Page Menu
Home
Phorge
Search
Configure Global Search
Log In
Files
F85949489
No One
Temporary
Actions
View File
Edit File
Delete File
View Transforms
Subscribe
Mute Notifications
Flag For Later
Award Token
Authored By
Unknown
Size
13 KB
Referenced Files
None
Subscribers
None
View Options
diff --git a/lib/pleroma/html.ex b/lib/pleroma/html.ex
index 7f1dbe28c..2c701adb5 100644
--- a/lib/pleroma/html.ex
+++ b/lib/pleroma/html.ex
@@ -1,259 +1,280 @@
# Pleroma: A lightweight social networking server
# Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
defmodule Pleroma.HTML do
alias HtmlSanitizeEx.Scrubber
defp get_scrubbers(scrubber) when is_atom(scrubber), do: [scrubber]
defp get_scrubbers(scrubbers) when is_list(scrubbers), do: scrubbers
defp get_scrubbers(_), do: [Pleroma.HTML.Scrubber.Default]
def get_scrubbers do
Pleroma.Config.get([:markup, :scrub_policy])
|> get_scrubbers
end
def filter_tags(html, nil) do
filter_tags(html, get_scrubbers())
end
def filter_tags(html, scrubbers) when is_list(scrubbers) do
Enum.reduce(scrubbers, html, fn scrubber, html ->
filter_tags(html, scrubber)
end)
end
def filter_tags(html, scrubber), do: Scrubber.scrub(html, scrubber)
def filter_tags(html), do: filter_tags(html, nil)
def strip_tags(html), do: Scrubber.scrub(html, Scrubber.StripTags)
def get_cached_scrubbed_html_for_activity(content, scrubbers, activity, key \\ "") do
key = "#{key}#{generate_scrubber_signature(scrubbers)}|#{activity.id}"
Cachex.fetch!(:scrubber_cache, key, fn _key ->
ensure_scrubbed_html(content, scrubbers, activity.data["object"]["fake"] || false)
end)
end
def get_cached_stripped_html_for_activity(content, activity, key) do
get_cached_scrubbed_html_for_activity(
content,
HtmlSanitizeEx.Scrubber.StripTags,
activity,
key
)
end
def ensure_scrubbed_html(
content,
scrubbers,
false = _fake
) do
{:commit, filter_tags(content, scrubbers)}
end
def ensure_scrubbed_html(
content,
scrubbers,
true = _fake
) do
{:ignore, filter_tags(content, scrubbers)}
end
defp generate_scrubber_signature(scrubber) when is_atom(scrubber) do
generate_scrubber_signature([scrubber])
end
defp generate_scrubber_signature(scrubbers) do
Enum.reduce(scrubbers, "", fn scrubber, signature ->
"#{signature}#{to_string(scrubber)}"
end)
end
def extract_first_external_url(_, nil), do: {:error, "No content"}
def extract_first_external_url(object, content) do
key = "URL|#{object.id}"
Cachex.fetch!(:scrubber_cache, key, fn _key ->
result =
content
|> Floki.filter_out("a.mention")
|> Floki.attribute("a", "href")
|> Enum.at(0)
{:commit, {:ok, result}}
end)
end
end
defmodule Pleroma.HTML.Scrubber.TwitterText do
@moduledoc """
An HTML scrubbing policy which limits to twitter-style text. Only
paragraphs, breaks and links are allowed through the filter.
"""
@markup Application.get_env(:pleroma, :markup)
@valid_schemes Pleroma.Config.get([:uri_schemes, :valid_schemes], [])
require HtmlSanitizeEx.Scrubber.Meta
alias HtmlSanitizeEx.Scrubber.Meta
Meta.remove_cdata_sections_before_scrub()
Meta.strip_comments()
# links
Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
- Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
+
+ Meta.allow_tag_with_this_attribute_values("a", "class", [
+ "hashtag",
+ "u-url",
+ "mention",
+ "u-url mention",
+ "mention u-url"
+ ])
Meta.allow_tag_with_this_attribute_values("a", "rel", [
"tag",
"nofollow",
"noopener",
"noreferrer"
])
+ Meta.allow_tag_with_these_attributes("a", ["name", "title"])
+
# paragraphs and linebreaks
Meta.allow_tag_with_these_attributes("br", [])
Meta.allow_tag_with_these_attributes("p", [])
# microformats
- Meta.allow_tag_with_these_attributes("span", ["class"])
+ Meta.allow_tag_with_this_attribute_values("span", "class", ["h-card"])
+ Meta.allow_tag_with_these_attributes("span", [])
# allow inline images for custom emoji
@allow_inline_images Keyword.get(@markup, :allow_inline_images)
if @allow_inline_images do
# restrict img tags to http/https only, because of MediaProxy.
Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
Meta.allow_tag_with_these_attributes("img", [
"width",
"height",
"title",
"alt"
])
end
Meta.strip_everything_not_covered()
end
defmodule Pleroma.HTML.Scrubber.Default do
@doc "The default HTML scrubbing policy: no "
require HtmlSanitizeEx.Scrubber.Meta
alias HtmlSanitizeEx.Scrubber.Meta
# credo:disable-for-previous-line
# No idea how to fix this one…
@markup Application.get_env(:pleroma, :markup)
@valid_schemes Pleroma.Config.get([:uri_schemes, :valid_schemes], [])
Meta.remove_cdata_sections_before_scrub()
Meta.strip_comments()
Meta.allow_tag_with_uri_attributes("a", ["href", "data-user", "data-tag"], @valid_schemes)
- Meta.allow_tag_with_these_attributes("a", ["name", "title", "class"])
+
+ Meta.allow_tag_with_this_attribute_values("a", "class", [
+ "hashtag",
+ "u-url",
+ "mention",
+ "u-url mention",
+ "mention u-url"
+ ])
Meta.allow_tag_with_this_attribute_values("a", "rel", [
"tag",
"nofollow",
"noopener",
"noreferrer"
])
+ Meta.allow_tag_with_these_attributes("a", ["name", "title"])
+
Meta.allow_tag_with_these_attributes("abbr", ["title"])
Meta.allow_tag_with_these_attributes("b", [])
Meta.allow_tag_with_these_attributes("blockquote", [])
Meta.allow_tag_with_these_attributes("br", [])
Meta.allow_tag_with_these_attributes("code", [])
Meta.allow_tag_with_these_attributes("del", [])
Meta.allow_tag_with_these_attributes("em", [])
Meta.allow_tag_with_these_attributes("i", [])
Meta.allow_tag_with_these_attributes("li", [])
Meta.allow_tag_with_these_attributes("ol", [])
Meta.allow_tag_with_these_attributes("p", [])
Meta.allow_tag_with_these_attributes("pre", [])
- Meta.allow_tag_with_these_attributes("span", ["class"])
Meta.allow_tag_with_these_attributes("strong", [])
Meta.allow_tag_with_these_attributes("u", [])
Meta.allow_tag_with_these_attributes("ul", [])
+ Meta.allow_tag_with_this_attribute_values("span", "class", ["h-card"])
+ Meta.allow_tag_with_these_attributes("span", [])
+
@allow_inline_images Keyword.get(@markup, :allow_inline_images)
if @allow_inline_images do
# restrict img tags to http/https only, because of MediaProxy.
Meta.allow_tag_with_uri_attributes("img", ["src"], ["http", "https"])
Meta.allow_tag_with_these_attributes("img", [
"width",
"height",
"title",
"alt"
])
end
@allow_tables Keyword.get(@markup, :allow_tables)
if @allow_tables do
Meta.allow_tag_with_these_attributes("table", [])
Meta.allow_tag_with_these_attributes("tbody", [])
Meta.allow_tag_with_these_attributes("td", [])
Meta.allow_tag_with_these_attributes("th", [])
Meta.allow_tag_with_these_attributes("thead", [])
Meta.allow_tag_with_these_attributes("tr", [])
end
@allow_headings Keyword.get(@markup, :allow_headings)
if @allow_headings do
Meta.allow_tag_with_these_attributes("h1", [])
Meta.allow_tag_with_these_attributes("h2", [])
Meta.allow_tag_with_these_attributes("h3", [])
Meta.allow_tag_with_these_attributes("h4", [])
Meta.allow_tag_with_these_attributes("h5", [])
end
@allow_fonts Keyword.get(@markup, :allow_fonts)
if @allow_fonts do
Meta.allow_tag_with_these_attributes("font", ["face"])
end
Meta.strip_everything_not_covered()
end
defmodule Pleroma.HTML.Transform.MediaProxy do
@moduledoc "Transforms inline image URIs to use MediaProxy."
alias Pleroma.Web.MediaProxy
def before_scrub(html), do: html
def scrub_attribute("img", {"src", "http" <> target}) do
media_url =
("http" <> target)
|> MediaProxy.url()
{"src", media_url}
end
def scrub_attribute(_tag, attribute), do: attribute
def scrub({"img", attributes, children}) do
attributes =
attributes
|> Enum.map(fn attr -> scrub_attribute("img", attr) end)
|> Enum.reject(&is_nil(&1))
{"img", attributes, children}
end
def scrub({:comment, _children}), do: ""
def scrub({tag, attributes, children}), do: {tag, attributes, children}
def scrub({_tag, children}), do: children
def scrub(text), do: text
end
diff --git a/test/html_test.exs b/test/html_test.exs
index 0b5d3d892..08738276e 100644
--- a/test/html_test.exs
+++ b/test/html_test.exs
@@ -1,92 +1,163 @@
# Pleroma: A lightweight social networking server
# Copyright © 2017-2018 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
defmodule Pleroma.HTMLTest do
alias Pleroma.HTML
use Pleroma.DataCase
@html_sample """
<b>this is in bold</b>
<p>this is a paragraph</p>
this is a linebreak<br />
this is a link with allowed "rel" attribute: <a href="http://example.com/" rel="tag">example.com</a>
this is a link with not allowed "rel" attribute: <a href="http://example.com/" rel="tag noallowed">example.com</a>
this is an image: <img src="http://example.com/image.jpg"><br />
<script>alert('hacked')</script>
"""
@html_onerror_sample """
<img src="http://example.com/image.jpg" onerror="alert('hacked')">
"""
+ @html_span_class_sample """
+ <span class="animate-spin">hi</span>
+ """
+
+ @html_span_microformats_sample """
+ <span class="h-card"><a class="u-url mention">@<span>foo</span></a></span>
+ """
+
+ @html_span_invalid_microformats_sample """
+ <span class="h-card"><a class="u-url mention animate-spin">@<span>foo</span></a></span>
+ """
+
describe "StripTags scrubber" do
test "works as expected" do
expected = """
this is in bold
this is a paragraph
this is a linebreak
this is a link with allowed "rel" attribute: example.com
this is a link with not allowed "rel" attribute: example.com
this is an image:
alert('hacked')
"""
assert expected == HTML.strip_tags(@html_sample)
end
test "does not allow attribute-based XSS" do
expected = "\n"
assert expected == HTML.strip_tags(@html_onerror_sample)
end
end
describe "TwitterText scrubber" do
test "normalizes HTML as expected" do
expected = """
this is in bold
<p>this is a paragraph</p>
this is a linebreak<br />
this is a link with allowed "rel" attribute: <a href="http://example.com/" rel="tag">example.com</a>
this is a link with not allowed "rel" attribute: <a href="http://example.com/">example.com</a>
this is an image: <img src="http://example.com/image.jpg" /><br />
alert('hacked')
"""
assert expected == HTML.filter_tags(@html_sample, Pleroma.HTML.Scrubber.TwitterText)
end
test "does not allow attribute-based XSS" do
expected = """
<img src="http://example.com/image.jpg" />
"""
assert expected == HTML.filter_tags(@html_onerror_sample, Pleroma.HTML.Scrubber.TwitterText)
end
+
+ test "does not allow spans with invalid classes" do
+ expected = """
+ <span>hi</span>
+ """
+
+ assert expected ==
+ HTML.filter_tags(@html_span_class_sample, Pleroma.HTML.Scrubber.TwitterText)
+ end
+
+ test "does allow microformats" do
+ expected = """
+ <span class="h-card"><a class="u-url mention">@<span>foo</span></a></span>
+ """
+
+ assert expected ==
+ HTML.filter_tags(@html_span_microformats_sample, Pleroma.HTML.Scrubber.TwitterText)
+ end
+
+ test "filters invalid microformats markup" do
+ expected = """
+ <span class="h-card"><a>@<span>foo</span></a></span>
+ """
+
+ assert expected ==
+ HTML.filter_tags(
+ @html_span_invalid_microformats_sample,
+ Pleroma.HTML.Scrubber.TwitterText
+ )
+ end
end
describe "default scrubber" do
test "normalizes HTML as expected" do
expected = """
<b>this is in bold</b>
<p>this is a paragraph</p>
this is a linebreak<br />
this is a link with allowed "rel" attribute: <a href="http://example.com/" rel="tag">example.com</a>
this is a link with not allowed "rel" attribute: <a href="http://example.com/">example.com</a>
this is an image: <img src="http://example.com/image.jpg" /><br />
alert('hacked')
"""
assert expected == HTML.filter_tags(@html_sample, Pleroma.HTML.Scrubber.Default)
end
test "does not allow attribute-based XSS" do
expected = """
<img src="http://example.com/image.jpg" />
"""
assert expected == HTML.filter_tags(@html_onerror_sample, Pleroma.HTML.Scrubber.Default)
end
+
+ test "does not allow spans with invalid classes" do
+ expected = """
+ <span>hi</span>
+ """
+
+ assert expected == HTML.filter_tags(@html_span_class_sample, Pleroma.HTML.Scrubber.Default)
+ end
+
+ test "does allow microformats" do
+ expected = """
+ <span class="h-card"><a class="u-url mention">@<span>foo</span></a></span>
+ """
+
+ assert expected ==
+ HTML.filter_tags(@html_span_microformats_sample, Pleroma.HTML.Scrubber.Default)
+ end
+
+ test "filters invalid microformats markup" do
+ expected = """
+ <span class="h-card"><a>@<span>foo</span></a></span>
+ """
+
+ assert expected ==
+ HTML.filter_tags(
+ @html_span_invalid_microformats_sample,
+ Pleroma.HTML.Scrubber.Default
+ )
+ end
end
end
File Metadata
Details
Attached
Mime Type
text/x-diff
Expires
Wed, Nov 18, 2:24 AM (22 h, 21 m)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
1839997
Default Alt Text
(13 KB)
Attached To
Mode
rPUBE pleroma-upstream
Attached
Detach File
Event Timeline