From 5baab519514286fbdb4e28bbaf262aa9577b8b14 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rapha=C3=ABl=20Vinot?= <raphael@vinot.info>
Date: Sun, 24 May 2020 15:35:30 +0200
Subject: [PATCH] fix: Proper fallback for the user agent.

---
 config/generic.json.sample | 6 ++++--
 lookyloo/lookyloo.py       | 7 ++++++-
 lookyloo/modules.py        | 1 +
 poetry.lock                | 2 +-
 4 files changed, 12 insertions(+), 4 deletions(-)

diff --git a/config/generic.json.sample b/config/generic.json.sample
index cb3b215c..0bab9024 100644
--- a/config/generic.json.sample
+++ b/config/generic.json.sample
@@ -3,10 +3,11 @@
   "splash_loglevel": "WARNING",
   "only_global_lookups": true,
   "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",
   "cache_clean_user": {},
   "time_delta_on_index": {
-    "weeks": 0,
-    "days": 1,
+    "weeks": 1,
+    "days": 0,
     "hours": 0
   },
   "enable_mail_notification": false,
@@ -23,6 +24,7 @@
     "splash_loglevel": "(Splash) INFO is *very* verbose.",
     "only_global_lookups": "Set it to True if your instance is publicly available so users aren't able to scan your internal network",
     "splash_url": "URL to connect to splash",
+    "default_user_agent": "Ultimate fallback if the capture form, or the asynchronous submission, don't provide a UA",
     "cache_clean_user": "Format: {username: password}",
     "time_delta_on_index": "Time interval of the capture displayed on the index",
     "enable_mail_notification": "Enable email notification or not",
diff --git a/lookyloo/lookyloo.py b/lookyloo/lookyloo.py
index d7eb7f52..d03a47b1 100644
--- a/lookyloo/lookyloo.py
+++ b/lookyloo/lookyloo.py
@@ -342,7 +342,12 @@ class Lookyloo():
                 return False
 
         cookies = load_cookies(cookies_pseudofile)
-        items = crawl(self.splash_url, url, cookies=cookies, depth=depth, user_agent=user_agent,
+        if not user_agent:
+            # Catch case where the UA is broken on the UI, and the async submission.
+            ua: str = self.get_config('default_user_agent')  # type: ignore
+        else:
+            ua = user_agent
+        items = crawl(self.splash_url, url, cookies=cookies, depth=depth, user_agent=ua,
                       log_enabled=True, log_level=self.get_config('splash_loglevel'))
         if not items:
             # broken
diff --git a/lookyloo/modules.py b/lookyloo/modules.py
index 68773d50..26d77457 100644
--- a/lookyloo/modules.py
+++ b/lookyloo/modules.py
@@ -48,6 +48,7 @@ class SaneJavaScript():
         "69e2da5cdc318fc237eaa243b6ea7ecc83b68dbdea8478dc69154abdda86ecb4e16c35891cc1facb3ce7e0cf19d5abf189c50f59c769777706f4558f6442abbc": "This is a 1*1 pixel GIF",
         "16dd1560fdd43c3eee7bcf622d940be93e7e74dee90286da37992d69cea844130911b97f41c71f8287b54f00bd3a388191112f490470cf27c374d524f49ba516": "This is a 1*1 pixel GIF",
         "01211111688dc2007519ff56603fbe345d057337b911c829aaee97b8d02e7d885e7a2c2d51730f54a04aebc1821897c8041f15e216f1c973ed313087fa91a3fb": "This is a 1*1 pixel GIF",
+        "71db01662075fac031dea18b2c766826c77dbab01400a8642cdc7059394841d5df9020076554c3beca6f808187d42e1a1acc98fad9a0e1ad32ae869145f53746": "This is a 1*1 pixel GIF",
         # "": "This is a 1*1 pixel GIF",
         "f1c33e72643ce366fd578e3b5d393799e8c9ea27b180987826af43b4fc00b65a4eaae5e6426a23448956fee99e3108c6a86f32fb4896c156e24af0571a11c498": "This is a 1*1 pixel PNG",
         "dc7c40381b3d22919e32c1b700ccb77b1b0aea2690642d01c1ac802561e135c01d5a4d2a0ea18efc0ec3362e8c549814a10a23563f1f56bd62aee0ced7e2bd99": "This is a 1*1 pixel PNG",
diff --git a/poetry.lock b/poetry.lock
index 8715a1a1..6e0d2692 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -870,7 +870,7 @@ scrapy = "^1.8.0"
 scrapy-splash = "^0.7.2"
 
 [package.source]
-reference = "9be949672aa7d90719e4d97c7b9aa70f106072bf"
+reference = "d0ea9acbd769c73a15c2142633acef2870d70e08"
 type = "git"
 url = "https://github.com/viper-framework/ScrapySplashWrapper.git"
 [[package]]