From 5af594bf0f620580bcc69e43b056871b0d908c2d Mon Sep 17 00:00:00 2001
From: syeopite <syeopite@syeopite.dev>
Date: Sun, 27 Jun 2021 15:01:22 -0700
Subject: [PATCH] Create and use random user agents on each request

---
 src/invidious.cr                            |  1 +
 src/invidious/helpers/random-user-agents.cr | 46 +++++++++++++++++++++
 src/invidious/helpers/utils.cr              |  2 +-
 3 files changed, 48 insertions(+), 1 deletion(-)
 create mode 100644 src/invidious/helpers/random-user-agents.cr

diff --git a/src/invidious.cr b/src/invidious.cr
index 89292f05..abe3cf15 100644
--- a/src/invidious.cr
+++ b/src/invidious.cr
@@ -66,6 +66,7 @@ SOFTWARE = {
   "branch"  => "#{CURRENT_BRANCH}",
 }
 
+UA_LIST = prepare_random_user_agents()
 YT_POOL = YoutubeConnectionPool.new(YT_URL, capacity: CONFIG.pool_size, timeout: 2.0, use_quic: CONFIG.use_quic)
 
 # CLI
diff --git a/src/invidious/helpers/random-user-agents.cr b/src/invidious/helpers/random-user-agents.cr
new file mode 100644
index 00000000..f16689fb
--- /dev/null
+++ b/src/invidious/helpers/random-user-agents.cr
@@ -0,0 +1,46 @@
+# Using different browsers would allow us to disguise our traffic even more.
+# However, due to the widely different version and operating system values as well as
+# their frequency on different OSes we're going to need a separate dataset
+# for everything.
+
+AGENT_COMPONENTS = {
+  "safari" => {
+    "os"       => ["Macintosh; Intel Mac OS X 10_15_7", "Macintosh; Intel Mac OS X 10_15_6"],
+    "versions" => ["14.1.1", "14.1", "14.0.3"],
+  },
+
+  "firefox" => {
+    "os" => ["Macintosh; Intel Mac OS X 10.15'", "Macintosh; Intel Mac OS X 10.14",
+             "Windows NT 10.0; Win64; x64", "X11; Ubuntu; Linux x86_64",
+             "X11; Linux x86_64"],
+    "versions" => ["88.0 ", "89.0"],
+  },
+}
+
+private def generate_user_agent
+  agent_component = AGENT_COMPONENTS.keys.sample(1)[0]
+  os = AGENT_COMPONENTS[agent_component]["os"].sample(1)[0]
+  version = AGENT_COMPONENTS[agent_component]["versions"].sample(1)[0]
+  base = "Mozilla/5.0 "
+
+  case agent_component
+  when "safari"
+    base += "(#{os}) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/#{version} Safari/605.1.15"
+  when "firefox"
+    base += "(#{os}; rv:#{version})) Gecko/20100101 Firefox/#{version}"
+  end
+
+  return base
+end
+
+# Prepare 10 user agents to randomly choose from in
+# order to avoid detection
+def prepare_random_user_agents
+  ua_list = [] of String
+  10.times { ua_list << generate_user_agent() }
+
+  # Conserve memory and deallocate AGENT_COMPONENTS
+  AGENT_COMPONENTS.clear
+
+  return ua_list
+end
diff --git a/src/invidious/helpers/utils.cr b/src/invidious/helpers/utils.cr
index 6ee07d7a..d26600c8 100644
--- a/src/invidious/helpers/utils.cr
+++ b/src/invidious/helpers/utils.cr
@@ -2,7 +2,7 @@ require "lsquic"
 require "db"
 
 def add_yt_headers(request)
-  request.headers["user-agent"] ||= "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.97 Safari/537.36"
+  request.headers["user-agent"] ||= UA_LIST.sample(1)[0]
   request.headers["accept-charset"] ||= "ISO-8859-1,utf-8;q=0.7,*;q=0.7"
   request.headers["accept"] ||= "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"
   request.headers["accept-language"] ||= "en-us,en;q=0.5"