1 # Pleroma: A lightweight social networking server
2 # Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
3 # SPDX-License-Identifier: AGPL-3.0-only
5 defmodule Pleroma.HTML do
6 alias HtmlSanitizeEx.Scrubber
8 defp get_scrubbers(scrubber) when is_atom(scrubber), do: [scrubber]
9 defp get_scrubbers(scrubbers) when is_list(scrubbers), do: scrubbers
10 defp get_scrubbers(_), do: [Pleroma.HTML.Scrubber.Default]
13 Pleroma.Config.get([:markup, :scrub_policy])
17 def filter_tags(html, nil) do
18 filter_tags(html, get_scrubbers())
21 def filter_tags(html, scrubbers) when is_list(scrubbers) do
22 Enum.reduce(scrubbers, html, fn scrubber, html ->
23 filter_tags(html, scrubber)
27 def filter_tags(html, scrubber), do: Scrubber.scrub(html, scrubber)
28 def filter_tags(html), do: filter_tags(html, nil)
29 def strip_tags(html), do: Scrubber.scrub(html, Scrubber.StripTags)
31 def get_cached_scrubbed_html_for_object(content, scrubbers, object, module) do
32 key = "#{module}#{generate_scrubber_signature(scrubbers)}|#{object.id}"
33 Cachex.fetch!(:scrubber_cache, key, fn _key -> ensure_scrubbed_html(content, scrubbers) end)
36 def get_cached_stripped_html_for_object(content, object, module) do
37 get_cached_scrubbed_html_for_object(
39 HtmlSanitizeEx.Scrubber.StripTags,
45 def ensure_scrubbed_html(
49 {:commit, filter_tags(content, scrubbers)}
52 defp generate_scrubber_signature(scrubber) when is_atom(scrubber) do
53 generate_scrubber_signature([scrubber])
56 defp generate_scrubber_signature(scrubbers) do
57 Enum.reduce(scrubbers, "", fn scrubber, signature ->
58 "#{signature}#{to_string(scrubber)}"
62 def extract_first_external_url(_, nil), do: {:error, "No content"}
64 def extract_first_external_url(object, content) do
65 key = "URL|#{object.id}"
67 Cachex.fetch!(:scrubber_cache, key, fn _key ->
70 |> Floki.filter_out("a.mention")
71 |> Floki.attribute("a", "href")
74 {:commit, {:ok, result}}
79 defmodule Pleroma.HTML.Scrubber.TwitterText do
81 An HTML scrubbing policy which limits to twitter-style text. Only
82 paragraphs, breaks and links are allowed through the filter.
85 @markup Application.get_env(:pleroma, :markup)
86 @valid_schemes Pleroma.Config.get([:uri_schemes, :valid_schemes], [])
88 require HtmlSanitizeEx.Scrubber.Meta
89 alias HtmlSanitizeEx.Scrubber.Meta
91 Meta.remove_cdata_sections_before_scrub()
95 Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
96 Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
98 Meta.allow_tag_with_this_attribute_values("a", "rel", [
105 # paragraphs and linebreaks
106 Meta.allow_tag_with_these_attributes("br", [])
107 Meta.allow_tag_with_these_attributes("p", [])
110 Meta.allow_tag_with_these_attributes("span", ["class"])
112 # allow inline images for custom emoji
113 @allow_inline_images Keyword.get(@markup, :allow_inline_images)
115 if @allow_inline_images do
116 # restrict img tags to http/https only, because of MediaProxy.
117 Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
119 Meta.allow_tag_with_these_attributes("img", [
127 Meta.strip_everything_not_covered()
130 defmodule Pleroma.HTML.Scrubber.Default do
131 @doc "The default HTML scrubbing policy: no "
133 require HtmlSanitizeEx.Scrubber.Meta
134 alias HtmlSanitizeEx.Scrubber.Meta
135 # credo:disable-for-previous-line
136 # No idea how to fix this one…
138 @markup Application.get_env(:pleroma, :markup)
139 @valid_schemes Pleroma.Config.get([:uri_schemes, :valid_schemes], [])
141 Meta.remove_cdata_sections_before_scrub()
142 Meta.strip_comments()
144 Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
145 Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
147 Meta.allow_tag_with_this_attribute_values("a", "rel", [
154 Meta.allow_tag_with_these_attributes("abbr", ["title"])
156 Meta.allow_tag_with_these_attributes("b", [])
157 Meta.allow_tag_with_these_attributes("blockquote", [])
158 Meta.allow_tag_with_these_attributes("br", [])
159 Meta.allow_tag_with_these_attributes("code", [])
160 Meta.allow_tag_with_these_attributes("del", [])
161 Meta.allow_tag_with_these_attributes("em", [])
162 Meta.allow_tag_with_these_attributes("i", [])
163 Meta.allow_tag_with_these_attributes("li", [])
164 Meta.allow_tag_with_these_attributes("ol", [])
165 Meta.allow_tag_with_these_attributes("p", [])
166 Meta.allow_tag_with_these_attributes("pre", [])
167 Meta.allow_tag_with_these_attributes("span", ["class"])
168 Meta.allow_tag_with_these_attributes("strong", [])
169 Meta.allow_tag_with_these_attributes("u", [])
170 Meta.allow_tag_with_these_attributes("ul", [])
172 @allow_inline_images Keyword.get(@markup, :allow_inline_images)
174 if @allow_inline_images do
175 # restrict img tags to http/https only, because of MediaProxy.
176 Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
178 Meta.allow_tag_with_these_attributes("img", [
186 @allow_tables Keyword.get(@markup, :allow_tables)
189 Meta.allow_tag_with_these_attributes("table", [])
190 Meta.allow_tag_with_these_attributes("tbody", [])
191 Meta.allow_tag_with_these_attributes("td", [])
192 Meta.allow_tag_with_these_attributes("th", [])
193 Meta.allow_tag_with_these_attributes("thead", [])
194 Meta.allow_tag_with_these_attributes("tr", [])
197 @allow_headings Keyword.get(@markup, :allow_headings)
199 if @allow_headings do
200 Meta.allow_tag_with_these_attributes("h1", [])
201 Meta.allow_tag_with_these_attributes("h2", [])
202 Meta.allow_tag_with_these_attributes("h3", [])
203 Meta.allow_tag_with_these_attributes("h4", [])
204 Meta.allow_tag_with_these_attributes("h5", [])
207 @allow_fonts Keyword.get(@markup, :allow_fonts)
210 Meta.allow_tag_with_these_attributes("font", ["face"])
213 Meta.strip_everything_not_covered()
216 defmodule Pleroma.HTML.Transform.MediaProxy do
217 @moduledoc "Transforms inline image URIs to use MediaProxy."
219 alias Pleroma.Web.MediaProxy
221 def before_scrub(html), do: html
223 def scrub_attribute("img", {"src", "http" <> target}) do
231 def scrub_attribute(_tag, attribute), do: attribute
233 def scrub({"img", attributes, children}) do
236 |> Enum.map(fn attr -> scrub_attribute("img", attr) end)
237 |> Enum.reject(&is_nil(&1))
239 {"img", attributes, children}
242 def scrub({:comment, _children}), do: ""
244 def scrub({tag, attributes, children}), do: {tag, attributes, children}
245 def scrub({_tag, children}), do: children
246 def scrub(text), do: text