nrd.li/akkoma
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

Fix atom leak in Rich Media Parser

Browse source
This commit is contained in:
Egor Kislitsyn 2020-06-09 21:49:24 +04:00 committed by rinpatch
parent 6c90fc8e70
commit 8b8b8599e9
7 changed files with 91 additions and 102 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
lib/pleroma/web/mastodon_api/views/status_view.ex
View file

@ -307,8 +307,8 @@ def render("card.json", %{rich_media: rich_media, page_url: page_url}) do
page_url_data = URI.parse(page_url)
page_url_data =
if rich_media[:url] != nil do
URI.merge(page_url_data, URI.parse(rich_media[:url]))
if is_binary(rich_media["url"]) do
URI.merge(page_url_data, URI.parse(rich_media["url"]))
else
page_url_data
end
@ -316,11 +316,9 @@ def render("card.json", %{rich_media: rich_media, page_url: page_url}) do
page_url = page_url_data |> to_string
image_url =
if rich_media[:image] != nil do
URI.merge(page_url_data, URI.parse(rich_media[:image]))
if is_binary(rich_media["image"]) do
URI.merge(page_url_data, URI.parse(rich_media["image"]))
|> to_string
else
nil
end
%{
@ -329,8 +327,8 @@ def render("card.json", %{rich_media: rich_media, page_url: page_url}) do
provider_url: page_url_data.scheme <> "://" <> page_url_data.host,
url: page_url,
image: image_url |> MediaProxy.url(),
title: rich_media[:title] || "",
description: rich_media[:description] || "",
title: rich_media["title"] || "",
description: rich_media["description"] || "",
pleroma: %{
opengraph: rich_media
}

6
lib/pleroma/web/rich_media/helpers.ex
View file

@ -9,7 +9,7 @@ defmodule Pleroma.Web.RichMedia.Helpers do
alias Pleroma.Object
alias Pleroma.Web.RichMedia.Parser
@spec validate_page_url(any()) :: :ok | :error
@spec validate_page_url(URI.t() | binary()) :: :ok | :error
defp validate_page_url(page_url) when is_binary(page_url) do
validate_tld = Application.get_env(:auto_linker, :opts)[:validate_tld]
@ -18,8 +18,8 @@ defp validate_page_url(page_url) when is_binary(page_url) do
|> parse_uri(page_url)
end
defp validate_page_url(%URI{host: host, scheme: scheme, authority: authority})
when scheme == "https" and not is_nil(authority) do
defp validate_page_url(%URI{host: host, scheme: "https", authority: authority})
when is_binary(authority) do
cond do
host in Config.get([:rich_media, :ignore_hosts], []) ->
:error

12
lib/pleroma/web/rich_media/parser.ex
View file

@ -83,7 +83,7 @@ defp parse_url(url) do
html
|> parse_html()
|> maybe_parse()
|> Map.put(:url, url)
|> Map.put("url", url)
|> clean_parsed_data()
|> check_parsed_data()
rescue
@ -103,8 +103,8 @@ defp maybe_parse(html) do
end)
end
defp check_parsed_data(%{title: title} = data)
when is_binary(title) and byte_size(title) > 0 do
defp check_parsed_data(%{"title" => title} = data)
when is_binary(title) and title != "" do
{:ok, data}
end
@ -115,11 +115,7 @@ defp check_parsed_data(data) do
defp clean_parsed_data(data) do
data
|> Enum.reject(fn {key, val} ->
with {:ok, _} <- Jason.encode(%{key => val}) do
false
else
_ -> true
end
not match?({:ok, _}, Jason.encode(%{key => val}))
end)
|> Map.new()
end

8
lib/pleroma/web/rich_media/parsers/meta_tags_parser.ex
View file

@ -29,19 +29,19 @@ defp normalize_attributes(html_node, prefix, key_name, value_name) do
{_tag, attributes, _children} = html_node
data =
Enum.into(attributes, %{}, fn {name, value} ->
Map.new(attributes, fn {name, value} ->
{name, String.trim_leading(value, "#{prefix}:")}
end)
%{String.to_atom(data[key_name]) => data[value_name]}
%{data[key_name] => data[value_name]}
end
defp maybe_put_title(%{title: _} = meta, _), do: meta
defp maybe_put_title(%{"title" => _} = meta, _), do: meta
defp maybe_put_title(meta, html) when meta != %{} do
case get_page_title(html) do
"" -> meta
title -> Map.put_new(meta, :title, title)
title -> Map.put_new(meta, "title", title)
end
end

18
lib/pleroma/web/rich_media/parsers/oembed_parser.ex
View file

@ -5,7 +5,7 @@
defmodule Pleroma.Web.RichMedia.Parsers.OEmbed do
def parse(html, _data) do
with elements = [_ | _] <- get_discovery_data(html),
{:ok, oembed_url} <- get_oembed_url(elements),
oembed_url when is_binary(oembed_url) <- get_oembed_url(elements),
{:ok, oembed_data} <- get_oembed_data(oembed_url) do
{:ok, oembed_data}
else
@ -17,19 +17,13 @@ defp get_discovery_data(html) do
html |> Floki.find("link[type='application/json+oembed']")
end
defp get_oembed_url(nodes) do
{"link", attributes, _children} = nodes |> hd()
{:ok, Enum.into(attributes, %{})["href"]}
defp get_oembed_url([{"link", attributes, _children} | _]) do
Enum.find_value(attributes, fn {k, v} -> if k == "href", do: v end)
end
defp get_oembed_data(url) do
{:ok, %Tesla.Env{body: json}} = Pleroma.HTTP.get(url, [], adapter: [pool: :media])
{:ok, data} = Jason.decode(json)
data = data |> Map.new(fn {k, v} -> {String.to_atom(k), v} end)
{:ok, data}
with {:ok, %Tesla.Env{body: json}} <- Pleroma.HTTP.get(url, [], adapter: [pool: :media]) do
Jason.decode(json)
end
end
end

75
test/web/rich_media/parser_test.exs
View file

@ -60,19 +60,19 @@ test "returns error when no metadata present" do
test "doesn't just add a title" do
assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/non-ogp") ==
{:error,
"Found metadata was invalid or incomplete: %{url: \"http://example.com/non-ogp\"}"}
"Found metadata was invalid or incomplete: %{\"url\" => \"http://example.com/non-ogp\"}"}
end
test "parses ogp" do
assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/ogp") ==
{:ok,
%{
image: "http://ia.media-imdb.com/images/rock.jpg",
title: "The Rock",
description:
"image" => "http://ia.media-imdb.com/images/rock.jpg",
"title" => "The Rock",
"description" =>
"Directed by Michael Bay. With Sean Connery, Nicolas Cage, Ed Harris, John Spencer.",
type: "video.movie",
url: "http://example.com/ogp"
"type" => "video.movie",
"url" => "http://example.com/ogp"
}}
end
@ -80,12 +80,12 @@ test "falls back to <title> when ogp:title is missing" do
assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/ogp-missing-title") ==
{:ok,
%{
image: "http://ia.media-imdb.com/images/rock.jpg",
title: "The Rock (1996)",
description:
"image" => "http://ia.media-imdb.com/images/rock.jpg",
"title" => "The Rock (1996)",
"description" =>
"Directed by Michael Bay. With Sean Connery, Nicolas Cage, Ed Harris, John Spencer.",
type: "video.movie",
url: "http://example.com/ogp-missing-title"
"type" => "video.movie",
"url" => "http://example.com/ogp-missing-title"
}}
end
@ -93,12 +93,12 @@ test "parses twitter card" do
assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/twitter-card") ==
{:ok,
%{
card: "summary",
site: "@flickr",
image: "https://farm6.staticflickr.com/5510/14338202952_93595258ff_z.jpg",
title: "Small Island Developing States Photo Submission",
description: "View the album on Flickr.",
url: "http://example.com/twitter-card"
"card" => "summary",
"site" => "@flickr",
"image" => "https://farm6.staticflickr.com/5510/14338202952_93595258ff_z.jpg",
"title" => "Small Island Developing States Photo Submission",
"description" => "View the album on Flickr.",
"url" => "http://example.com/twitter-card"
}}
end
@ -106,27 +106,28 @@ test "parses OEmbed" do
assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/oembed") ==
{:ok,
%{
author_name: "‮‭‬bees‬",
author_url: "https://www.flickr.com/photos/bees/",
cache_age: 3600,
flickr_type: "photo",
height: "768",
html:
"author_name" => "‮‭‬bees‬",
"author_url" => "https://www.flickr.com/photos/bees/",
"cache_age" => 3600,
"flickr_type" => "photo",
"height" => "768",
"html" =>
"<a data-flickr-embed=\"true\" href=\"https://www.flickr.com/photos/bees/2362225867/\" title=\"Bacon Lollys by ‮‭‬bees‬, on Flickr\"><img src=\"https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_b.jpg\" width=\"1024\" height=\"768\" alt=\"Bacon Lollys\"></a><script async src=\"https://embedr.flickr.com/assets/client-code.js\" charset=\"utf-8\"></script>",
license: "All Rights Reserved",
license_id: 0,
provider_name: "Flickr",
provider_url: "https://www.flickr.com/",
thumbnail_height: 150,
thumbnail_url: "https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_q.jpg",
thumbnail_width: 150,
title: "Bacon Lollys",
type: "photo",
url: "http://example.com/oembed",
version: "1.0",
web_page: "https://www.flickr.com/photos/bees/2362225867/",
web_page_short_url: "https://flic.kr/p/4AK2sc",
width: "1024"
"license" => "All Rights Reserved",
"license_id" => 0,
"provider_name" => "Flickr",
"provider_url" => "https://www.flickr.com/",
"thumbnail_height" => 150,
"thumbnail_url" =>
"https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_q.jpg",
"thumbnail_width" => 150,
"title" => "Bacon Lollys",
"type" => "photo",
"url" => "http://example.com/oembed",
"version" => "1.0",
"web_page" => "https://www.flickr.com/photos/bees/2362225867/",
"web_page_short_url" => "https://flic.kr/p/4AK2sc",
"width" => "1024"
}}
end

60
test/web/rich_media/parsers/twitter_card_test.exs
View file

@ -19,11 +19,11 @@ test "parses twitter card with only name attributes" do
assert TwitterCard.parse(html, %{}) ==
{:ok,
%{
"app:id:googleplay": "com.nytimes.android",
"app:name:googleplay": "NYTimes",
"app:url:googleplay": "nytimes://reader/id/100000006583622",
site: nil,
title:
"app:id:googleplay" => "com.nytimes.android",
"app:name:googleplay" => "NYTimes",
"app:url:googleplay" => "nytimes://reader/id/100000006583622",
"site" => nil,
"title" =>
"She Was Arrested at 14. Then Her Photo Went to a Facial Recognition Database. - The New York Times"
}}
end
@ -36,15 +36,15 @@ test "parses twitter card with only property attributes" do
assert TwitterCard.parse(html, %{}) ==
{:ok,
%{
card: "summary_large_image",
description:
"card" => "summary_large_image",
"description" =>
"With little oversight, the N.Y.P.D. has been using powerful surveillance technology on photos of children and teenagers.",
image:
"image" =>
"https://static01.nyt.com/images/2019/08/01/nyregion/01nypd-juveniles-promo/01nypd-juveniles-promo-videoSixteenByNineJumbo1600.jpg",
"image:alt": "",
title:
"image:alt" => "",
"title" =>
"She Was Arrested at 14. Then Her Photo Went to a Facial Recognition Database.",
url:
"url" =>
"https://www.nytimes.com/2019/08/01/nyregion/nypd-facial-recognition-children-teenagers.html"
}}
end
@ -57,19 +57,19 @@ test "parses twitter card with name & property attributes" do
assert TwitterCard.parse(html, %{}) ==
{:ok,
%{
"app:id:googleplay": "com.nytimes.android",
"app:name:googleplay": "NYTimes",
"app:url:googleplay": "nytimes://reader/id/100000006583622",
card: "summary_large_image",
description:
"app:id:googleplay" => "com.nytimes.android",
"app:name:googleplay" => "NYTimes",
"app:url:googleplay" => "nytimes://reader/id/100000006583622",
"card" => "summary_large_image",
"description" =>
"With little oversight, the N.Y.P.D. has been using powerful surveillance technology on photos of children and teenagers.",
image:
"image" =>
"https://static01.nyt.com/images/2019/08/01/nyregion/01nypd-juveniles-promo/01nypd-juveniles-promo-videoSixteenByNineJumbo1600.jpg",
"image:alt": "",
site: nil,
title:
"image:alt" => "",
"site" => nil,
"title" =>
"She Was Arrested at 14. Then Her Photo Went to a Facial Recognition Database.",
url:
"url" =>
"https://www.nytimes.com/2019/08/01/nyregion/nypd-facial-recognition-children-teenagers.html"
}}
end
@ -86,11 +86,11 @@ test "respect only first title tag on the page" do
assert TwitterCard.parse(html, %{}) ==
{:ok,
%{
site: "@atlasobscura",
title:
"site" => "@atlasobscura",
"title" =>
"The Missing Grave of Margaret Corbin, Revolutionary War Veteran - Atlas Obscura",
card: "summary_large_image",
image: image_path
"card" => "summary_large_image",
"image" => image_path
}}
end
@ -102,12 +102,12 @@ test "takes first founded title in html head if there is html markup error" do
assert TwitterCard.parse(html, %{}) ==
{:ok,
%{
site: nil,
title:
"site" => nil,
"title" =>
"She Was Arrested at 14. Then Her Photo Went to a Facial Recognition Database. - The New York Times",
"app:id:googleplay": "com.nytimes.android",
"app:name:googleplay": "NYTimes",
"app:url:googleplay": "nytimes://reader/id/100000006583622"
"app:id:googleplay" => "com.nytimes.android",
"app:name:googleplay" => "NYTimes",
"app:url:googleplay" => "nytimes://reader/id/100000006583622"
}}
end
end