Add support for Anti-Captcha

src/invidious.cr

@@ -212,6 +212,19 @@ spawn do
   end
 end
 
+if CONFIG.captcha_key
+  spawn do
+    bypass_captcha(CONFIG.captcha_key, logger) do |cookies|
+      cookies.each do |cookie|
+        config.cookies << cookie
+      end
+
+      # Persist cookies between runs
+      File.write("config/config.yml", config.to_yaml)
+    end
+  end
+end
+
 connection_channel = Channel({Bool, Channel(PQ::Notification)}).new(32)
 spawn do
   connections = [] of Channel(PQ::Notification)

src/invidious/helpers/helpers.cr

@@ -192,6 +192,27 @@ struct Config
     end
   end
 
+  module StringToCookies
+    def self.to_yaml(value : HTTP::Cookies, yaml : YAML::Nodes::Builder)
+      (value.map { |c| "#{c.name}=#{c.value}" }).join("; ").to_yaml(yaml)
+    end
+
+    def self.from_yaml(ctx : YAML::ParseContext, node : YAML::Nodes::Node) : HTTP::Cookies
+      unless node.is_a?(YAML::Nodes::Scalar)
+        node.raise "Expected scalar, not #{node.class}"
+      end
+
+      cookies = HTTP::Cookies.new
+      node.value.split(";").each do |cookie|
+        next if cookie.strip.empty?
+        name, value = cookie.split("=", 2)
+        cookies << HTTP::Cookie.new(name.strip, value.strip)
+      end
+
+      cookies
+    end
+  end
+
   def disabled?(option)
     case disabled = CONFIG.disable_proxy
     when Bool
@@ -236,6 +257,8 @@ struct Config
     host_binding:      {type: String, default: "0.0.0.0"},                                                  # Host to bind (overrided by command line argument)
     pool_size:         {type: Int32, default: 100},                                                         # Pool size for HTTP requests to youtube.com and ytimg.com (each domain has a separate pool of `pool_size`)
     admin_email:       {type: String, default: "omarroth@protonmail.com"},                                  # Email for bug reports
+    cookies:           {type: HTTP::Cookies, default: HTTP::Cookies.new, converter: StringToCookies},       # Saved cookies in "name1=value1; name2=value2..." format
+    captcha_key:       {type: String?, default: nil},                                                       # Key for Anti-Captcha
   })
 end
 

src/invidious/helpers/jobs.cr

@@ -228,11 +228,78 @@ def update_decrypt_function
       yield decrypt_function
     rescue ex
       next
-    end
-
+    ensure
       sleep 1.minute
       Fiber.yield
     end
+  end
+end
+
+def bypass_captcha(captcha_key, logger)
+  loop do
+    begin
+      response = YT_POOL.client &.get("/watch?v=CvFH_6DNRCY&gl=US&hl=en&disable_polymer=1&has_verified=1&bpctr=9999999999")
+      if response.body.includes?("To continue with your YouTube experience, please fill out the form below.")
+        html = XML.parse_html(response.body)
+        form = html.xpath_node(%(//form[@action="/das_captcha"])).not_nil!
+        site_key = form.xpath_node(%(.//div[@class="g-recaptcha"])).try &.["data-sitekey"]
+
+        inputs = {} of String => String
+        form.xpath_nodes(%(.//input[@name])).map do |node|
+          inputs[node["name"]] = node["value"]
+        end
+
+        headers = response.cookies.add_request_headers(HTTP::Headers.new)
+
+        response = JSON.parse(HTTP::Client.post("https://api.anti-captcha.com/createTask", body: {
+          "clientKey" => CONFIG.captcha_key,
+          "task"      => {
+            "type" => "NoCaptchaTaskProxyless",
+            # "type"          => "NoCaptchaTask",
+            "websiteURL" => "https://www.youtube.com/watch?v=CvFH_6DNRCY&gl=US&hl=en&disable_polymer=1&has_verified=1&bpctr=9999999999",
+            "websiteKey" => site_key,
+            # "proxyType"     => "http",
+            # "proxyAddress"  => CONFIG.proxy_address,
+            # "proxyPort"     => CONFIG.proxy_port,
+            # "proxyLogin"    => CONFIG.proxy_user,
+            # "proxyPassword" => CONFIG.proxy_pass,
+            # "userAgent"     => random_user_agent,
+          },
+        }.to_json).body)
+
+        if response["error"]?
+          raise response["error"].as_s
+        end
+
+        task_id = response["taskId"].as_i
+
+        loop do
+          sleep 10.seconds
+
+          response = JSON.parse(HTTP::Client.post("https://api.anti-captcha.com/getTaskResult", body: {
+            "clientKey" => CONFIG.captcha_key,
+            "taskId"    => task_id,
+          }.to_json).body)
+
+          if response["status"]?.try &.== "ready"
+            break
+          elsif response["errorId"]?.try &.as_i != 0
+            raise response["errorDescription"].as_s
+          end
+        end
+
+        inputs["g-recaptcha-response"] = response["solution"]["gRecaptchaResponse"].as_s
+        response = YT_POOL.client &.post("/das_captcha", headers, form: inputs)
+
+        yield response.cookies.select { |cookie| cookie.name != "PREF" }
+      end
+    rescue ex
+      logger.puts("Exception: #{ex.message}")
+    ensure
+      sleep 1.minute
+      Fiber.yield
+    end
+  end
 end
 
 def find_working_proxies(regions)

src/invidious/helpers/proxy.cr

@@ -99,6 +99,7 @@ class HTTPClient < HTTP::Client
       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"
+      request.headers["cookie"] = "#{(CONFIG.cookies.map { |c| "#{c.name}=#{c.value}" }).join("; ")}; #{request.headers["cookie"]?}"
     end
 
     super