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   require HtmlSanitizeEx.Scrubber.Meta
  40   alias HtmlSanitizeEx.Scrubber.Meta
  41
  42   @valid_schemes [
  43     "https://",
  44     "http://",
  45     "dat://",
  46     "dweb://",
  47     "gopher://",
  48     "ipfs://",
  49     "ipns://",
  50     "irc:",
  51     "ircs:",
  52     "magnet:",
  53     "mailto:",
  54     "mumble:",
  55     "ssb://",
  56     "xmpp:"
  57   ]
  58
  59   Meta.remove_cdata_sections_before_scrub()
  60   Meta.strip_comments()
  61
  62   # links
  63   Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes)
  64   Meta.allow_tag_with_these_attributes("a", ["name", "title"])
  65
  66   # paragraphs and linebreaks
  67   Meta.allow_tag_with_these_attributes("br", [])
  68   Meta.allow_tag_with_these_attributes("p", [])
  69
  70   # microformats
  71   Meta.allow_tag_with_these_attributes("span", [])
  72
  73   # allow inline images for custom emoji
  74   @markup Application.get_env(:pleroma, :markup)
  75   @allow_inline_images Keyword.get(@markup, :allow_inline_images)
  76
  77   if @allow_inline_images do
  78     Meta.allow_tag_with_uri_attributes("img", ["src"], @valid_schemes)
  79
  80     Meta.allow_tag_with_these_attributes("img", [
  81       "width",
  82       "height",
  83       "title",
  84       "alt"
  85     ])
  86   end
  87
  88   Meta.strip_everything_not_covered()
  89 end
  90
  91 defmodule Pleroma.HTML.Scrubber.Default do
  92   @doc "The default HTML scrubbing policy: no "
  93
  94   require HtmlSanitizeEx.Scrubber.Meta
  95   alias HtmlSanitizeEx.Scrubber.Meta
  96
  97   @valid_schemes ["http", "https"]
  98
  99   Meta.remove_cdata_sections_before_scrub()
 100   Meta.strip_comments()
 101
 102   Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes)
 103   Meta.allow_tag_with_these_attributes("a", ["name", "title"])
 104
 105   Meta.allow_tag_with_these_attributes("b", [])
 106   Meta.allow_tag_with_these_attributes("blockquote", [])
 107   Meta.allow_tag_with_these_attributes("br", [])
 108   Meta.allow_tag_with_these_attributes("code", [])
 109   Meta.allow_tag_with_these_attributes("del", [])
 110   Meta.allow_tag_with_these_attributes("em", [])
 111   Meta.allow_tag_with_these_attributes("i", [])
 112   Meta.allow_tag_with_these_attributes("li", [])
 113   Meta.allow_tag_with_these_attributes("ol", [])
 114   Meta.allow_tag_with_these_attributes("p", [])
 115   Meta.allow_tag_with_these_attributes("pre", [])
 116   Meta.allow_tag_with_these_attributes("span", [])
 117   Meta.allow_tag_with_these_attributes("strong", [])
 118   Meta.allow_tag_with_these_attributes("u", [])
 119   Meta.allow_tag_with_these_attributes("ul", [])
 120
 121   @markup Application.get_env(:pleroma, :markup)
 122   @allow_inline_images Keyword.get(@markup, :allow_inline_images)
 123
 124   if @allow_inline_images do
 125     Meta.allow_tag_with_uri_attributes("img", ["src"], @valid_schemes)
 126
 127     Meta.allow_tag_with_these_attributes("img", [
 128       "width",
 129       "height",
 130       "title",
 131       "alt"
 132     ])
 133   end
 134
 135   @allow_tables Keyword.get(@markup, :allow_tables)
 136
 137   if @allow_tables do
 138     Meta.allow_tag_with_these_attributes("table", [])
 139     Meta.allow_tag_with_these_attributes("tbody", [])
 140     Meta.allow_tag_with_these_attributes("td", [])
 141     Meta.allow_tag_with_these_attributes("th", [])
 142     Meta.allow_tag_with_these_attributes("thead", [])
 143     Meta.allow_tag_with_these_attributes("tr", [])
 144   end
 145
 146   @allow_headings Keyword.get(@markup, :allow_headings)
 147
 148   if @allow_headings do
 149     Meta.allow_tag_with_these_attributes("h1", [])
 150     Meta.allow_tag_with_these_attributes("h2", [])
 151     Meta.allow_tag_with_these_attributes("h3", [])
 152     Meta.allow_tag_with_these_attributes("h4", [])
 153     Meta.allow_tag_with_these_attributes("h5", [])
 154   end
 155
 156   @allow_fonts Keyword.get(@markup, :allow_fonts)
 157
 158   if @allow_fonts do
 159     Meta.allow_tag_with_these_attributes("font", ["face"])
 160   end
 161
 162   Meta.strip_everything_not_covered()
 163 end
 164
 165 defmodule Pleroma.HTML.Transform.MediaProxy do
 166   @moduledoc "Transforms inline image URIs to use MediaProxy."
 167
 168   alias Pleroma.Web.MediaProxy
 169
 170   def before_scrub(html), do: html
 171
 172   def scrub_attribute("img", {"src", "http" <> target}) do
 173     media_url =
 174       ("http" <> target)
 175       |> MediaProxy.url()
 176
 177     {"src", media_url}
 178   end
 179
 180   def scrub_attribute(tag, attribute), do: attribute
 181
 182   def scrub({"img", attributes, children}) do
 183     attributes =
 184       attributes
 185       |> Enum.map(fn attr -> scrub_attribute("img", attr) end)
 186       |> Enum.reject(&is_nil(&1))
 187
 188     {"img", attributes, children}
 189   end
 190
 191   def scrub({tag, attributes, children}), do: {tag, attributes, children}
 192   def scrub({tag, children}), do: children
 193   def scrub(text), do: text
 194 end