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

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