diff --git a/config/config.exs b/config/config.exs index 66aee3264a..d7a82ff639 100644 --- a/config/config.exs +++ b/config/config.exs @@ -409,8 +409,8 @@ ignore_hosts: [], ignore_tld: ["local", "localdomain", "lan"], parsers: [ - Pleroma.Web.RichMedia.Parsers.TwitterCard, - Pleroma.Web.RichMedia.Parsers.OEmbed + Pleroma.Web.RichMedia.Parsers.OEmbed, + Pleroma.Web.RichMedia.Parsers.TwitterCard ], failure_backoff: 60_000, ttl_setters: [Pleroma.Web.RichMedia.Parser.TTL.AwsSignedUrl] diff --git a/lib/pleroma/web/mastodon_api/views/status_view.ex b/lib/pleroma/web/mastodon_api/views/status_view.ex index bac897a57b..5eb09f0a1e 100644 --- a/lib/pleroma/web/mastodon_api/views/status_view.ex +++ b/lib/pleroma/web/mastodon_api/views/status_view.ex @@ -367,36 +367,8 @@ def render("show.json", _) do nil end - def render("card.json", %{rich_media: rich_media, page_url: page_url}) do - page_url_data = URI.parse(page_url) - - page_url_data = - if is_binary(rich_media["url"]) do - URI.merge(page_url_data, URI.parse(rich_media["url"])) - else - page_url_data - end - - page_url = page_url_data |> to_string - - image_url = - if is_binary(rich_media["image"]) do - URI.merge(page_url_data, URI.parse(rich_media["image"])) - |> to_string - end - - %{ - type: "link", - provider_name: page_url_data.host, - 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"] || "", - pleroma: %{ - opengraph: rich_media - } - } + def render("card.json", %{rich_media: rich_media, page_url: _page_url}) do + rich_media end def render("card.json", _), do: nil diff --git a/lib/pleroma/web/plugs/http_security_plug.ex b/lib/pleroma/web/plugs/http_security_plug.ex index 0025b042a4..7be45dce83 100644 --- a/lib/pleroma/web/plugs/http_security_plug.ex +++ b/lib/pleroma/web/plugs/http_security_plug.ex @@ -79,7 +79,8 @@ def headers do "frame-ancestors 'none'", "style-src 'self' 'unsafe-inline'", "font-src 'self'", - "manifest-src 'self'" + "manifest-src 'self'", + "frame-src 'self' https:" ] @csp_start [Enum.join(static_csp_rules, ";") <> ";"] diff --git a/lib/pleroma/web/rich_media/helpers.ex b/lib/pleroma/web/rich_media/helpers.ex index 566fc8c8a9..249730aea6 100644 --- a/lib/pleroma/web/rich_media/helpers.ex +++ b/lib/pleroma/web/rich_media/helpers.ex @@ -8,6 +8,7 @@ defmodule Pleroma.Web.RichMedia.Helpers do alias Pleroma.HTML alias Pleroma.Object alias Pleroma.Web.RichMedia.Parser + alias Pleroma.Web.RichMedia.Parser.Card @options [ pool: :media, @@ -15,6 +16,8 @@ defmodule Pleroma.Web.RichMedia.Helpers do recv_timeout: 2_000 ] + @headers [{"user-agent", Pleroma.Application.user_agent() <> "; Bot"}] + @spec validate_page_url(URI.t() | binary()) :: :ok | :error defp validate_page_url(page_url) when is_binary(page_url) do validate_tld = Config.get([Pleroma.Formatter, :validate_tld]) @@ -55,12 +58,23 @@ defp get_tld(host) do |> hd end + defp strip_card(%Card{} = card) do + card + |> Map.from_struct() + |> Map.new(fn {k, v} -> {Atom.to_string(k), v} end) + end + + defp strip_card(%{} = card) do + Map.new(card, fn {k, v} -> {Atom.to_string(k), v} end) + end + def fetch_data_for_object(object) do with true <- Config.get([:rich_media, :enabled]), {:ok, page_url} <- HTML.extract_first_external_url_from_object(object), :ok <- validate_page_url(page_url), - {:ok, rich_media} <- Parser.parse(page_url) do + {:ok, rich_media} <- Parser.parse(page_url), + rich_media <- strip_card(rich_media) do %{page_url: page_url, rich_media: rich_media} else _ -> %{} @@ -78,8 +92,12 @@ def fetch_data_for_activity(%Activity{data: %{"type" => "Create"}} = activity) d def fetch_data_for_activity(_), do: %{} + def oembed_get(url) do + Pleroma.HTTP.get(url, @headers, @options) + end + def rich_media_get(url) do - headers = [{"user-agent", Pleroma.Application.user_agent() <> "; Bot"}] + headers = @headers head_check = case Pleroma.HTTP.head(url, headers, @options) do diff --git a/lib/pleroma/web/rich_media/parser.ex b/lib/pleroma/web/rich_media/parser.ex index d6b54943b4..d7b0cab6fe 100644 --- a/lib/pleroma/web/rich_media/parser.ex +++ b/lib/pleroma/web/rich_media/parser.ex @@ -4,6 +4,7 @@ defmodule Pleroma.Web.RichMedia.Parser do require Logger + alias Pleroma.Web.RichMedia.Parser.Card @cachex Pleroma.Config.get([:cachex, :provider], Cachex) @@ -131,13 +132,34 @@ defp get_ttl_from_image(data, url) do end def parse_url(url) do + case maybe_fetch_oembed(url) do + {:ok, %Card{} = card} -> {:ok, card} + _ -> fetch_document(url) + end + end + + defp maybe_fetch_oembed(url) do + with {:ok, oembed_url} <- OEmbedProviders.oembed_url(url), + {:ok, %Tesla.Env{body: json}} <- + Pleroma.Web.RichMedia.Helpers.oembed_get(oembed_url), + {:ok, data} <- Jason.decode(json), + %Card{} = card <- Card.from_oembed(data, url) do + {:ok, card} + else + {:error, error} -> {:error, error} + error -> {:error, error} + end + end + + defp fetch_document(url) do with {:ok, %Tesla.Env{body: html}} <- Pleroma.Web.RichMedia.Helpers.rich_media_get(url), {:ok, html} <- Floki.parse_document(html) do html |> maybe_parse() |> Map.put("url", url) |> clean_parsed_data() - |> check_parsed_data() + |> Card.from_meta_tags(url) + |> check_card() end end @@ -150,13 +172,13 @@ defp maybe_parse(html) do end) end - defp check_parsed_data(%{"title" => title} = data) + defp check_card(%Card{title: title} = card) when is_binary(title) and title != "" do - {:ok, data} + {:ok, card} end - defp check_parsed_data(data) do - {:error, {:invalid_metadata, data}} + defp check_card(card) do + {:error, {:invalid_metadata, card}} end defp clean_parsed_data(data) do diff --git a/lib/pleroma/web/rich_media/parser/card.ex b/lib/pleroma/web/rich_media/parser/card.ex new file mode 100644 index 0000000000..5ea7194858 --- /dev/null +++ b/lib/pleroma/web/rich_media/parser/card.ex @@ -0,0 +1,75 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2021 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.RichMedia.Parser.Card do + @types ["link", "photo", "video", "rich"] + + # https://docs.joinmastodon.org/entities/card/ + defstruct url: nil, + title: nil, + description: "", + type: "link", + author_name: "", + author_url: "", + provider_name: "", + provider_url: "", + html: "", + width: 0, + height: 0, + image: nil, + embed_url: "", + blurhash: nil + + def from_oembed(%{"type" => type, "title" => title} = oembed, url) when type in @types do + %__MODULE__{ + url: url, + title: title, + description: "", + type: type, + author_name: oembed["author_name"], + author_url: oembed["author_url"], + provider_name: oembed["provider_name"], + provider_url: oembed["provider_url"], + html: oembed["html"], + width: oembed["width"], + height: oembed["height"], + image: oembed["thumbnail_url"] |> proxy(), + embed_url: oembed["url"] |> proxy() + } + end + + def from_oembed(_oembed, _url), do: nil + + def from_meta_tags(rich_media, page_url) do + page_url_data = URI.parse(page_url) + + page_url_data = + if is_binary(rich_media["url"]) do + URI.merge(page_url_data, URI.parse(rich_media["url"])) + else + page_url_data + end + + page_url = page_url_data |> to_string + + image_url = + if is_binary(rich_media["image"]) do + URI.merge(page_url_data, URI.parse(rich_media["image"])) + |> to_string + end + + %__MODULE__{ + type: "link", + provider_name: page_url_data.host, + provider_url: page_url_data.scheme <> "://" <> page_url_data.host, + url: page_url, + image: image_url |> proxy(), + title: rich_media["title"] || "", + description: rich_media["description"] || "" + } + end + + defp proxy(url) when is_binary(url), do: Pleroma.Web.MediaProxy.url(url) + defp proxy(_), do: nil +end diff --git a/mix.exs b/mix.exs index 436381f325..be423d6dfe 100644 --- a/mix.exs +++ b/mix.exs @@ -198,6 +198,7 @@ defp deps do {:open_api_spex, git: "https://git.pleroma.social/pleroma/elixir-libraries/open_api_spex.git", ref: "f296ac0924ba3cf79c7a588c4c252889df4c2edd"}, + {:oembed_providers, "~> 0.1.0"}, ## dev & test {:ex_doc, "~> 0.22", only: :dev, runtime: false}, diff --git a/mix.lock b/mix.lock index 99be818262..5e6036a333 100644 --- a/mix.lock +++ b/mix.lock @@ -52,6 +52,7 @@ "gen_stage": {:hex, :gen_stage, "0.14.3", "d0c66f1c87faa301c1a85a809a3ee9097a4264b2edf7644bf5c123237ef732bf", [:mix], [], "hexpm"}, "gen_state_machine": {:hex, :gen_state_machine, "2.0.5", "9ac15ec6e66acac994cc442dcc2c6f9796cf380ec4b08267223014be1c728a95", [:mix], [], "hexpm"}, "gettext": {:hex, :gettext, "0.18.0", "406d6b9e0e3278162c2ae1de0a60270452c553536772167e2d701f028116f870", [:mix], [], "hexpm", "c3f850be6367ebe1a08616c2158affe4a23231c70391050bf359d5f92f66a571"}, + "glob": {:hex, :glob, "1.0.0", "b4d54d66e7797ce037cdd18f2587fc9932187355340e222cafe125cd333d7a0a", [:rebar3], [], "hexpm", "ca25de25ac5a762ba6c979718ae6afef8402cfc9155b87479d215fbe676801e1"}, "gun": {:git, "https://github.com/ninenines/gun.git", "921c47146b2d9567eac7e9a4d2ccc60fffd4f327", [ref: "921c47146b2d9567eac7e9a4d2ccc60fffd4f327"]}, "hackney": {:git, "https://git.pleroma.social/pleroma/elixir-libraries/hackney.git", "7d7119f0651515d6d7669c78393fd90950a3ec6e", [ref: "7d7119f0651515d6d7669c78393fd90950a3ec6e"]}, "html_entities": {:hex, :html_entities, "0.5.1", "1c9715058b42c35a2ab65edc5b36d0ea66dd083767bef6e3edb57870ef556549", [:mix], [], "hexpm", "30efab070904eb897ff05cd52fa61c1025d7f8ef3a9ca250bc4e6513d16c32de"}, @@ -82,6 +83,7 @@ "nimble_pool": {:hex, :nimble_pool, "0.1.0", "ffa9d5be27eee2b00b0c634eb649aa27f97b39186fec3c493716c2a33e784ec6", [:mix], [], "hexpm", "343a1eaa620ddcf3430a83f39f2af499fe2370390d4f785cd475b4df5acaf3f9"}, "nodex": {:git, "https://git.pleroma.social/pleroma/nodex", "cb6730f943cfc6aad674c92161be23a8411f15d1", [ref: "cb6730f943cfc6aad674c92161be23a8411f15d1"]}, "oban": {:hex, :oban, "2.3.4", "ec7509b9af2524d55f529cb7aee93d36131ae0bf0f37706f65d2fe707f4d9fd8", [:mix], [{:ecto_sql, ">= 3.4.3", [hex: :ecto_sql, repo: "hexpm", optional: false]}, {:jason, "~> 1.1", [hex: :jason, repo: "hexpm", optional: false]}, {:postgrex, "~> 0.14", [hex: :postgrex, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "c70ca0434758fd1805422ea4446af5e910ddc697c0c861549c8f0eb0cfbd2fdf"}, + "oembed_providers": {:hex, :oembed_providers, "0.1.0", "9b336ee5f3ca20ee4ed005383c74b154d30d0abeb98e95828855c0e2841ae46b", [:mix], [{:glob, "~> 1.0", [hex: :glob, repo: "hexpm", optional: false]}, {:jason, "~> 1.2", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "ac1dda0f743aa6fdead3eef59decfefc9de91d550bf0805b8fce16ed10d421ba"}, "open_api_spex": {:git, "https://git.pleroma.social/pleroma/elixir-libraries/open_api_spex.git", "f296ac0924ba3cf79c7a588c4c252889df4c2edd", [ref: "f296ac0924ba3cf79c7a588c4c252889df4c2edd"]}, "p1_utils": {:hex, :p1_utils, "1.0.18", "3fe224de5b2e190d730a3c5da9d6e8540c96484cf4b4692921d1e28f0c32b01c", [:rebar3], [], "hexpm", "1fc8773a71a15553b179c986b22fbeead19b28fe486c332d4929700ffeb71f88"}, "parse_trans": {:git, "https://github.com/uwiger/parse_trans.git", "76abb347c3c1d00fb0ccf9e4b43e22b3d2288484", [tag: "3.3.0"]},