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