total rebaseHEADmaster

5 files changed, 88 insertions, 42 deletions

diff --git a/priv/scrubbers/default.ex b/priv/scrubbers/default.ex
index e10e3ec..a75a646 100755..100644
--- a/priv/scrubbers/default.ex
+++ b/priv/scrubbers/default.ex
@@ -33,35 +33,56 @@ defmodule Pleroma.HTML.Scrubber.Default do
     "ugc"
   ])
 
-  Meta.allow_tag_with_these_attributes(:a, ["name", "title"])
-
-  Meta.allow_tag_with_these_attributes(:abbr, ["title"])
-
-  Meta.allow_tag_with_these_attributes(:b, [])
-  Meta.allow_tag_with_these_attributes(:blockquote, [])
-  Meta.allow_tag_with_these_attributes(:br, [])
-  Meta.allow_tag_with_these_attributes(:code, [])
-  Meta.allow_tag_with_these_attributes(:del, [])
-  Meta.allow_tag_with_these_attributes(:em, [])
-  Meta.allow_tag_with_these_attributes(:hr, [])
-  Meta.allow_tag_with_these_attributes(:i, [])
-  Meta.allow_tag_with_these_attributes(:li, [])
-  Meta.allow_tag_with_these_attributes(:ol, [])
-  Meta.allow_tag_with_these_attributes(:p, [])
-  Meta.allow_tag_with_these_attributes(:pre, [])
-  Meta.allow_tag_with_these_attributes(:strong, [])
-  Meta.allow_tag_with_these_attributes(:sub, [])
-  Meta.allow_tag_with_these_attributes(:sup, [])
-  Meta.allow_tag_with_these_attributes(:ruby, [])
-  Meta.allow_tag_with_these_attributes(:rb, [])
-  Meta.allow_tag_with_these_attributes(:rp, [])
-  Meta.allow_tag_with_these_attributes(:rt, [])
-  Meta.allow_tag_with_these_attributes(:rtc, [])
-  Meta.allow_tag_with_these_attributes(:u, [])
-  Meta.allow_tag_with_these_attributes(:ul, [])
-
-  Meta.allow_tag_with_this_attribute_values(:span, "class", ["h-card", "recipients-inline"])
-  Meta.allow_tag_with_these_attributes(:span, [])
+  Meta.allow_tag_with_these_attributes(:a, ["name", "title", "lang"])
+
+  Meta.allow_tag_with_these_attributes(:abbr, ["title", "lang"])
+  Meta.allow_tag_with_these_attributes(:acronym, ["title", "lang"])
+
+  # sort(1)-ed list
+  Meta.allow_tag_with_these_attributes(:bdi, [])
+  Meta.allow_tag_with_these_attributes(:bdo, ["dir"])
+  Meta.allow_tag_with_these_attributes(:big, ["lang"])
+  Meta.allow_tag_with_these_attributes(:b, ["lang"])
+  Meta.allow_tag_with_these_attributes(:blockquote, ["lang"])
+  Meta.allow_tag_with_these_attributes(:br, ["lang"])
+  Meta.allow_tag_with_these_attributes(:cite, ["lang"])
+  Meta.allow_tag_with_these_attributes(:code, ["lang"])
+  Meta.allow_tag_with_these_attributes(:del, ["lang"])
+  Meta.allow_tag_with_these_attributes(:dfn, ["lang"])
+  Meta.allow_tag_with_these_attributes(:em, ["lang"])
+  Meta.allow_tag_with_these_attributes(:hr, ["lang"])
+  Meta.allow_tag_with_these_attributes(:i, ["lang"])
+  Meta.allow_tag_with_these_attributes(:ins, ["lang"])
+  Meta.allow_tag_with_these_attributes(:kbd, ["lang"])
+  Meta.allow_tag_with_these_attributes(:li, ["lang"])
+  Meta.allow_tag_with_these_attributes(:ol, ["lang"])
+  Meta.allow_tag_with_these_attributes(:p, ["lang"])
+  Meta.allow_tag_with_these_attributes(:pre, ["lang"])
+  Meta.allow_tag_with_these_attributes(:q, ["lang"])
+  Meta.allow_tag_with_these_attributes(:rb, ["lang"])
+  Meta.allow_tag_with_these_attributes(:rp, ["lang"])
+  Meta.allow_tag_with_these_attributes(:rtc, ["lang"])
+  Meta.allow_tag_with_these_attributes(:rt, ["lang"])
+  Meta.allow_tag_with_these_attributes(:ruby, ["lang"])
+  Meta.allow_tag_with_these_attributes(:samp, ["lang"])
+  Meta.allow_tag_with_these_attributes(:s, ["lang"])
+  Meta.allow_tag_with_these_attributes(:small, ["lang"])
+  Meta.allow_tag_with_these_attributes(:strong, ["lang"])
+  Meta.allow_tag_with_these_attributes(:sub, ["lang"])
+  Meta.allow_tag_with_these_attributes(:sup, ["lang"])
+  Meta.allow_tag_with_these_attributes(:tt, ["lang"])
+  Meta.allow_tag_with_these_attributes(:u, ["lang"])
+  Meta.allow_tag_with_these_attributes(:ul, ["lang"])
+  Meta.allow_tag_with_these_attributes(:var, ["lang"])
+  Meta.allow_tag_with_these_attributes(:wbr, ["lang"])
+
+  Meta.allow_tag_with_this_attribute_values(:span, "class", [
+    "h-card",
+    "recipients-inline",
+    "quote-inline"
+  ])
+
+  Meta.allow_tag_with_these_attributes(:span, ["lang"])
 
   Meta.allow_tag_with_this_attribute_values(:code, "class", ["inline"])
 
@@ -77,29 +98,30 @@ defmodule Pleroma.HTML.Scrubber.Default do
       "width",
       "height",
       "title",
-      "alt"
+      "alt",
+      "lang"
     ])
   end
 
   if Pleroma.Config.get([:markup, :allow_tables]) do
-    Meta.allow_tag_with_these_attributes(:table, [])
-    Meta.allow_tag_with_these_attributes(:tbody, [])
-    Meta.allow_tag_with_these_attributes(:td, [])
-    Meta.allow_tag_with_these_attributes(:th, [])
-    Meta.allow_tag_with_these_attributes(:thead, [])
-    Meta.allow_tag_with_these_attributes(:tr, [])
+    Meta.allow_tag_with_these_attributes(:table, ["lang"])
+    Meta.allow_tag_with_these_attributes(:tbody, ["lang"])
+    Meta.allow_tag_with_these_attributes(:td, ["lang"])
+    Meta.allow_tag_with_these_attributes(:th, ["lang"])
+    Meta.allow_tag_with_these_attributes(:thead, ["lang"])
+    Meta.allow_tag_with_these_attributes(:tr, ["lang"])
   end
 
   if Pleroma.Config.get([:markup, :allow_headings]) do
-    Meta.allow_tag_with_these_attributes(:h1, [])
-    Meta.allow_tag_with_these_attributes(:h2, [])
-    Meta.allow_tag_with_these_attributes(:h3, [])
-    Meta.allow_tag_with_these_attributes(:h4, [])
-    Meta.allow_tag_with_these_attributes(:h5, [])
+    Meta.allow_tag_with_these_attributes(:h1, ["lang"])
+    Meta.allow_tag_with_these_attributes(:h2, ["lang"])
+    Meta.allow_tag_with_these_attributes(:h3, ["lang"])
+    Meta.allow_tag_with_these_attributes(:h4, ["lang"])
+    Meta.allow_tag_with_these_attributes(:h5, ["lang"])
   end
 
   if Pleroma.Config.get([:markup, :allow_fonts]) do
-    Meta.allow_tag_with_these_attributes(:font, ["face"])
+    Meta.allow_tag_with_these_attributes(:font, ["face", "lang"])
   end
 
   Meta.strip_everything_not_covered()
diff --git a/priv/scrubbers/links_only.ex b/priv/scrubbers/links_only.ex
index 7f434fb..7f434fb 100755..100644
--- a/priv/scrubbers/links_only.ex
+++ b/priv/scrubbers/links_only.ex
diff --git a/priv/scrubbers/media_proxy.ex b/priv/scrubbers/media_proxy.ex
index 9935b38..9935b38 100755..100644
--- a/priv/scrubbers/media_proxy.ex
+++ b/priv/scrubbers/media_proxy.ex
diff --git a/priv/scrubbers/search_indexing.ex b/priv/scrubbers/search_indexing.ex
new file mode 100644
index 0000000..02756ab
--- /dev/null
+++ b/priv/scrubbers/search_indexing.ex
@@ -0,0 +1,24 @@
+# Pleroma: A lightweight social networking server
+# Copyright © 2017-2022 Pleroma Authors <https://pleroma.social/>
+# SPDX-License-Identifier: AGPL-3.0-only
+
+defmodule Pleroma.HTML.Scrubber.SearchIndexing do
+  @moduledoc """
+  An HTML scrubbing policy that scrubs things for searching.
+  """
+
+  require FastSanitize.Sanitizer.Meta
+  alias FastSanitize.Sanitizer.Meta
+
+  # Explicitly remove mentions
+  def scrub({:a, attrs, children}) do
+    if(Enum.any?(attrs, fn {att, val} -> att == "class" and String.contains?(val, "mention") end),
+      do: nil,
+      # Strip the tag itself, leave only children (text, presumably)
+      else: children
+    )
+  end
+
+  Meta.strip_comments()
+  Meta.strip_everything_not_covered()
+end
diff --git a/priv/scrubbers/twitter_text.ex b/priv/scrubbers/twitter_text.ex
index 6e23b3e..6e23b3e 100755..100644
--- a/priv/scrubbers/twitter_text.ex
+++ b/priv/scrubbers/twitter_text.ex