spc-pleroma/lib/pleroma/web/rich_media/card.ex

158 lines
3.9 KiB
Elixir
Raw Normal View History

RichMedia refactor Rich Media parsing was previously handled on-demand with a 2 second HTTP request timeout and retained only in Cachex. Every time a Pleroma instance is restarted it will have to request and parse the data for each status with a URL detected. When fetching a batch of statuses they were processed in parallel to attempt to keep the maximum latency at 2 seconds, but often resulted in a timeline appearing to hang during loading due to a URL that could not be successfully reached. URLs which had images links that expire (Amazon AWS) were parsed and inserted with a TTL to ensure the image link would not break. Rich Media data is now cached in the database and fetched asynchronously. Cachex is used as a read-through cache. When the data becomes available we stream an update to the clients. If the result is returned quickly the experience is almost seamless. Activities were already processed for their Rich Media data during ingestion to warm the cache, so users should not normally encounter the asynchronous loading of the Rich Media data. Implementation notes: - The async worker is a Task with a globally unique process name to prevent duplicate processing of the same URL - The Task will attempt to fetch the data 3 times with increasing sleep time between attempts - The HTTP request obeys the default HTTP request timeout value instead of 2 seconds - URLs that cannot be successfully parsed due to an unexpected error receives a negative cache entry for 15 minutes - URLs that fail with an expected error will receive a negative cache with no TTL - Activities that have no detected URLs insert a nil value in the Cachex :scrubber_cache so we do not repeat parsing the object content with Floki every time the activity is rendered - Expiring image URLs are handled with an Oban job - There is no automatic cleanup of the Rich Media data in the database, but it is safe to delete at any time - The post draft/preview feature makes the URL processing synchronous so the rendered post preview will have an accurate rendering Overall performance of timelines and creating new posts which contain URLs is greatly improved.
2024-02-11 21:11:52 +00:00
defmodule Pleroma.Web.RichMedia.Card do
use Ecto.Schema
import Ecto.Changeset
import Ecto.Query
alias Pleroma.Activity
alias Pleroma.HTML
alias Pleroma.Object
alias Pleroma.Repo
alias Pleroma.Web.RichMedia.Backfill
alias Pleroma.Web.RichMedia.Parser
@cachex Pleroma.Config.get([:cachex, :provider], Cachex)
@config_impl Application.compile_env(:pleroma, [__MODULE__, :config_impl], Pleroma.Config)
@type t :: %__MODULE__{}
schema "rich_media_card" do
field(:url_hash, :binary)
field(:fields, :map)
timestamps()
end
@doc false
def changeset(card, attrs) do
card
|> cast(attrs, [:url_hash, :fields])
|> validate_required([:url_hash, :fields])
|> unique_constraint(:url_hash)
end
@spec create(String.t(), map()) :: {:ok, t()}
def create(url, fields) do
url_hash = url_to_hash(url)
fields = Map.put_new(fields, "url", url)
%__MODULE__{}
|> changeset(%{url_hash: url_hash, fields: fields})
|> Repo.insert(on_conflict: {:replace, [:fields]}, conflict_target: :url_hash)
end
@spec delete(String.t()) :: {:ok, Ecto.Schema.t()} | {:error, Ecto.Changeset.t()} | :ok
def delete(url) do
url_hash = url_to_hash(url)
@cachex.del(:rich_media_cache, url_hash)
case get_by_url(url) do
2024-05-08 02:27:18 +00:00
%__MODULE__{} = card -> Repo.delete(card)
RichMedia refactor Rich Media parsing was previously handled on-demand with a 2 second HTTP request timeout and retained only in Cachex. Every time a Pleroma instance is restarted it will have to request and parse the data for each status with a URL detected. When fetching a batch of statuses they were processed in parallel to attempt to keep the maximum latency at 2 seconds, but often resulted in a timeline appearing to hang during loading due to a URL that could not be successfully reached. URLs which had images links that expire (Amazon AWS) were parsed and inserted with a TTL to ensure the image link would not break. Rich Media data is now cached in the database and fetched asynchronously. Cachex is used as a read-through cache. When the data becomes available we stream an update to the clients. If the result is returned quickly the experience is almost seamless. Activities were already processed for their Rich Media data during ingestion to warm the cache, so users should not normally encounter the asynchronous loading of the Rich Media data. Implementation notes: - The async worker is a Task with a globally unique process name to prevent duplicate processing of the same URL - The Task will attempt to fetch the data 3 times with increasing sleep time between attempts - The HTTP request obeys the default HTTP request timeout value instead of 2 seconds - URLs that cannot be successfully parsed due to an unexpected error receives a negative cache entry for 15 minutes - URLs that fail with an expected error will receive a negative cache with no TTL - Activities that have no detected URLs insert a nil value in the Cachex :scrubber_cache so we do not repeat parsing the object content with Floki every time the activity is rendered - Expiring image URLs are handled with an Oban job - There is no automatic cleanup of the Rich Media data in the database, but it is safe to delete at any time - The post draft/preview feature makes the URL processing synchronous so the rendered post preview will have an accurate rendering Overall performance of timelines and creating new posts which contain URLs is greatly improved.
2024-02-11 21:11:52 +00:00
nil -> :ok
end
end
@spec get_by_url(String.t() | nil) :: t() | nil | :error
def get_by_url(url) when is_binary(url) do
if @config_impl.get([:rich_media, :enabled]) do
url_hash = url_to_hash(url)
@cachex.fetch!(:rich_media_cache, url_hash, fn _ ->
result =
__MODULE__
|> where(url_hash: ^url_hash)
|> Repo.one()
case result do
%__MODULE__{} = card -> {:commit, card}
_ -> {:ignore, nil}
end
end)
else
:error
end
end
def get_by_url(nil), do: nil
@spec get_or_backfill_by_url(String.t(), map()) :: t() | nil
def get_or_backfill_by_url(url, backfill_opts \\ %{}) do
case get_by_url(url) do
%__MODULE__{} = card ->
card
nil ->
backfill_opts = Map.put(backfill_opts, :url, url)
Backfill.start(backfill_opts)
nil
:error ->
nil
end
end
@spec get_by_object(Object.t()) :: t() | nil | :error
def get_by_object(object) do
case HTML.extract_first_external_url_from_object(object) do
nil -> nil
url -> get_or_backfill_by_url(url)
end
end
@spec get_by_activity(Activity.t()) :: t() | nil | :error
# Fake/Draft activity
def get_by_activity(%Activity{id: "pleroma:fakeid"} = activity) do
with %Object{} = object <- Object.normalize(activity, fetch: false),
url when not is_nil(url) <- HTML.extract_first_external_url_from_object(object) do
case get_by_url(url) do
# Cache hit
%__MODULE__{} = card ->
card
# Cache miss, but fetch for rendering the Draft
_ ->
with {:ok, fields} <- Parser.parse(url),
{:ok, card} <- create(url, fields) do
card
else
_ -> nil
end
end
else
_ ->
nil
end
end
def get_by_activity(activity) do
with %Object{} = object <- Object.normalize(activity, fetch: false),
{_, nil} <- {:cached, get_cached_url(object, activity.id)} do
nil
else
{:cached, url} ->
get_or_backfill_by_url(url, %{activity_id: activity.id})
_ ->
:error
end
end
@spec url_to_hash(String.t()) :: String.t()
def url_to_hash(url) do
:crypto.hash(:sha256, url) |> Base.encode16(case: :lower)
end
defp get_cached_url(object, activity_id) do
key = "URL|#{activity_id}"
@cachex.fetch!(:scrubber_cache, key, fn _ ->
url = HTML.extract_first_external_url_from_object(object)
Activity.HTML.add_cache_key_for(activity_id, key)
{:commit, url}
end)
end
end