Implement blocklists for MediaProxy

CHANGELOG.md

@@ -4,6 +4,12 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 
+## Unreleased
+
+## Added
+- Added a new configuration option to the MediaProxy feature that allows the blocking of specific domains from using the media proxy or being explicitly allowed by the Content-Security-Policy.
+  - Please make sure instances you wanted to block media from are not in the MediaProxy `whitelist`, and instead use `blocklist`.
+
 ## 2023.05
 
 ## Added

config/config.exs

@@ -443,7 +443,8 @@
     # Note: max_read_duration defaults to Pleroma.ReverseProxy.max_read_duration_default/1
     max_read_duration: 30_000
   ],
-  whitelist: []
+  whitelist: [],
+  blocklist: []
 
 config :pleroma, Pleroma.Web.MediaProxy.Invalidation.Http,
   method: :purge,

config/description.exs

@@ -1558,7 +1558,21 @@
       %{
         key: :whitelist,
         type: {:list, :string},
-        description: "List of hosts with scheme to bypass the MediaProxy",
+        description: """
+        List of hosts with scheme to bypass the MediaProxy.\n
+        The media will be fetched by the client, directly from the remote server.\n
+        To allow this, it will Content-Security-Policy exceptions for each instance listed.\n
+        This is to be used for instances you trust and do not want to cache media for.
+        """,
+        suggestions: ["http://example.com"]
+      },
+      %{
+        key: :blocklist,
+        type: {:list, :string},
+        description: """
+        List of hosts with scheme which will not go through the MediaProxy, and will not be explicitly allowed by the Content-Security-Policy.
+        This is to be used for instances where you do not want their media to go through your server or to be accessed by clients.
+        """,
         suggestions: ["http://example.com"]
       }
     ]

lib/pleroma/web/media_proxy.ex

@@ -52,7 +52,7 @@ def url(url) do
 
   @spec url_proxiable?(String.t()) :: boolean()
   def url_proxiable?(url) do
-    not local?(url) and not whitelisted?(url)
+    not local?(url) and not whitelisted?(url) and not blocked?(url)
   end
 
   def preview_url(url, preview_params \\ []) do
@@ -83,6 +83,11 @@ def whitelisted?(url) do
     domain in mediaproxy_whitelist_domains
   end
 
+  def blocked?(url) do
+    %{host: domain} = URI.parse(url)
+    domain in Config.get([:media_proxy, :whitelist])
+  end
+
   defp maybe_get_domain_from_url("http" <> _ = url) do
     URI.parse(url).host
   end

test/pleroma/web/media_proxy_test.exs

@@ -199,6 +199,15 @@ test "mediaproxy whitelist" do
       assert unencoded == url
     end
 
+    test "mediaproxy blocklist" do
+      clear_config([:media_proxy, :whitelist], ["https://google.com"])
+      clear_config([:media_proxy, :blocklist], ["https://feld.me"])
+      url = "https://feld.me/foo.png"
+
+      unencoded = MediaProxy.url(url)
+      assert unencoded == url
+    end
+
     # TODO: delete after removing support bare domains for media proxy whitelist
     test "mediaproxy whitelist bare domains whitelist (deprecated)" do
       clear_config([:media_proxy, :whitelist], ["google.com", "feld.me"])
@@ -220,6 +229,18 @@ test "does not change whitelisted urls" do
       assert String.starts_with?(encoded, media_url)
     end
 
+    test "does not change blocked urls" do
+      clear_config([:media_proxy, :whitelist], ["mycdn.akamai.com"])
+      clear_config([:media_proxy, :base_url], "https://cache.pleroma.social")
+
+      media_url = "https://mycdn.akamai.com"
+
+      url = "#{media_url}/static/logo.png"
+      encoded = MediaProxy.url(url)
+
+      assert String.starts_with?(encoded, media_url)
+    end
+
     test "ensure Pleroma.Upload base_url is always whitelisted" do
       media_url = "https://media.pleroma.social"
       clear_config([Pleroma.Upload, :base_url], media_url)

test/pleroma/web/plugs/http_security_plug_test.exs

@@ -128,6 +128,12 @@ test "with media_proxy bare domains whitelist (deprecated)", %{conn: conn} do
       clear_config([:media_proxy, :whitelist], ["example4.com", "example5.com"])
       assert_media_img_src(conn, "example5.com example4.com")
     end
+
+    test "with media_proxy blocklist", %{conn: conn} do
+      clear_config([:media_proxy, :whitelist], ["https://example6.com", "https://example7.com"])
+      clear_config([:media_proxy, :blocklist], ["https://example8.com"])
+      assert_media_img_src(conn, "https://example7.com https://example6.com")
+    end
   end
 
   defp assert_media_img_src(conn, url) do