X-Git-Url: http://git.squeep.com/?a=blobdiff_plain;f=lib%2Fpleroma%2Fhtml.ex;h=f5c6e5033fb572b0f02e717101dceafbd3c394c9;hb=4a278cd80a02fa1882db17397bb97b155c76570e;hp=0ec73a91d39c3763e34f8cb98c8af7203b22ec3a;hpb=255f46d7ab124d86a71e994deffca5f4f438b49b;p=akkoma

diff --git a/lib/pleroma/html.ex b/lib/pleroma/html.ex
index 0ec73a91d..f5c6e5033 100644
--- a/lib/pleroma/html.ex
+++ b/lib/pleroma/html.ex
@@ -1,14 +1,217 @@
+# Pleroma: A lightweight social networking server
+# Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
+# SPDX-License-Identifier: AGPL-3.0-only
+
 defmodule Pleroma.HTML do
   alias HtmlSanitizeEx.Scrubber
 
+  defp get_scrubbers(scrubber) when is_atom(scrubber), do: [scrubber]
+  defp get_scrubbers(scrubbers) when is_list(scrubbers), do: scrubbers
+  defp get_scrubbers(_), do: [Pleroma.HTML.Scrubber.Default]
+
+  def get_scrubbers() do
+    Pleroma.Config.get([:markup, :scrub_policy])
+    |> get_scrubbers
+  end
+
+  def filter_tags(html, nil) do
+    filter_tags(html, get_scrubbers())
+  end
+
+  def filter_tags(html, scrubbers) when is_list(scrubbers) do
+    Enum.reduce(scrubbers, html, fn scrubber, html ->
+      filter_tags(html, scrubber)
+    end)
+  end
+
+  def filter_tags(html, scrubber), do: Scrubber.scrub(html, scrubber)
+  def filter_tags(html), do: filter_tags(html, nil)
+  def strip_tags(html), do: Scrubber.scrub(html, Scrubber.StripTags)
+
+  def get_cached_scrubbed_html_for_object(content, scrubbers, object, module) do
+    key = "#{module}#{generate_scrubber_signature(scrubbers)}|#{object.id}"
+    Cachex.fetch!(:scrubber_cache, key, fn _key -> ensure_scrubbed_html(content, scrubbers) end)
+  end
+
+  def get_cached_stripped_html_for_object(content, object, module) do
+    get_cached_scrubbed_html_for_object(
+      content,
+      HtmlSanitizeEx.Scrubber.StripTags,
+      object,
+      module
+    )
+  end
+
+  def ensure_scrubbed_html(
+        content,
+        scrubbers
+      ) do
+    {:commit, filter_tags(content, scrubbers)}
+  end
+
+  defp generate_scrubber_signature(scrubber) when is_atom(scrubber) do
+    generate_scrubber_signature([scrubber])
+  end
+
+  defp generate_scrubber_signature(scrubbers) do
+    Enum.reduce(scrubbers, "", fn scrubber, signature ->
+      "#{signature}#{to_string(scrubber)}"
+    end)
+  end
+end
+
+defmodule Pleroma.HTML.Scrubber.TwitterText do
+  @moduledoc """
+  An HTML scrubbing policy which limits to twitter-style text.  Only
+  paragraphs, breaks and links are allowed through the filter.
+  """
+
   @markup Application.get_env(:pleroma, :markup)
+  @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
+  @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
+
+  require HtmlSanitizeEx.Scrubber.Meta
+  alias HtmlSanitizeEx.Scrubber.Meta
+
+  Meta.remove_cdata_sections_before_scrub()
+  Meta.strip_comments()
+
+  # links
+  Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
+  Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
 
-  def filter_tags(html) do
-    scrubber = Keyword.get(@markup, :scrub_policy)
-    html |> Scrubber.scrub(scrubber)
+  # paragraphs and linebreaks
+  Meta.allow_tag_with_these_attributes("br", [])
+  Meta.allow_tag_with_these_attributes("p", [])
+
+  # microformats
+  Meta.allow_tag_with_these_attributes("span", ["class"])
+
+  # allow inline images for custom emoji
+  @allow_inline_images Keyword.get(@markup, :allow_inline_images)
+
+  if @allow_inline_images do
+    # restrict img tags to http/https only, because of MediaProxy.
+    Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
+
+    Meta.allow_tag_with_these_attributes("img", [
+      "width",
+      "height",
+      "title",
+      "alt"
+    ])
   end
 
-  def strip_tags(html) do
-    html |> Scrubber.scrub(Scrubber.StripTags)
+  Meta.strip_everything_not_covered()
+end
+
+defmodule Pleroma.HTML.Scrubber.Default do
+  @doc "The default HTML scrubbing policy: no "
+
+  require HtmlSanitizeEx.Scrubber.Meta
+  alias HtmlSanitizeEx.Scrubber.Meta
+
+  @markup Application.get_env(:pleroma, :markup)
+  @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
+  @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
+
+  Meta.remove_cdata_sections_before_scrub()
+  Meta.strip_comments()
+
+  Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
+  Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
+
+  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("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("span", ["class"])
+  Meta.allow_tag_with_these_attributes("strong", [])
+  Meta.allow_tag_with_these_attributes("u", [])
+  Meta.allow_tag_with_these_attributes("ul", [])
+
+  @allow_inline_images Keyword.get(@markup, :allow_inline_images)
+
+  if @allow_inline_images do
+    # restrict img tags to http/https only, because of MediaProxy.
+    Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
+
+    Meta.allow_tag_with_these_attributes("img", [
+      "width",
+      "height",
+      "title",
+      "alt"
+    ])
   end
+
+  @allow_tables Keyword.get(@markup, :allow_tables)
+
+  if @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", [])
+  end
+
+  @allow_headings Keyword.get(@markup, :allow_headings)
+
+  if @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", [])
+  end
+
+  @allow_fonts Keyword.get(@markup, :allow_fonts)
+
+  if @allow_fonts do
+    Meta.allow_tag_with_these_attributes("font", ["face"])
+  end
+
+  Meta.strip_everything_not_covered()
+end
+
+defmodule Pleroma.HTML.Transform.MediaProxy do
+  @moduledoc "Transforms inline image URIs to use MediaProxy."
+
+  alias Pleroma.Web.MediaProxy
+
+  def before_scrub(html), do: html
+
+  def scrub_attribute("img", {"src", "http" <> target}) do
+    media_url =
+      ("http" <> target)
+      |> MediaProxy.url()
+
+    {"src", media_url}
+  end
+
+  def scrub_attribute(_tag, attribute), do: attribute
+
+  def scrub({"img", attributes, children}) do
+    attributes =
+      attributes
+      |> Enum.map(fn attr -> scrub_attribute("img", attr) end)
+      |> Enum.reject(&is_nil(&1))
+
+    {"img", attributes, children}
+  end
+
+  def scrub({:comment, _children}), do: ""
+
+  def scrub({tag, attributes, children}), do: {tag, attributes, children}
+  def scrub({_tag, children}), do: children
+  def scrub(text), do: text
 end