2019-01-28 07:04:54 +01:00
|
|
|
# Pleroma: A lightweight social networking server
|
2020-03-03 23:44:49 +01:00
|
|
|
# Copyright _ 2017-2020 Pleroma Authors <https://pleroma.social/>
|
2019-01-28 07:04:54 +01:00
|
|
|
# SPDX-License-Identifier: AGPL-3.0-only
|
|
|
|
|
|
|
|
defmodule Pleroma.Web.RichMedia.Helpers do
|
2019-02-09 16:16:26 +01:00
|
|
|
alias Pleroma.Activity
|
2019-06-25 21:25:37 +02:00
|
|
|
alias Pleroma.Config
|
2019-02-09 16:16:26 +01:00
|
|
|
alias Pleroma.HTML
|
2019-03-05 03:52:23 +01:00
|
|
|
alias Pleroma.Object
|
2019-01-28 07:04:54 +01:00
|
|
|
alias Pleroma.Web.RichMedia.Parser
|
|
|
|
|
2020-08-03 19:37:31 +02:00
|
|
|
@rich_media_options [
|
|
|
|
pool: :media,
|
|
|
|
max_body: 2_000_000
|
|
|
|
]
|
|
|
|
|
2020-06-09 19:49:24 +02:00
|
|
|
@spec validate_page_url(URI.t() | binary()) :: :ok | :error
|
2019-03-04 19:38:23 +01:00
|
|
|
defp validate_page_url(page_url) when is_binary(page_url) do
|
2020-07-22 00:18:17 +02:00
|
|
|
validate_tld = Pleroma.Config.get([Pleroma.Formatter, :validate_tld])
|
2019-06-26 05:24:12 +02:00
|
|
|
|
2019-06-25 21:25:37 +02:00
|
|
|
page_url
|
2020-07-22 00:18:17 +02:00
|
|
|
|> Linkify.Parser.url?(validate_tld: validate_tld)
|
2019-06-25 21:25:37 +02:00
|
|
|
|> parse_uri(page_url)
|
|
|
|
end
|
2019-06-18 15:08:18 +02:00
|
|
|
|
2020-06-09 19:49:24 +02:00
|
|
|
defp validate_page_url(%URI{host: host, scheme: "https", authority: authority})
|
|
|
|
when is_binary(authority) do
|
2019-06-25 14:52:53 +02:00
|
|
|
cond do
|
2019-06-25 21:25:37 +02:00
|
|
|
host in Config.get([:rich_media, :ignore_hosts], []) ->
|
2019-06-25 14:52:53 +02:00
|
|
|
:error
|
|
|
|
|
2019-06-25 21:25:37 +02:00
|
|
|
get_tld(host) in Config.get([:rich_media, :ignore_tld], []) ->
|
|
|
|
:error
|
2019-06-25 14:52:53 +02:00
|
|
|
|
|
|
|
true ->
|
2019-06-25 21:25:37 +02:00
|
|
|
:ok
|
2019-03-04 19:38:23 +01:00
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
defp validate_page_url(_), do: :error
|
2019-02-24 20:13:46 +01:00
|
|
|
|
2019-06-25 21:25:37 +02:00
|
|
|
defp parse_uri(true, url) do
|
|
|
|
url
|
|
|
|
|> URI.parse()
|
|
|
|
|> validate_page_url
|
|
|
|
end
|
|
|
|
|
|
|
|
defp parse_uri(_, _), do: :error
|
|
|
|
|
|
|
|
defp get_tld(host) do
|
|
|
|
host
|
|
|
|
|> String.split(".")
|
|
|
|
|> Enum.reverse()
|
|
|
|
|> hd
|
|
|
|
end
|
|
|
|
|
2020-07-30 19:57:26 +02:00
|
|
|
def fetch_data_for_object(object) do
|
2019-06-25 21:25:37 +02:00
|
|
|
with true <- Config.get([:rich_media, :enabled]),
|
2019-05-17 20:49:43 +02:00
|
|
|
false <- object.data["sensitive"] || false,
|
2020-07-30 19:57:26 +02:00
|
|
|
{:ok, page_url} <-
|
2020-09-07 12:19:19 +02:00
|
|
|
HTML.extract_first_external_url_from_object(object),
|
2019-02-24 20:13:46 +01:00
|
|
|
:ok <- validate_page_url(page_url),
|
2019-01-28 07:04:54 +01:00
|
|
|
{:ok, rich_media} <- Parser.parse(page_url) do
|
|
|
|
%{page_url: page_url, rich_media: rich_media}
|
|
|
|
else
|
|
|
|
_ -> %{}
|
|
|
|
end
|
|
|
|
end
|
2019-03-23 03:26:49 +01:00
|
|
|
|
2020-07-30 19:57:26 +02:00
|
|
|
def fetch_data_for_activity(%Activity{data: %{"type" => "Create"}} = activity) do
|
|
|
|
with true <- Config.get([:rich_media, :enabled]),
|
|
|
|
%Object{} = object <- Object.normalize(activity) do
|
|
|
|
fetch_data_for_object(object)
|
|
|
|
else
|
|
|
|
_ -> %{}
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
2019-03-23 03:26:49 +01:00
|
|
|
def fetch_data_for_activity(_), do: %{}
|
2019-05-13 04:02:00 +02:00
|
|
|
|
2020-04-14 18:43:53 +02:00
|
|
|
def perform(:fetch, %Activity{} = activity) do
|
|
|
|
fetch_data_for_activity(activity)
|
|
|
|
:ok
|
|
|
|
end
|
2020-08-03 19:37:31 +02:00
|
|
|
|
|
|
|
def rich_media_get(url) do
|
|
|
|
headers = [{"user-agent", Pleroma.Application.user_agent() <> "; Bot"}]
|
|
|
|
|
|
|
|
options =
|
|
|
|
if Application.get_env(:tesla, :adapter) == Tesla.Adapter.Hackney do
|
|
|
|
Keyword.merge(@rich_media_options,
|
|
|
|
recv_timeout: 2_000,
|
|
|
|
with_body: true
|
|
|
|
)
|
|
|
|
else
|
|
|
|
@rich_media_options
|
|
|
|
end
|
|
|
|
|
2020-09-14 14:38:00 +02:00
|
|
|
head_check =
|
|
|
|
case Pleroma.HTTP.head(url, headers, adapter: options) do
|
|
|
|
# If the HEAD request didn't reach the server for whatever reason,
|
|
|
|
# we assume the GET that comes right after won't either
|
|
|
|
{:error, _} = e ->
|
|
|
|
e
|
|
|
|
|
|
|
|
{:ok, %Tesla.Env{status: 200, headers: headers}} ->
|
|
|
|
with :ok <- check_content_type(headers),
|
|
|
|
:ok <- check_content_length(headers),
|
|
|
|
do: :ok
|
|
|
|
|
|
|
|
_ ->
|
|
|
|
:ok
|
|
|
|
end
|
|
|
|
|
|
|
|
with :ok <- head_check, do: Pleroma.HTTP.get(url, headers, adapter: options)
|
|
|
|
end
|
|
|
|
|
|
|
|
defp check_content_type(headers) do
|
|
|
|
case List.keyfind(headers, "content-type", 0) do
|
|
|
|
{_, content_type} ->
|
|
|
|
case Plug.Conn.Utils.media_type(content_type) do
|
|
|
|
{:ok, "text", "html", _} -> :ok
|
|
|
|
_ -> {:error, {:content_type, content_type}}
|
|
|
|
end
|
|
|
|
|
|
|
|
_ ->
|
|
|
|
:ok
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
@max_body @rich_media_options[:max_body]
|
|
|
|
defp check_content_length(headers) do
|
|
|
|
case List.keyfind(headers, "content-length", 0) do
|
|
|
|
{_, maybe_content_length} ->
|
|
|
|
case Integer.parse(maybe_content_length) do
|
|
|
|
{content_length, ""} when content_length <= @max_body -> :ok
|
|
|
|
{_, ""} -> {:error, :body_too_large}
|
|
|
|
_ -> :ok
|
|
|
|
end
|
|
|
|
|
|
|
|
_ ->
|
|
|
|
:ok
|
|
|
|
end
|
2020-08-03 19:37:31 +02:00
|
|
|
end
|
2019-01-28 07:04:54 +01:00
|
|
|
end
|