqa
+
+from .helpers import get_homedir, load_configs, get_config, safe_create_dir, get_socket_path, try_make_file  # noqa

lookyloo/default/abstractmanager.py

@@ -29,13 +29,20 @@ class AbstractManager(ABC):
 
     @staticmethod
     def is_running() -> List[Tuple[str, float]]:
-        r = Redis(unix_socket_path=get_socket_path('cache'), db=1, decode_responses=True)
-        return r.zrangebyscore('running', '-inf', '+inf', withscores=True)
+        try:
+            r = Redis(unix_socket_path=get_socket_path('cache'), db=1, decode_responses=True)
+            return r.zrangebyscore('running', '-inf', '+inf', withscores=True)
+        except ConnectionError:
+            print('Unable to connect to redis, the system is down.')
+            return []
 
     @staticmethod
     def force_shutdown():
-        r = Redis(unix_socket_path=get_socket_path('cache'), db=1, decode_responses=True)
-        r.set('shutdown', 1)
+        try:
+            r = Redis(unix_socket_path=get_socket_path('cache'), db=1, decode_responses=True)
+            r.set('shutdown', 1)
+        except ConnectionError:
+            print('Unable to connect to redis, the system is down.')
 
     def set_running(self) -> None:
         self.__redis.zincrby('running', 1, self.script_name)

lookyloo/default/exceptions.py

@@ -0,0 +1,18 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+
+class LookylooException(Exception):
+    pass
+
+
+class MissingEnv(LookylooException):
+    pass
+
+
+class CreateDirectoryException(LookylooException):
+    pass
+
+
+class ConfigError(LookylooException):
+    pass

lookyloo/default/helpers.py

@@ -0,0 +1,101 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+import json
+import logging
+import os
+from functools import lru_cache
+from pathlib import Path
+from typing import Any, Dict, Optional, Union
+
+from . import env_global_name
+from .exceptions import ConfigError, CreateDirectoryException, MissingEnv
+
+configs: Dict[str, Dict[str, Any]] = {}
+logger = logging.getLogger('Helpers')
+
+
+@lru_cache(64)
+def get_homedir() -> Path:
+    if not os.environ.get(env_global_name):
+        # Try to open a .env file in the home directory if it exists.
+        if (Path(__file__).resolve().parent.parent.parent / '.env').exists():
+            with (Path(__file__).resolve().parent.parent.parent / '.env').open() as f:
+                for line in f:
+                    key, value = line.strip().split('=', 1)
+                    if value[0] in ['"', "'"]:
+                        value = value[1:-1]
+                    os.environ[key] = value
+
+    if not os.environ.get(env_global_name):
+        guessed_home = Path(__file__).resolve().parent.parent.parent
+        raise MissingEnv(f"{env_global_name} is missing. \
+Run the following command (assuming you run the code from the clonned repository):\
+    export {env_global_name}='{guessed_home}'")
+    return Path(os.environ[env_global_name])
+
+
+@lru_cache(64)
+def load_configs(path_to_config_files: Optional[Union[str, Path]]=None):
+    global configs
+    if configs:
+        return
+    if path_to_config_files:
+        if isinstance(path_to_config_files, str):
+            config_path = Path(path_to_config_files)
+        else:
+            config_path = path_to_config_files
+    else:
+        config_path = get_homedir() / 'config'
+    if not config_path.exists():
+        raise ConfigError(f'Configuration directory {config_path} does not exists.')
+    elif not config_path.is_dir():
+        raise ConfigError(f'Configuration directory {config_path} is not a directory.')
+
+    configs = {}
+    for path in config_path.glob('*.json'):
+        with path.open() as _c:
+            configs[path.stem] = json.load(_c)
+
+
+@lru_cache(64)
+def get_config(config_type: str, entry: str, quiet: bool=False) -> Any:
+    """Get an entry from the given config_type file. Automatic fallback to the sample file"""
+    global configs
+    if not configs:
+        load_configs()
+    if config_type in configs:
+        if entry in configs[config_type]:
+            return configs[config_type][entry]
+        else:
+            if not quiet:
+                logger.warning(f'Unable to find {entry} in config file.')
+    else:
+        if not quiet:
+            logger.warning(f'No {config_type} config file available.')
+    if not quiet:
+        logger.warning(f'Falling back on sample config, please initialize the {config_type} config file.')
+    with (get_homedir() / 'config' / f'{config_type}.json.sample').open() as _c:
+        sample_config = json.load(_c)
+    return sample_config[entry]
+
+
+def safe_create_dir(to_create: Path) -> None:
+    if to_create.exists() and not to_create.is_dir():
+        raise CreateDirectoryException(f'The path {to_create} already exists and is not a directory')
+    to_create.mkdir(parents=True, exist_ok=True)
+
+
+def get_socket_path(name: str) -> str:
+    mapping = {
+        'cache': Path('cache', 'cache.sock'),
+        'indexing': Path('indexing', 'indexing.sock'),
+    }
+    return str(get_homedir() / mapping[name])
+
+
+def try_make_file(filename: Path):
+    try:
+        filename.touch(exist_ok=False)
+        return True
+    except FileExistsError:
+        return False

lookyloo/exceptions.py

@@ -1,27 +1,13 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
-
-class LookylooException(Exception):
-    pass
-
-
-class MissingEnv(LookylooException):
-    pass
+from .default import LookylooException
 
 
 class NoValidHarFile(LookylooException):
     pass
 
 
-class CreateDirectoryException(LookylooException):
-    pass
-
-
-class ConfigError(LookylooException):
-    pass
-
-
 class MissingUUID(LookylooException):
     pass
 

lookyloo/helpers.py

@@ -18,9 +18,8 @@ from publicsuffix2 import PublicSuffixList, fetch  # type: ignore
 from pytaxonomies import Taxonomies
 from requests.exceptions import HTTPError
 
-from .exceptions import ConfigError, CreateDirectoryException, MissingEnv
+from .default import get_homedir, safe_create_dir, get_config
 
-configs: Dict[str, Dict[str, Any]] = {}
 logger = logging.getLogger('Lookyloo - Helpers')
 
 
@@ -71,26 +70,6 @@ def get_public_suffix_list():
     return psl
 
 
-@lru_cache(64)
-def get_homedir() -> Path:
-    if not os.environ.get('LOOKYLOO_HOME'):
-        # Try to open a .env file in the home directory if it exists.
-        if (Path(__file__).resolve().parent.parent / '.env').exists():
-            with (Path(__file__).resolve().parent.parent / '.env').open() as f:
-                for line in f:
-                    key, value = line.strip().split('=', 1)
-                    if value[0] in ['"', "'"]:
-                        value = value[1:-1]
-                    os.environ[key] = value
-
-    if not os.environ.get('LOOKYLOO_HOME'):
-        guessed_home = Path(__file__).resolve().parent.parent
-        raise MissingEnv(f"LOOKYLOO_HOME is missing. \
-Run the following command (assuming you run the code from the clonned repository):\
-    export LOOKYLOO_HOME='{guessed_home}'")
-    return Path(os.environ['LOOKYLOO_HOME'])
-
-
 @lru_cache(64)
 def get_captures_dir() -> Path:
     capture_dir = get_homedir() / 'scraped'
@@ -104,66 +83,6 @@ def get_email_template() -> str:
         return f.read()
 
 
-@lru_cache(64)
-def load_configs(path_to_config_files: Optional[Union[str, Path]]=None):
-    global configs
-    if configs:
-        return
-    if path_to_config_files:
-        if isinstance(path_to_config_files, str):
-            config_path = Path(path_to_config_files)
-        else:
-            config_path = path_to_config_files
-    else:
-        config_path = get_homedir() / 'config'
-    if not config_path.exists():
-        raise ConfigError(f'Configuration directory {config_path} does not exists.')
-    elif not config_path.is_dir():
-        raise ConfigError(f'Configuration directory {config_path} is not a directory.')
-
-    configs = {}
-    for path in config_path.glob('*.json'):
-        with path.open() as _c:
-            configs[path.stem] = json.load(_c)
-
-
-@lru_cache(64)
-def get_config(config_type: str, entry: str, quiet: bool=False) -> Any:
-    """Get an entry from the given config_type file. Automatic fallback to the sample file"""
-    global configs
-    if not configs:
-        load_configs()
-    if config_type in configs:
-        if entry in configs[config_type]:
-            return configs[config_type][entry]
-        else:
-            if not quiet:
-                logger.warning(f'Unable to find {entry} in config file.')
-    else:
-        if not quiet:
-            logger.warning(f'No {config_type} config file available.')
-    if not quiet:
-        logger.warning(f'Falling back on sample config, please initialize the {config_type} config file.')
-    with (get_homedir() / 'config' / f'{config_type}.json.sample').open() as _c:
-        sample_config = json.load(_c)
-    return sample_config[entry]
-
-
-def safe_create_dir(to_create: Path) -> None:
-    if to_create.exists() and not to_create.is_dir():
-        raise CreateDirectoryException(f'The path {to_create} already exists and is not a directory')
-    to_create.mkdir(parents=True, exist_ok=True)
-
-
-def get_socket_path(name: str) -> str:
-    mapping = {
-        'cache': Path('cache', 'cache.sock'),
-        'indexing': Path('indexing', 'indexing.sock'),
-        'storage': Path('storage', 'storage.sock'),
-    }
-    return str(get_homedir() / mapping[name])
-
-
 def get_user_agents(directory: str='user_agents') -> Dict[str, Any]:
     ua_files_path = sorted((get_homedir() / directory).glob('**/*.json'), reverse=True)
     with ua_files_path[0].open() as f:
@@ -223,14 +142,6 @@ def uniq_domains(uniq_urls):
     return domains
 
 
-def try_make_file(filename: Path):
-    try:
-        filename.touch(exist_ok=False)
-        return True
-    except FileExistsError:
-        return False
-
-
 @lru_cache(64)
 def get_useragent_for_requests():
     version = pkg_resources.get_distribution('lookyloo').version

lookyloo/indexing.py

@@ -12,7 +12,8 @@ from har2tree import CrawledTree
 from redis import ConnectionPool, Redis
 from redis.connection import UnixDomainSocketConnection
 
-from .helpers import get_public_suffix_list, get_socket_path, get_config
+from .default import get_socket_path, get_config
+from .helpers import get_public_suffix_list
 
 
 class Indexing():

lookyloo/lookyloo.py

@@ -25,11 +25,11 @@ from werkzeug.useragents import UserAgent
 
 from .capturecache import CaptureCache, CapturesIndex
 from .context import Context
-from .exceptions import (LookylooException, MissingCaptureDirectory,
+from .default import LookylooException, get_homedir, get_config, get_socket_path
+from .exceptions import (MissingCaptureDirectory,
                          MissingUUID, TreeNeedsRebuild, NoValidHarFile)
-from .helpers import (CaptureStatus, get_captures_dir, get_config,
-                      get_email_template, get_homedir, get_resources_hashes,
-                      get_socket_path, get_splash_url, get_taxonomies, uniq_domains)
+from .helpers import (CaptureStatus, get_captures_dir, get_email_template,
+                      get_resources_hashes, get_splash_url, get_taxonomies, uniq_domains)
 from .indexing import Indexing
 from .modules import (MISP, PhishingInitiative, UniversalWhois,
                       UrlScan, VirusTotal, Phishtank)

lookyloo/modules/misp.py

@@ -11,7 +11,8 @@ from har2tree import HostNode, URLNode, Har2TreeError
 from pymisp import MISPAttribute, MISPEvent, PyMISP
 from pymisp.tools import FileObject, URLObject
 
-from ..helpers import get_config, get_homedir, get_public_suffix_list
+from ..default import get_config, get_homedir
+from ..helpers import get_public_suffix_list
 if TYPE_CHECKING:
     from ..capturecache import CaptureCache
 

lookyloo/modules/phishtank.py

@@ -10,8 +10,7 @@ from typing import Any, Dict, Optional, List
 from har2tree import CrawledTree
 from pyphishtanklookup import PhishtankLookup
 
-from ..exceptions import ConfigError
-from ..helpers import get_homedir
+from ..default import ConfigError, get_homedir
 
 # Note: stop doing requests 48 after the capture was intially done.
 

lookyloo/modules/pi.py

@@ -11,8 +11,7 @@ from typing import Any, Dict, Optional
 from har2tree import CrawledTree
 from pyeupi import PyEUPI
 
-from ..exceptions import ConfigError
-from ..helpers import get_homedir
+from ..default import ConfigError, get_homedir
 
 
 class PhishingInitiative():

lookyloo/modules/sanejs.py

@@ -8,7 +8,7 @@ from typing import Any, Dict, Iterable, List, Union
 
 from pysanejs import SaneJS
 
-from ..helpers import get_config, get_homedir
+from ..default import get_config, get_homedir
 
 
 class SaneJavaScript():

lookyloo/modules/urlscan.py

@@ -10,8 +10,8 @@ from typing import Any, Dict
 
 import requests
 
-from ..exceptions import ConfigError
-from ..helpers import get_config, get_homedir, get_useragent_for_requests
+from ..default import ConfigError, get_config, get_homedir
+from ..helpers import get_useragent_for_requests
 
 
 class UrlScan():

lookyloo/modules/uwhois.py

@@ -7,7 +7,7 @@ from typing import Any, Dict
 
 from har2tree import CrawledTree, Har2TreeError, HostNode
 
-from ..helpers import get_config
+from ..default import get_config
 
 
 class UniversalWhois():

lookyloo/modules/vt.py

@@ -12,8 +12,7 @@ import vt  # type: ignore
 from har2tree import CrawledTree
 from vt.error import APIError  # type: ignore
 
-from ..exceptions import ConfigError
-from ..helpers import get_homedir
+from ..default import ConfigError, get_homedir
 
 
 class VirusTotal():

tools/3rdparty.py

@@ -3,7 +3,7 @@
 
 import requests
 
-from lookyloo.helpers import get_homedir
+from lookyloo.default import get_homedir
 
 d3js_version = '7'
 datatables_version = "1.11.3"

tools/change_captures_dir.py

@@ -6,7 +6,8 @@ from pathlib import Path
 
 from redis import Redis
 
-from lookyloo.helpers import get_captures_dir, safe_create_dir, get_socket_path
+from lookyloo.default import safe_create_dir, get_socket_path
+from lookyloo.helpers import get_captures_dir
 
 
 def rename_captures():

tools/generate_sri.py

@@ -7,7 +7,7 @@ import json
 
 from typing import Dict
 
-from lookyloo.helpers import get_homedir
+from lookyloo.default import get_homedir
 
 if __name__ == '__main__':
     dest_dir = get_homedir() / 'website' / 'web'

tools/manual_parse_ua_list.py

@@ -14,7 +14,7 @@ try:
 except ImportError:
     HAS_CF = False
 
-from lookyloo.helpers import get_homedir, safe_create_dir
+from lookyloo.default import get_homedir, safe_create_dir
 
 
 def update_user_agents() -> None:
@@ -63,6 +63,7 @@ def ua_parser(html_content: str) -> Dict[str, Any]:
         to_store['by_frequency'].append({'os': os, 'browser': browser, 'useragent': ua['useragent']})
     return to_store
 
+
 def main():
     to_parse = Path('Most Common User Agents - Tech Blog (wh).html')
 
@@ -77,5 +78,6 @@ def main():
     with open(ua_file_name, 'w') as f:
         json.dump(to_store, f, indent=2)
 
+
 if __name__ == '__main__':
     main()

tools/monitoring.py

@@ -11,8 +11,8 @@ from redis.exceptions import ConnectionError
 from rich.console import Console
 from rich.padding import Padding
 
-from lookyloo.helpers import get_socket_path, splash_status
-from lookyloo.abstractmanager import AbstractManager
+from lookyloo.default import get_socket_path, AbstractManager
+from lookyloo.helpers import splash_status
 
 # NOTE: run with watch:
 #   watch --color tools/monitoring.py

tools/validate_config_files.py

@@ -5,7 +5,7 @@ import json
 import logging
 import argparse
 
-from lookyloo.helpers import get_homedir
+from lookyloo.default import get_homedir
 
 
 def validate_generic_config_file():

website/web/__init__.py

@@ -21,8 +21,9 @@ from flask_restx import Api  # type: ignore
 from pymisp import MISPEvent, MISPServerError
 from werkzeug.security import check_password_hash
 
+from lookyloo.default import get_config
 from lookyloo.exceptions import MissingUUID, NoValidHarFile
-from lookyloo.helpers import (CaptureStatus, get_config, get_taxonomies,
+from lookyloo.helpers import (CaptureStatus, get_taxonomies,
                               get_user_agents, load_cookies, splash_status)
 from lookyloo.lookyloo import Indexing, Lookyloo
 

website/web/helpers.py

@@ -11,7 +11,7 @@ from typing import Dict, List, Union
 import flask_login  # type: ignore
 from werkzeug.security import generate_password_hash
 
-from lookyloo.helpers import get_config, get_homedir
+from lookyloo.default import get_config, get_homedir
 
 
 def src_request_ip(request) -> str: