git.squeep.com Git - akkoma/blob - lib/pleroma/html.ex

   1 defmodule Pleroma.HTML do
   2   alias HtmlSanitizeEx.Scrubber
   3
   4   defp get_scrubbers(scrubber) when is_atom(scrubber), do: [scrubber]
   5   defp get_scrubbers(scrubbers) when is_list(scrubbers), do: scrubbers
   6   defp get_scrubbers(_), do: [Pleroma.HTML.Scrubber.Default]
   7
   8   def get_scrubbers() do
   9     Pleroma.Config.get([:markup, :scrub_policy])
  10     |> get_scrubbers
  11   end
  12
  13   def filter_tags(html, nil) do
  14     get_scrubbers()
  15     |> Enum.reduce(html, fn scrubber, html ->
  16       filter_tags(html, scrubber)
  17     end)
  18   end
  19
  20   def filter_tags(html, scrubber) do
  21     html |> Scrubber.scrub(scrubber)
  22   end
  23
  24   def filter_tags(html), do: filter_tags(html, nil)
  25
  26   def strip_tags(html) do
  27     html |> Scrubber.scrub(Scrubber.StripTags)
  28   end
  29 end
  30
  31 defmodule Pleroma.HTML.Scrubber.TwitterText do
  32   @moduledoc """
  33   An HTML scrubbing policy which limits to twitter-style text.  Only
  34   paragraphs, breaks and links are allowed through the filter.
  35   """
  36
  37   @markup Application.get_env(:pleroma, :markup)
  38   @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
  39   @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
  40
  41   require HtmlSanitizeEx.Scrubber.Meta
  42   alias HtmlSanitizeEx.Scrubber.Meta
  43
  44   Meta.remove_cdata_sections_before_scrub()
  45   Meta.strip_comments()
  46
  47   # links
  48   Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes)
  49   Meta.allow_tag_with_these_attributes("a", ["name", "title"])
  50
  51   # paragraphs and linebreaks
  52   Meta.allow_tag_with_these_attributes("br", [])
  53   Meta.allow_tag_with_these_attributes("p", [])
  54
  55   # microformats
  56   Meta.allow_tag_with_these_attributes("span", [])
  57
  58   # allow inline images for custom emoji
  59   @allow_inline_images Keyword.get(@markup, :allow_inline_images)
  60
  61   if @allow_inline_images do
  62     # restrict img tags to http/https only, because of MediaProxy.
  63     Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
  64
  65     Meta.allow_tag_with_these_attributes("img", [
  66       "width",
  67       "height",
  68       "title",
  69       "alt"
  70     ])
  71   end
  72
  73   Meta.strip_everything_not_covered()
  74 end
  75
  76 defmodule Pleroma.HTML.Scrubber.Default do
  77   @doc "The default HTML scrubbing policy: no "
  78
  79   require HtmlSanitizeEx.Scrubber.Meta
  80   alias HtmlSanitizeEx.Scrubber.Meta
  81
  82   @markup Application.get_env(:pleroma, :markup)
  83   @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
  84   @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
  85
  86   Meta.remove_cdata_sections_before_scrub()
  87   Meta.strip_comments()
  88
  89   Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes)
  90   Meta.allow_tag_with_these_attributes("a", ["name", "title"])
  91
  92   Meta.allow_tag_with_these_attributes("abbr", ["title"])
  93
  94   Meta.allow_tag_with_these_attributes("b", [])
  95   Meta.allow_tag_with_these_attributes("blockquote", [])
  96   Meta.allow_tag_with_these_attributes("br", [])
  97   Meta.allow_tag_with_these_attributes("code", [])
  98   Meta.allow_tag_with_these_attributes("del", [])
  99   Meta.allow_tag_with_these_attributes("em", [])
 100   Meta.allow_tag_with_these_attributes("i", [])
 101   Meta.allow_tag_with_these_attributes("li", [])
 102   Meta.allow_tag_with_these_attributes("ol", [])
 103   Meta.allow_tag_with_these_attributes("p", [])
 104   Meta.allow_tag_with_these_attributes("pre", [])
 105   Meta.allow_tag_with_these_attributes("span", [])
 106   Meta.allow_tag_with_these_attributes("strong", [])
 107   Meta.allow_tag_with_these_attributes("u", [])
 108   Meta.allow_tag_with_these_attributes("ul", [])
 109
 110   @allow_inline_images Keyword.get(@markup, :allow_inline_images)
 111
 112   if @allow_inline_images do
 113     # restrict img tags to http/https only, because of MediaProxy.
 114     Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
 115
 116     Meta.allow_tag_with_these_attributes("img", [
 117       "width",
 118       "height",
 119       "title",
 120       "alt"
 121     ])
 122   end
 123
 124   @allow_tables Keyword.get(@markup, :allow_tables)
 125
 126   if @allow_tables do
 127     Meta.allow_tag_with_these_attributes("table", [])
 128     Meta.allow_tag_with_these_attributes("tbody", [])
 129     Meta.allow_tag_with_these_attributes("td", [])
 130     Meta.allow_tag_with_these_attributes("th", [])
 131     Meta.allow_tag_with_these_attributes("thead", [])
 132     Meta.allow_tag_with_these_attributes("tr", [])
 133   end
 134
 135   @allow_headings Keyword.get(@markup, :allow_headings)
 136
 137   if @allow_headings do
 138     Meta.allow_tag_with_these_attributes("h1", [])
 139     Meta.allow_tag_with_these_attributes("h2", [])
 140     Meta.allow_tag_with_these_attributes("h3", [])
 141     Meta.allow_tag_with_these_attributes("h4", [])
 142     Meta.allow_tag_with_these_attributes("h5", [])
 143   end
 144
 145   @allow_fonts Keyword.get(@markup, :allow_fonts)
 146
 147   if @allow_fonts do
 148     Meta.allow_tag_with_these_attributes("font", ["face"])
 149   end
 150
 151   Meta.strip_everything_not_covered()
 152 end
 153
 154 defmodule Pleroma.HTML.Transform.MediaProxy do
 155   @moduledoc "Transforms inline image URIs to use MediaProxy."
 156
 157   alias Pleroma.Web.MediaProxy
 158
 159   def before_scrub(html), do: html
 160
 161   def scrub_attribute("img", {"src", "http" <> target}) do
 162     media_url =
 163       ("http" <> target)
 164       |> MediaProxy.url()
 165
 166     {"src", media_url}
 167   end
 168
 169   def scrub_attribute(tag, attribute), do: attribute
 170
 171   def scrub({"img", attributes, children}) do
 172     attributes =
 173       attributes
 174       |> Enum.map(fn attr -> scrub_attribute("img", attr) end)
 175       |> Enum.reject(&is_nil(&1))
 176
 177     {"img", attributes, children}
 178   end
 179
 180   def scrub({:comment, children}), do: ""
 181
 182   def scrub({tag, attributes, children}), do: {tag, attributes, children}
 183   def scrub({tag, children}), do: children
 184   def scrub(text), do: text
 185 end