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