pleroma

commit: 5a39866388c411f2bcee9848352f8c420513f34f
parent 6256822afd368e5f6b410d47c5ff9b584e50a461
Author: Ekaterina Vaartis <vaartis@kotobank.ch>
Date:   Sat, 27 Aug 2022 01:43:59 +0300

Specifically strip mentions for search indexing

Diffstat:
M
lib/mix/tasks/pleroma/search/meilisearch.ex
1
+
M
lib/pleroma/search/meilisearch.ex
3
++-
A
priv/scrubbers/search_indexing.ex
24
++++++++++++++++++++++++

3 files changed, 27 insertions(+), 1 deletion(-)
diff --git a/lib/mix/tasks/pleroma/search/meilisearch.ex b/lib/mix/tasks/pleroma/search/meilisearch.ex
@@ -13,6 +13,7 @@ defmodule Mix.Tasks.Pleroma.Search.Meilisearch do
 
   def run(["index"]) do
     start_pleroma()
+    Pleroma.HTML.compile_scrubbers()
 
     meili_version =
       (
diff --git a/lib/pleroma/search/meilisearch.ex b/lib/pleroma/search/meilisearch.ex
@@ -122,7 +122,8 @@ defmodule Pleroma.Search.Meilisearch do
         end
 
       content =
-        with {:ok, scrubbed} <- FastSanitize.strip_tags(content_str),
+        with {:ok, scrubbed} <-
+               FastSanitize.Sanitizer.scrub(content_str, Pleroma.HTML.Scrubber.SearchIndexing),
              trimmed <- String.trim(scrubbed) do
           trimmed
         end
diff --git a/priv/scrubbers/search_indexing.ex 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