diff --git a/changelog.d/3904.security b/changelog.d/3904.security new file mode 100644 index 000000000..04836d4e8 --- /dev/null +++ b/changelog.d/3904.security @@ -0,0 +1 @@ +HTTP Security: By default, don't allow unsafe-eval. The setting needs to be changed to allow Flash emulation. diff --git a/changelog.d/3907.skip b/changelog.d/3907.skip new file mode 100644 index 000000000..e69de29bb diff --git a/changelog.d/add-rbl-mrf.add b/changelog.d/add-rbl-mrf.add new file mode 100644 index 000000000..363270fb9 --- /dev/null +++ b/changelog.d/add-rbl-mrf.add @@ -0,0 +1 @@ +Add DNSRBL MRF diff --git a/changelog.d/anti-mentionspam-mrf.add b/changelog.d/anti-mentionspam-mrf.add new file mode 100644 index 000000000..9466f85f4 --- /dev/null +++ b/changelog.d/anti-mentionspam-mrf.add @@ -0,0 +1 @@ +Add Anti-mention Spam MRF backported from Rebased diff --git a/changelog.d/logger-metadata.add b/changelog.d/logger-metadata.add new file mode 100644 index 000000000..6c627a972 --- /dev/null +++ b/changelog.d/logger-metadata.add @@ -0,0 +1 @@ +Logger metadata is now attached to some logs to help with troubleshooting and analysis diff --git a/changelog.d/oban-queues.change b/changelog.d/oban-queues.change new file mode 100644 index 000000000..16df6409a --- /dev/null +++ b/changelog.d/oban-queues.change @@ -0,0 +1 @@ +Oban queues have refactored to simplify the queue design diff --git a/changelog.d/prometheus-docs.change b/changelog.d/prometheus-docs.change new file mode 100644 index 000000000..a9bd1e2e9 --- /dev/null +++ b/changelog.d/prometheus-docs.change @@ -0,0 +1 @@ +Update the documentation for configuring Prometheus metrics. diff --git a/changelog.d/promexdocs.add b/changelog.d/promexdocs.add new file mode 100644 index 000000000..dda972994 --- /dev/null +++ b/changelog.d/promexdocs.add @@ -0,0 +1 @@ +PromEx documentation diff --git a/changelog.d/qdrant_search.add b/changelog.d/qdrant_search.add new file mode 100644 index 000000000..9801131d1 --- /dev/null +++ b/changelog.d/qdrant_search.add @@ -0,0 +1 @@ +Add Qdrant/OpenAI embedding search diff --git a/changelog.d/show-reposter-replies.add b/changelog.d/show-reposter-replies.add new file mode 100644 index 000000000..3b852ec3b --- /dev/null +++ b/changelog.d/show-reposter-replies.add @@ -0,0 +1 @@ +Display reposted replies with exclude_replies: true \ No newline at end of file diff --git a/changelog.d/support-honk-image-summaries.add b/changelog.d/support-honk-image-summaries.add new file mode 100644 index 000000000..052c03f95 --- /dev/null +++ b/changelog.d/support-honk-image-summaries.add @@ -0,0 +1 @@ +Support honk-style attachment summaries as alt-text. diff --git a/config/config.exs b/config/config.exs index b93de52e1..039d68379 100644 --- a/config/config.exs +++ b/config/config.exs @@ -135,13 +135,13 @@ config :logger, :console, level: :debug, format: "\n$time $metadata[$level] $message\n", - metadata: [:request_id] + metadata: [:actor, :path, :type, :user] config :logger, :ex_syslogger, level: :debug, ident: "pleroma", format: "$metadata[$level] $message", - metadata: [:request_id] + metadata: [:actor, :path, :type, :user] config :mime, :types, %{ "application/xml" => ["xml"], @@ -410,6 +410,11 @@ accept: [], reject: [] +config :pleroma, :mrf_dnsrbl, + nameserver: "127.0.0.1", + port: 53, + zone: "bl.pleroma.com" + # threshold of 7 days config :pleroma, :mrf_object_age, threshold: 604_800, @@ -430,6 +435,8 @@ mention_parent: true, mention_quoted: true +config :pleroma, :mrf_antimentionspam, user_age_limit: 30_000 + config :pleroma, :rich_media, enabled: true, ignore_hosts: [], @@ -512,7 +519,8 @@ sts: false, sts_max_age: 31_536_000, ct_max_age: 2_592_000, - referrer_policy: "same-origin" + referrer_policy: "same-origin", + allow_unsafe_eval: false config :cors_plug, max_age: 86_400, @@ -574,24 +582,14 @@ log: false, queues: [ activity_expiration: 10, - token_expiration: 5, - filter_expiration: 1, - backup: 1, federator_incoming: 5, federator_outgoing: 5, ingestion_queue: 50, web_push: 50, - mailer: 10, transmogrifier: 20, - scheduled_activities: 10, - poll_notifications: 10, background: 5, - remote_fetcher: 2, - attachments_cleanup: 1, - new_users_digest: 1, - mute_expire: 5, search_indexing: [limit: 10, paused: true], - rich_media_expiration: 2 + slow: 1 ], plugins: [Oban.Plugins.Pruner], crontab: [ @@ -926,6 +924,19 @@ config :pleroma, Pleroma.Uploaders.Uploader, timeout: 30_000 +config :pleroma, Pleroma.Search.QdrantSearch, + qdrant_url: "http://127.0.0.1:6333/", + qdrant_api_key: "", + openai_url: "http://127.0.0.1:11345", + # The healthcheck url has to be set to nil when used with the real openai + # API, as it doesn't have a healthcheck endpoint. + openai_healthcheck_url: "http://127.0.0.1:11345/health", + openai_model: "snowflake/snowflake-arctic-embed-xs", + openai_api_key: "", + qdrant_index_configuration: %{ + vectors: %{size: 384, distance: "Cosine"} + } + # Import environment specific config. This must remain at the bottom # of this file so it overrides the configuration defined above. import_config "#{Mix.env()}.exs" diff --git a/config/dev.exs b/config/dev.exs index fe8de5045..f23719fe3 100644 --- a/config/dev.exs +++ b/config/dev.exs @@ -35,8 +35,8 @@ # configured to run both http and https servers on # different ports. -# Do not include metadata nor timestamps in development logs -config :logger, :console, format: "[$level] $message\n" +# Do not include timestamps in development logs +config :logger, :console, format: "$metadata[$level] $message\n" # Set a higher stacktrace during development. Avoid configuring such # in production as building large stacktraces may be expensive. diff --git a/config/test.exs b/config/test.exs index 3345bb3a9..6c88ad3c6 100644 --- a/config/test.exs +++ b/config/test.exs @@ -154,6 +154,7 @@ config :pleroma, Pleroma.ScheduledActivity, config_impl: Pleroma.UnstubbedConfigMock config :pleroma, Pleroma.Web.RichMedia.Helpers, config_impl: Pleroma.StaticStubbedConfigMock config :pleroma, Pleroma.Uploaders.IPFS, config_impl: Pleroma.UnstubbedConfigMock +config :pleroma, Pleroma.Web.Plugs.HTTPSecurityPlug, config_impl: Pleroma.StaticStubbedConfigMock peer_module = if String.to_integer(System.otp_release()) >= 25 do diff --git a/docs/configuration/cheatsheet.md b/docs/configuration/cheatsheet.md index 00dc2f941..8f6d6aeea 100644 --- a/docs/configuration/cheatsheet.md +++ b/docs/configuration/cheatsheet.md @@ -473,6 +473,7 @@ This will make Pleroma listen on `127.0.0.1` port `8080` and generate urls start * ``ct_max_age``: The maximum age for the `Expect-CT` header if sent. * ``referrer_policy``: The referrer policy to use, either `"same-origin"` or `"no-referrer"`. * ``report_uri``: Adds the specified url to `report-uri` and `report-to` group in CSP header. +* `allow_unsafe_eval`: Adds `wasm-unsafe-eval` to the CSP header. Needed for some non-essential frontend features like Flash emulation. ### Pleroma.Web.Plugs.RemoteIp diff --git a/docs/configuration/search.md b/docs/configuration/search.md index 0316c9bf4..d34f84d4f 100644 --- a/docs/configuration/search.md +++ b/docs/configuration/search.md @@ -10,6 +10,30 @@ To use built-in search that has no external dependencies, set the search module While it has no external dependencies, it has problems with performance and relevancy. +## QdrantSearch + +This uses the vector search engine [Qdrant](https://qdrant.tech) to search the posts in a vector space. This needs a way to generate embeddings and uses the [OpenAI API](https://platform.openai.com/docs/guides/embeddings/what-are-embeddings). This is implemented by several project besides OpenAI itself, including the python-based fastembed-server found in `supplemental/search/fastembed-api`. + +The default settings will support a setup where both the fastembed server and Qdrant run on the same system as pleroma. To use it, set the search provider and run the fastembed server, see the README in `supplemental/search/fastembed-api`: + +> config :pleroma, Pleroma.Search, module: Pleroma.Search.QdrantSearch + +Then, start the Qdrant server, see [here](https://qdrant.tech/documentation/quick-start/) for instructions. + +You will also need to create the Qdrant index once by running `mix pleroma.search.indexer create_index`. Running `mix pleroma.search.indexer index` will retroactively index the last 100_000 activities. + +### Indexing and model options + +To see the available configuration options, check out the QdrantSearch section in `config/config.exs`. + +The default indexing option work for the default model (`snowflake-arctic-embed-xs`). To optimize for a low memory footprint, adjust the index configuration as described in the [Qdrant docs](https://qdrant.tech/documentation/guides/optimize/). See also [this blog post](https://qdrant.tech/articles/memory-consumption/) that goes into detail. + +Different embedding models will need different vector size settings. You can see a list of the models supported by the fastembed server [here](https://qdrant.github.io/fastembed/examples/Supported_Models), including their vector dimensions. These vector dimensions need to be set in the `qdrant_index_configuration`. + +E.g, If you want to use `sentence-transformers/all-MiniLM-L6-v2` as a model, you will not need to adjust things, because it and `snowflake-arctic-embed-xs` are both 384 dimensional models. If you want to use `snowflake/snowflake-arctic-embed-l`, you will need to adjust the `size` parameter in the `qdrant_index_configuration` to 1024, as it has a dimension of 1024. + +When using a different model, you will need do drop the index and recreate it (`mix pleroma.search.indexer drop_index` and `mix pleroma.search.indexer create_index`), as the different embeddings are not compatible with each other. + ## Meilisearch Note that it's quite a bit more memory hungry than PostgreSQL (around 4-5G for ~1.2 million diff --git a/docs/development/API/prometheus.md b/docs/development/API/prometheus.md index a5158d905..140291fe0 100644 --- a/docs/development/API/prometheus.md +++ b/docs/development/API/prometheus.md @@ -1,44 +1,47 @@ -# Prometheus Metrics +# Prometheus / OpenTelemetry Metrics -Pleroma includes support for exporting metrics via the [prometheus_ex](https://github.com/deadtrickster/prometheus.ex) library. +Pleroma includes support for exporting metrics via the [prom_ex](https://github.com/akoutmos/prom_ex) library. +The metrics are exposed by a dedicated webserver/port to improve privacy and security. Config example: ``` -config :prometheus, Pleroma.Web.Endpoint.MetricsExporter, - enabled: true, - auth: {:basic, "myusername", "mypassword"}, - ip_whitelist: ["127.0.0.1"], - path: "/api/pleroma/app_metrics", - format: :text -``` - -* `enabled` (Pleroma extension) enables the endpoint -* `ip_whitelist` (Pleroma extension) could be used to restrict access only to specified IPs -* `auth` sets the authentication (`false` for no auth; configurable to HTTP Basic Auth, see [prometheus-plugs](https://github.com/deadtrickster/prometheus-plugs#exporting) documentation) -* `format` sets the output format (`:text` or `:protobuf`) -* `path` sets the path to app metrics page - - -## `/api/pleroma/app_metrics` - -### Exports Prometheus application metrics - -* Method: `GET` -* Authentication: not required by default (see configuration options above) -* Params: none -* Response: text - -## Grafana - -### Config example - -The following is a config example to use with [Grafana](https://grafana.com) +config :pleroma, Pleroma.PromEx, + disabled: false, + manual_metrics_start_delay: :no_delay, + drop_metrics_groups: [], + grafana: [ + host: System.get_env("GRAFANA_HOST", "http://localhost:3000"), + auth_token: System.get_env("GRAFANA_TOKEN"), + upload_dashboards_on_start: false, + folder_name: "BEAM", + annotate_app_lifecycle: true + ], + metrics_server: [ + port: 4021, + path: "/metrics", + protocol: :http, + pool_size: 5, + cowboy_opts: [], + auth_strategy: :none + ], + datasource: "Prometheus" ``` - - job_name: 'beam' - metrics_path: /api/pleroma/app_metrics - scheme: https + +PromEx supports the ability to automatically publish dashboards to your Grafana server as well as register Annotations. If you do not wish to configure this capability you must generate the dashboard JSON files and import them directly. You can find the mix commands in the upstream [documentation](https://hexdocs.pm/prom_ex/Mix.Tasks.PromEx.Dashboard.Export.html). You can find the list of modules enabled in Pleroma for which you should generate dashboards for by examining the contents of the `lib/pleroma/prom_ex.ex` module. + +## prometheus.yml + +The following is a bare minimum config example to use with [Prometheus](https://prometheus.io) or Prometheus-compatible software like [VictoriaMetrics](https://victoriametrics.com). + +``` +global: + scrape_interval: 15s + +scrape_configs: + - job_name: 'pleroma' + scheme: http static_configs: - - targets: ['pleroma.soykaf.com'] + - targets: ['pleroma.soykaf.com:4021'] ``` diff --git a/lib/mix/tasks/pleroma/search/indexer.ex b/lib/mix/tasks/pleroma/search/indexer.ex new file mode 100644 index 000000000..81a9fced6 --- /dev/null +++ b/lib/mix/tasks/pleroma/search/indexer.ex @@ -0,0 +1,80 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2021 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Mix.Tasks.Pleroma.Search.Indexer do + import Mix.Pleroma + import Ecto.Query + + alias Pleroma.Workers.SearchIndexingWorker + + def run(["create_index"]) do + start_pleroma() + + with :ok <- Pleroma.Config.get([Pleroma.Search, :module]).create_index() do + IO.puts("Index created") + else + e -> IO.puts("Could not create index: #{inspect(e)}") + end + end + + def run(["drop_index"]) do + start_pleroma() + + with :ok <- Pleroma.Config.get([Pleroma.Search, :module]).drop_index() do + IO.puts("Index dropped") + else + e -> IO.puts("Could not drop index: #{inspect(e)}") + end + end + + def run(["index" | options]) do + {options, [], []} = + OptionParser.parse( + options, + strict: [ + limit: :integer + ] + ) + + start_pleroma() + + limit = Keyword.get(options, :limit, 100_000) + + per_step = 1000 + chunks = max(div(limit, per_step), 1) + + 1..chunks + |> Enum.each(fn step -> + q = + from(a in Pleroma.Activity, + limit: ^per_step, + offset: ^per_step * (^step - 1), + select: [:id], + order_by: [desc: :id] + ) + + {:ok, ids} = + Pleroma.Repo.transaction(fn -> + Pleroma.Repo.stream(q, timeout: :infinity) + |> Enum.map(fn a -> + a.id + end) + end) + + IO.puts("Got #{length(ids)} activities, adding to indexer") + + ids + |> Enum.chunk_every(100) + |> Enum.each(fn chunk -> + IO.puts("Adding #{length(chunk)} activities to indexing queue") + + chunk + |> Enum.map(fn id -> + SearchIndexingWorker.new(%{"op" => "add_to_index", "activity" => id}) + end) + |> Oban.insert_all() + end) + end) + end +end diff --git a/lib/pleroma/application.ex b/lib/pleroma/application.ex index d266d1836..0d9757b44 100644 --- a/lib/pleroma/application.ex +++ b/lib/pleroma/application.ex @@ -14,6 +14,7 @@ defmodule Pleroma.Application do @name Mix.Project.config()[:name] @version Mix.Project.config()[:version] @repository Mix.Project.config()[:source_url] + @compile_env Mix.env() def name, do: @name def version, do: @version @@ -51,7 +52,11 @@ def start(_type, _args) do Pleroma.HTML.compile_scrubbers() Pleroma.Config.Oban.warn() Config.DeprecationWarnings.warn() - Pleroma.Web.Plugs.HTTPSecurityPlug.warn_if_disabled() + + if @compile_env != :test do + Pleroma.Web.Plugs.HTTPSecurityPlug.warn_if_disabled() + end + Pleroma.ApplicationRequirements.verify!() load_custom_modules() Pleroma.Docs.JSON.compile() diff --git a/lib/pleroma/scheduled_activity.ex b/lib/pleroma/scheduled_activity.ex index 63c6cb45b..c361d7d89 100644 --- a/lib/pleroma/scheduled_activity.ex +++ b/lib/pleroma/scheduled_activity.ex @@ -204,7 +204,7 @@ def due_activities(offset \\ 0) do def job_query(scheduled_activity_id) do from(j in Oban.Job, - where: j.queue == "scheduled_activities", + where: j.queue == "federator_outgoing", where: fragment("args ->> 'activity_id' = ?::text", ^to_string(scheduled_activity_id)) ) end diff --git a/lib/pleroma/search/database_search.ex b/lib/pleroma/search/database_search.ex index 11e99e7f1..c6fe8a9bd 100644 --- a/lib/pleroma/search/database_search.ex +++ b/lib/pleroma/search/database_search.ex @@ -48,6 +48,12 @@ def add_to_index(_activity), do: :ok @impl true def remove_from_index(_object), do: :ok + @impl true + def create_index, do: :ok + + @impl true + def drop_index, do: :ok + @impl true def healthcheck_endpoints, do: nil diff --git a/lib/pleroma/search/meilisearch.ex b/lib/pleroma/search/meilisearch.ex index 08c2f3d86..9bba5b30f 100644 --- a/lib/pleroma/search/meilisearch.ex +++ b/lib/pleroma/search/meilisearch.ex @@ -10,6 +10,12 @@ defmodule Pleroma.Search.Meilisearch do @behaviour Pleroma.Search.SearchBackend + @impl true + def create_index, do: :ok + + @impl true + def drop_index, do: :ok + defp meili_headers do private_key = Config.get([Pleroma.Search.Meilisearch, :private_key]) diff --git a/lib/pleroma/search/qdrant_search.ex b/lib/pleroma/search/qdrant_search.ex new file mode 100644 index 000000000..b659bb682 --- /dev/null +++ b/lib/pleroma/search/qdrant_search.ex @@ -0,0 +1,182 @@ +defmodule Pleroma.Search.QdrantSearch do + @behaviour Pleroma.Search.SearchBackend + import Ecto.Query + + alias Pleroma.Activity + alias Pleroma.Config.Getting, as: Config + + alias __MODULE__.OpenAIClient + alias __MODULE__.QdrantClient + + import Pleroma.Search.Meilisearch, only: [object_to_search_data: 1] + import Pleroma.Search.DatabaseSearch, only: [maybe_fetch: 3] + + @impl true + def create_index do + payload = Config.get([Pleroma.Search.QdrantSearch, :qdrant_index_configuration]) + + with {:ok, %{status: 200}} <- QdrantClient.put("/collections/posts", payload) do + :ok + else + e -> {:error, e} + end + end + + @impl true + def drop_index do + with {:ok, %{status: 200}} <- QdrantClient.delete("/collections/posts") do + :ok + else + e -> {:error, e} + end + end + + def get_embedding(text) do + with {:ok, %{body: %{"data" => [%{"embedding" => embedding}]}}} <- + OpenAIClient.post("/v1/embeddings", %{ + input: text, + model: Config.get([Pleroma.Search.QdrantSearch, :openai_model]) + }) do + {:ok, embedding} + else + _ -> + {:error, "Failed to get embedding"} + end + end + + defp actor_from_activity(%{data: %{"actor" => actor}}) do + actor + end + + defp actor_from_activity(_), do: nil + + defp build_index_payload(activity, embedding) do + actor = actor_from_activity(activity) + published_at = activity.data["published"] + + %{ + points: [ + %{ + id: activity.id |> FlakeId.from_string() |> Ecto.UUID.cast!(), + vector: embedding, + payload: %{actor: actor, published_at: published_at} + } + ] + } + end + + defp build_search_payload(embedding, options) do + base = %{ + vector: embedding, + limit: options[:limit] || 20, + offset: options[:offset] || 0 + } + + if author = options[:author] do + Map.put(base, :filter, %{ + must: [%{key: "actor", match: %{value: author.ap_id}}] + }) + else + base + end + end + + @impl true + def add_to_index(activity) do + # This will only index public or unlisted notes + maybe_search_data = object_to_search_data(activity.object) + + if activity.data["type"] == "Create" and maybe_search_data do + with {:ok, embedding} <- get_embedding(maybe_search_data.content), + {:ok, %{status: 200}} <- + QdrantClient.put( + "/collections/posts/points", + build_index_payload(activity, embedding) + ) do + :ok + else + e -> {:error, e} + end + else + :ok + end + end + + @impl true + def remove_from_index(object) do + activity = Activity.get_by_object_ap_id_with_object(object.data["id"]) + id = activity.id |> FlakeId.from_string() |> Ecto.UUID.cast!() + + with {:ok, %{status: 200}} <- + QdrantClient.post("/collections/posts/points/delete", %{"points" => [id]}) do + :ok + else + e -> {:error, e} + end + end + + @impl true + def search(user, original_query, options) do + query = "Represent this sentence for searching relevant passages: #{original_query}" + + with {:ok, embedding} <- get_embedding(query), + {:ok, %{body: %{"result" => result}}} <- + QdrantClient.post( + "/collections/posts/points/search", + build_search_payload(embedding, options) + ) do + ids = + Enum.map(result, fn %{"id" => id} -> + Ecto.UUID.dump!(id) + end) + + from(a in Activity, where: a.id in ^ids) + |> Activity.with_preloaded_object() + |> Activity.restrict_deactivated_users() + |> Ecto.Query.order_by([a], fragment("array_position(?, ?)", ^ids, a.id)) + |> Pleroma.Repo.all() + |> maybe_fetch(user, original_query) + else + _ -> + [] + end + end + + @impl true + def healthcheck_endpoints do + qdrant_health = + Config.get([Pleroma.Search.QdrantSearch, :qdrant_url]) + |> URI.parse() + |> Map.put(:path, "/healthz") + |> URI.to_string() + + openai_health = Config.get([Pleroma.Search.QdrantSearch, :openai_healthcheck_url]) + + [qdrant_health, openai_health] |> Enum.filter(& &1) + end +end + +defmodule Pleroma.Search.QdrantSearch.OpenAIClient do + use Tesla + alias Pleroma.Config.Getting, as: Config + + plug(Tesla.Middleware.BaseUrl, Config.get([Pleroma.Search.QdrantSearch, :openai_url])) + plug(Tesla.Middleware.JSON) + + plug(Tesla.Middleware.Headers, [ + {"Authorization", + "Bearer #{Pleroma.Config.get([Pleroma.Search.QdrantSearch, :openai_api_key])}"} + ]) +end + +defmodule Pleroma.Search.QdrantSearch.QdrantClient do + use Tesla + alias Pleroma.Config.Getting, as: Config + + plug(Tesla.Middleware.BaseUrl, Config.get([Pleroma.Search.QdrantSearch, :qdrant_url])) + plug(Tesla.Middleware.JSON) + + plug(Tesla.Middleware.Headers, [ + {"api-key", Pleroma.Config.get([Pleroma.Search.QdrantSearch, :qdrant_api_key])} + ]) +end diff --git a/lib/pleroma/search/search_backend.ex b/lib/pleroma/search/search_backend.ex index 13c887bc2..f4ed13c36 100644 --- a/lib/pleroma/search/search_backend.ex +++ b/lib/pleroma/search/search_backend.ex @@ -22,6 +22,16 @@ defmodule Pleroma.Search.SearchBackend do """ @callback remove_from_index(object :: Pleroma.Object.t()) :: :ok | {:error, any()} + @doc """ + Create the index + """ + @callback create_index() :: :ok | {:error, any()} + + @doc """ + Drop the index + """ + @callback drop_index() :: :ok | {:error, any()} + @doc """ Healthcheck endpoints of search backend infrastructure to monitor for controlling processing of jobs in the Oban queue. diff --git a/lib/pleroma/web/activity_pub/activity_pub.ex b/lib/pleroma/web/activity_pub/activity_pub.ex index 643877268..5bb0fba6e 100644 --- a/lib/pleroma/web/activity_pub/activity_pub.ex +++ b/lib/pleroma/web/activity_pub/activity_pub.ex @@ -979,8 +979,9 @@ defp restrict_media(query, _), do: query defp restrict_replies(query, %{exclude_replies: true}) do from( - [_activity, object] in query, - where: fragment("?->>'inReplyTo' is null", object.data) + [activity, object] in query, + where: + fragment("?->>'inReplyTo' is null or ?->>'type' = 'Announce'", object.data, activity.data) ) end diff --git a/lib/pleroma/web/activity_pub/activity_pub_controller.ex b/lib/pleroma/web/activity_pub/activity_pub_controller.ex index e38a94966..d2b2cae0b 100644 --- a/lib/pleroma/web/activity_pub/activity_pub_controller.ex +++ b/lib/pleroma/web/activity_pub/activity_pub_controller.ex @@ -52,6 +52,7 @@ defmodule Pleroma.Web.ActivityPub.ActivityPubController do when action in [:activity, :object] ) + plug(:log_inbox_metadata when action in [:inbox]) plug(:set_requester_reachable when action in [:inbox]) plug(:relay_active? when action in [:relay]) @@ -521,6 +522,13 @@ defp set_requester_reachable(%Plug.Conn{} = conn, _) do conn end + defp log_inbox_metadata(conn = %{params: %{"actor" => actor, "type" => type}}, _) do + Logger.metadata(actor: actor, type: type) + conn + end + + defp log_inbox_metadata(conn, _), do: conn + def upload_media(%{assigns: %{user: %User{} = user}} = conn, %{"file" => file} = data) do with {:ok, object} <- ActivityPub.upload( diff --git a/lib/pleroma/web/activity_pub/mrf/anti_mention_spam_policy.ex b/lib/pleroma/web/activity_pub/mrf/anti_mention_spam_policy.ex new file mode 100644 index 000000000..531e75ce8 --- /dev/null +++ b/lib/pleroma/web/activity_pub/mrf/anti_mention_spam_policy.ex @@ -0,0 +1,87 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2022 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.ActivityPub.MRF.AntiMentionSpamPolicy do + alias Pleroma.Config + alias Pleroma.User + require Pleroma.Constants + + @behaviour Pleroma.Web.ActivityPub.MRF.Policy + + defp user_has_posted?(%User{} = u), do: u.note_count > 0 + + defp user_has_age?(%User{} = u) do + user_age_limit = Config.get([:mrf_antimentionspam, :user_age_limit], 30_000) + diff = NaiveDateTime.utc_now() |> NaiveDateTime.diff(u.inserted_at, :millisecond) + diff >= user_age_limit + end + + defp good_reputation?(%User{} = u) do + user_has_age?(u) and user_has_posted?(u) + end + + # copied from HellthreadPolicy + defp get_recipient_count(message) do + recipients = (message["to"] || []) ++ (message["cc"] || []) + + follower_collection = + User.get_cached_by_ap_id(message["actor"] || message["attributedTo"]).follower_address + + if Enum.member?(recipients, Pleroma.Constants.as_public()) do + recipients = + recipients + |> List.delete(Pleroma.Constants.as_public()) + |> List.delete(follower_collection) + + {:public, length(recipients)} + else + recipients = + recipients + |> List.delete(follower_collection) + + {:not_public, length(recipients)} + end + end + + defp object_has_recipients?(%{"object" => object} = activity) do + {_, object_count} = get_recipient_count(object) + {_, activity_count} = get_recipient_count(activity) + object_count + activity_count > 0 + end + + defp object_has_recipients?(object) do + {_, count} = get_recipient_count(object) + count > 0 + end + + @impl true + def filter(%{"type" => "Create", "actor" => actor} = activity) do + with {:ok, %User{local: false} = u} <- User.get_or_fetch_by_ap_id(actor), + {:has_mentions, true} <- {:has_mentions, object_has_recipients?(activity)}, + {:good_reputation, true} <- {:good_reputation, good_reputation?(u)} do + {:ok, activity} + else + {:ok, %User{local: true}} -> + {:ok, activity} + + {:has_mentions, false} -> + {:ok, activity} + + {:good_reputation, false} -> + {:reject, "[AntiMentionSpamPolicy] User rejected"} + + {:error, _} -> + {:reject, "[AntiMentionSpamPolicy] Failed to get or fetch user by ap_id"} + + e -> + {:reject, "[AntiMentionSpamPolicy] Unhandled error #{inspect(e)}"} + end + end + + # in all other cases, pass through + def filter(message), do: {:ok, message} + + @impl true + def describe, do: {:ok, %{}} +end diff --git a/lib/pleroma/web/activity_pub/mrf/dnsrbl_policy.ex b/lib/pleroma/web/activity_pub/mrf/dnsrbl_policy.ex new file mode 100644 index 000000000..9543cc545 --- /dev/null +++ b/lib/pleroma/web/activity_pub/mrf/dnsrbl_policy.ex @@ -0,0 +1,142 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2024 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.ActivityPub.MRF.DNSRBLPolicy do + @moduledoc """ + Dynamic activity filtering based on an RBL database + + This MRF makes queries to a custom DNS server which will + respond with values indicating the classification of the domain + the activity originated from. This method has been widely used + in the email anti-spam industry for very fast reputation checks. + + e.g., if the DNS response is 127.0.0.1 or empty, the domain is OK + Other values such as 127.0.0.2 may be used for specific classifications. + + Information for why the host is blocked can be stored in a corresponding TXT record. + + This method is fail-open so if the queries fail the activites are accepted. + + An example of software meant for this purpsoe is rbldnsd which can be found + at http://www.corpit.ru/mjt/rbldnsd.html or mirrored at + https://git.pleroma.social/feld/rbldnsd + + It is highly recommended that you run your own copy of rbldnsd and use an + external mechanism to sync/share the contents of the zone file. This is + important to keep the latency on the queries as low as possible and prevent + your DNS server from being attacked so it fails and content is permitted. + """ + + @behaviour Pleroma.Web.ActivityPub.MRF.Policy + + alias Pleroma.Config + + require Logger + + @query_retries 1 + @query_timeout 500 + + @impl true + def filter(%{"actor" => actor} = object) do + actor_info = URI.parse(actor) + + with {:ok, object} <- check_rbl(actor_info, object) do + {:ok, object} + else + _ -> {:reject, "[DNSRBLPolicy]"} + end + end + + @impl true + def filter(object), do: {:ok, object} + + @impl true + def describe do + mrf_dnsrbl = + Config.get(:mrf_dnsrbl) + |> Enum.into(%{}) + + {:ok, %{mrf_dnsrbl: mrf_dnsrbl}} + end + + @impl true + def config_description do + %{ + key: :mrf_dnsrbl, + related_policy: "Pleroma.Web.ActivityPub.MRF.DNSRBLPolicy", + label: "MRF DNSRBL", + description: "DNS RealTime Blackhole Policy", + children: [ + %{ + key: :nameserver, + type: {:string}, + description: "DNSRBL Nameserver to Query (IP or hostame)", + suggestions: ["127.0.0.1"] + }, + %{ + key: :port, + type: {:string}, + description: "Nameserver port", + suggestions: ["53"] + }, + %{ + key: :zone, + type: {:string}, + description: "Root zone for querying", + suggestions: ["bl.pleroma.com"] + } + ] + } + end + + defp check_rbl(%{host: actor_host}, object) do + with false <- match?(^actor_host, Pleroma.Web.Endpoint.host()), + zone when not is_nil(zone) <- Keyword.get(Config.get([:mrf_dnsrbl]), :zone) do + query = + Enum.join([actor_host, zone], ".") + |> String.to_charlist() + + rbl_response = rblquery(query) + + if Enum.empty?(rbl_response) do + {:ok, object} + else + Task.start(fn -> + reason = rblquery(query, :txt) || "undefined" + + Logger.warning( + "DNSRBL Rejected activity from #{actor_host} for reason: #{inspect(reason)}" + ) + end) + + :error + end + else + _ -> {:ok, object} + end + end + + defp get_rblhost_ip(rblhost) do + case rblhost |> String.to_charlist() |> :inet_parse.address() do + {:ok, _} -> rblhost |> String.to_charlist() |> :inet_parse.address() + _ -> {:ok, rblhost |> String.to_charlist() |> :inet_res.lookup(:in, :a) |> Enum.random()} + end + end + + defp rblquery(query, type \\ :a) do + config = Config.get([:mrf_dnsrbl]) + + case get_rblhost_ip(config[:nameserver]) do + {:ok, rblnsip} -> + :inet_res.lookup(query, :in, type, + nameservers: [{rblnsip, config[:port]}], + timeout: @query_timeout, + retry: @query_retries + ) + + _ -> + [] + end + end +end diff --git a/lib/pleroma/web/activity_pub/object_validators/attachment_validator.ex b/lib/pleroma/web/activity_pub/object_validators/attachment_validator.ex index 72975f348..5ee9e7549 100644 --- a/lib/pleroma/web/activity_pub/object_validators/attachment_validator.ex +++ b/lib/pleroma/web/activity_pub/object_validators/attachment_validator.ex @@ -15,6 +15,7 @@ defmodule Pleroma.Web.ActivityPub.ObjectValidators.AttachmentValidator do field(:type, :string, default: "Link") field(:mediaType, ObjectValidators.MIME, default: "application/octet-stream") field(:name, :string) + field(:summary, :string) field(:blurhash, :string) embeds_many :url, UrlObjectValidator, primary_key: false do @@ -44,7 +45,7 @@ def changeset(struct, data) do |> fix_url() struct - |> cast(data, [:id, :type, :mediaType, :name, :blurhash]) + |> cast(data, [:id, :type, :mediaType, :name, :summary, :blurhash]) |> cast_embed(:url, with: &url_changeset/2, required: true) |> validate_inclusion(:type, ~w[Link Document Audio Image Video]) |> validate_required([:type, :mediaType]) diff --git a/lib/pleroma/web/api_spec/schemas/attachment.ex b/lib/pleroma/web/api_spec/schemas/attachment.ex index 2871b5f99..4104ed25c 100644 --- a/lib/pleroma/web/api_spec/schemas/attachment.ex +++ b/lib/pleroma/web/api_spec/schemas/attachment.ex @@ -50,7 +50,11 @@ defmodule Pleroma.Web.ApiSpec.Schemas.Attachment do pleroma: %Schema{ type: :object, properties: %{ - mime_type: %Schema{type: :string, description: "mime type of the attachment"} + mime_type: %Schema{type: :string, description: "mime type of the attachment"}, + name: %Schema{ + type: :string, + description: "Name of the attachment, typically the filename" + } } } }, diff --git a/lib/pleroma/web/endpoint.ex b/lib/pleroma/web/endpoint.ex index 2e2104904..fef907ace 100644 --- a/lib/pleroma/web/endpoint.ex +++ b/lib/pleroma/web/endpoint.ex @@ -38,6 +38,8 @@ defmodule Pleroma.Web.Endpoint do plug(Plug.Telemetry, event_prefix: [:phoenix, :endpoint]) + plug(Pleroma.Web.Plugs.LoggerMetadataPath) + plug(Pleroma.Web.Plugs.SetLocalePlug) plug(CORSPlug) plug(Pleroma.Web.Plugs.HTTPSecurityPlug) diff --git a/lib/pleroma/web/federator.ex b/lib/pleroma/web/federator.ex index 1f2c3835a..4b30fd21d 100644 --- a/lib/pleroma/web/federator.ex +++ b/lib/pleroma/web/federator.ex @@ -44,7 +44,7 @@ def incoming_ap_doc(%{params: params, req_headers: req_headers}) do end def incoming_ap_doc(%{"type" => "Delete"} = params) do - ReceiverWorker.enqueue("incoming_ap_doc", %{"params" => params}, priority: 3) + ReceiverWorker.enqueue("incoming_ap_doc", %{"params" => params}, priority: 3, queue: :slow) end def incoming_ap_doc(params) do diff --git a/lib/pleroma/web/mastodon_api/views/status_view.ex b/lib/pleroma/web/mastodon_api/views/status_view.ex index c945290c1..0c16749a4 100644 --- a/lib/pleroma/web/mastodon_api/views/status_view.ex +++ b/lib/pleroma/web/mastodon_api/views/status_view.ex @@ -624,6 +624,19 @@ def render("attachment.json", %{attachment: attachment}) do to_string(attachment["id"] || hash_id) end + description = + if attachment["summary"] do + HTML.strip_tags(attachment["summary"]) + else + attachment["name"] + end + + name = if attachment["summary"], do: attachment["name"] + + pleroma = + %{mime_type: media_type} + |> Maps.put_if_present(:name, name) + %{ id: attachment_id, url: href, @@ -631,8 +644,8 @@ def render("attachment.json", %{attachment: attachment}) do preview_url: href_preview, text_url: href, type: type, - description: attachment["name"], - pleroma: %{mime_type: media_type}, + description: description, + pleroma: pleroma, blurhash: attachment["blurhash"] } |> Maps.put_if_present(:meta, meta) diff --git a/lib/pleroma/web/plugs/http_security_plug.ex b/lib/pleroma/web/plugs/http_security_plug.ex index a27dcd0ab..38f6c511e 100644 --- a/lib/pleroma/web/plugs/http_security_plug.ex +++ b/lib/pleroma/web/plugs/http_security_plug.ex @@ -3,26 +3,27 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Web.Plugs.HTTPSecurityPlug do - alias Pleroma.Config import Plug.Conn require Logger + @config_impl Application.compile_env(:pleroma, [__MODULE__, :config_impl], Pleroma.Config) + def init(opts), do: opts def call(conn, _options) do - if Config.get([:http_security, :enabled]) do + if @config_impl.get([:http_security, :enabled]) do conn |> merge_resp_headers(headers()) - |> maybe_send_sts_header(Config.get([:http_security, :sts])) + |> maybe_send_sts_header(@config_impl.get([:http_security, :sts])) else conn end end def primary_frontend do - with %{"name" => frontend} <- Config.get([:frontends, :primary]), - available <- Config.get([:frontends, :available]), + with %{"name" => frontend} <- @config_impl.get([:frontends, :primary]), + available <- @config_impl.get([:frontends, :available]), %{} = primary_frontend <- Map.get(available, frontend) do {:ok, primary_frontend} end @@ -37,8 +38,8 @@ def custom_http_frontend_headers do end def headers do - referrer_policy = Config.get([:http_security, :referrer_policy]) - report_uri = Config.get([:http_security, :report_uri]) + referrer_policy = @config_impl.get([:http_security, :referrer_policy]) + report_uri = @config_impl.get([:http_security, :report_uri]) custom_http_frontend_headers = custom_http_frontend_headers() headers = [ @@ -86,10 +87,10 @@ def headers do @csp_start [Enum.join(static_csp_rules, ";") <> ";"] defp csp_string do - scheme = Config.get([Pleroma.Web.Endpoint, :url])[:scheme] + scheme = @config_impl.get([Pleroma.Web.Endpoint, :url])[:scheme] static_url = Pleroma.Web.Endpoint.static_url() websocket_url = Pleroma.Web.Endpoint.websocket_url() - report_uri = Config.get([:http_security, :report_uri]) + report_uri = @config_impl.get([:http_security, :report_uri]) img_src = "img-src 'self' data: blob:" media_src = "media-src 'self'" @@ -97,8 +98,8 @@ defp csp_string do # Strict multimedia CSP enforcement only when MediaProxy is enabled {img_src, media_src, connect_src} = - if Config.get([:media_proxy, :enabled]) && - !Config.get([:media_proxy, :proxy_opts, :redirect_on_failure]) do + if @config_impl.get([:media_proxy, :enabled]) && + !@config_impl.get([:media_proxy, :proxy_opts, :redirect_on_failure]) do sources = build_csp_multimedia_source_list() { @@ -115,17 +116,21 @@ defp csp_string do end connect_src = - if Config.get(:env) == :dev do + if @config_impl.get([:env]) == :dev do [connect_src, " http://localhost:3035/"] else connect_src end script_src = - if Config.get(:env) == :dev do - "script-src 'self' 'unsafe-eval'" + if @config_impl.get([:http_security, :allow_unsafe_eval]) do + if @config_impl.get([:env]) == :dev do + "script-src 'self' 'unsafe-eval'" + else + "script-src 'self' 'wasm-unsafe-eval'" + end else - "script-src 'self' 'wasm-unsafe-eval'" + "script-src 'self'" end report = if report_uri, do: ["report-uri ", report_uri, ";report-to csp-endpoint"] @@ -161,11 +166,11 @@ defp build_csp_param_from_whitelist(url), do: url defp build_csp_multimedia_source_list do media_proxy_whitelist = [:media_proxy, :whitelist] - |> Config.get() + |> @config_impl.get() |> build_csp_from_whitelist([]) - captcha_method = Config.get([Pleroma.Captcha, :method]) - captcha_endpoint = Config.get([captcha_method, :endpoint]) + captcha_method = @config_impl.get([Pleroma.Captcha, :method]) + captcha_endpoint = @config_impl.get([captcha_method, :endpoint]) base_endpoints = [ @@ -173,7 +178,7 @@ defp build_csp_multimedia_source_list do [Pleroma.Upload, :base_url], [Pleroma.Uploaders.S3, :public_endpoint] ] - |> Enum.map(&Config.get/1) + |> Enum.map(&@config_impl.get/1) [captcha_endpoint | base_endpoints] |> Enum.map(&build_csp_param/1) @@ -200,7 +205,7 @@ defp build_csp_param(url) when is_binary(url) do end def warn_if_disabled do - unless Config.get([:http_security, :enabled]) do + unless Pleroma.Config.get([:http_security, :enabled]) do Logger.warning(" .i;;;;i. iYcviii;vXY: @@ -245,8 +250,8 @@ def warn_if_disabled do end defp maybe_send_sts_header(conn, true) do - max_age_sts = Config.get([:http_security, :sts_max_age]) - max_age_ct = Config.get([:http_security, :ct_max_age]) + max_age_sts = @config_impl.get([:http_security, :sts_max_age]) + max_age_ct = @config_impl.get([:http_security, :ct_max_age]) merge_resp_headers(conn, [ {"strict-transport-security", "max-age=#{max_age_sts}; includeSubDomains"}, diff --git a/lib/pleroma/web/plugs/logger_metadata_path.ex b/lib/pleroma/web/plugs/logger_metadata_path.ex new file mode 100644 index 000000000..a5553cfc8 --- /dev/null +++ b/lib/pleroma/web/plugs/logger_metadata_path.ex @@ -0,0 +1,12 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2022 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.Plugs.LoggerMetadataPath do + def init(opts), do: opts + + def call(conn, _) do + Logger.metadata(path: conn.request_path) + conn + end +end diff --git a/lib/pleroma/web/plugs/logger_metadata_user.ex b/lib/pleroma/web/plugs/logger_metadata_user.ex new file mode 100644 index 000000000..6a5c0041d --- /dev/null +++ b/lib/pleroma/web/plugs/logger_metadata_user.ex @@ -0,0 +1,18 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2022 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.Plugs.LoggerMetadataUser do + alias Pleroma.User + + def init(opts), do: opts + + def call(%{assigns: %{user: user = %User{}}} = conn, _) do + Logger.metadata(user: user.nickname) + conn + end + + def call(conn, _) do + conn + end +end diff --git a/lib/pleroma/web/router.ex b/lib/pleroma/web/router.ex index 368a04df0..56c457e90 100644 --- a/lib/pleroma/web/router.ex +++ b/lib/pleroma/web/router.ex @@ -29,6 +29,7 @@ defmodule Pleroma.Web.Router do pipeline :browser do plug(:accepts, ["html"]) plug(:fetch_session) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :oauth do @@ -67,12 +68,14 @@ defmodule Pleroma.Web.Router do plug(:fetch_session) plug(:authenticate) plug(OpenApiSpex.Plug.PutApiSpec, module: Pleroma.Web.ApiSpec) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :no_auth_or_privacy_expectations_api do plug(:base_api) plug(:after_auth) plug(Pleroma.Web.Plugs.IdempotencyPlug) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end # Pipeline for app-related endpoints (no user auth checks — app-bound tokens must be supported) @@ -83,12 +86,14 @@ defmodule Pleroma.Web.Router do pipeline :api do plug(:expect_public_instance_or_user_authentication) plug(:no_auth_or_privacy_expectations_api) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :authenticated_api do plug(:expect_user_authentication) plug(:no_auth_or_privacy_expectations_api) plug(Pleroma.Web.Plugs.EnsureAuthenticatedPlug) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :admin_api do @@ -99,6 +104,7 @@ defmodule Pleroma.Web.Router do plug(Pleroma.Web.Plugs.EnsureAuthenticatedPlug) plug(Pleroma.Web.Plugs.UserIsStaffPlug) plug(Pleroma.Web.Plugs.IdempotencyPlug) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :require_admin do @@ -179,6 +185,7 @@ defmodule Pleroma.Web.Router do plug(:browser) plug(:authenticate) plug(Pleroma.Web.Plugs.EnsureUserTokenAssignsPlug) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :well_known do @@ -193,6 +200,7 @@ defmodule Pleroma.Web.Router do pipeline :pleroma_api do plug(:accepts, ["html", "json"]) plug(OpenApiSpex.Plug.PutApiSpec, module: Pleroma.Web.ApiSpec) + plug(Pleroma.Web.Plugs.LoggerMetadataUser) end pipeline :mailbox_preview do diff --git a/lib/pleroma/workers/attachments_cleanup_worker.ex b/lib/pleroma/workers/attachments_cleanup_worker.ex index 4c1764053..0b570b70b 100644 --- a/lib/pleroma/workers/attachments_cleanup_worker.ex +++ b/lib/pleroma/workers/attachments_cleanup_worker.ex @@ -8,7 +8,7 @@ defmodule Pleroma.Workers.AttachmentsCleanupWorker do alias Pleroma.Object alias Pleroma.Repo - use Pleroma.Workers.WorkerHelper, queue: "attachments_cleanup" + use Pleroma.Workers.WorkerHelper, queue: "slow" @impl Oban.Worker def perform(%Job{ diff --git a/lib/pleroma/workers/backup_worker.ex b/lib/pleroma/workers/backup_worker.ex index a485ddb4b..54ac31a3c 100644 --- a/lib/pleroma/workers/backup_worker.ex +++ b/lib/pleroma/workers/backup_worker.ex @@ -3,7 +3,7 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Workers.BackupWorker do - use Oban.Worker, queue: :backup, max_attempts: 1 + use Oban.Worker, queue: :slow, max_attempts: 1 alias Oban.Job alias Pleroma.User.Backup diff --git a/lib/pleroma/workers/cron/new_users_digest_worker.ex b/lib/pleroma/workers/cron/new_users_digest_worker.ex index 1c3e445aa..d2abb2d3b 100644 --- a/lib/pleroma/workers/cron/new_users_digest_worker.ex +++ b/lib/pleroma/workers/cron/new_users_digest_worker.ex @@ -9,7 +9,7 @@ defmodule Pleroma.Workers.Cron.NewUsersDigestWorker do import Ecto.Query - use Pleroma.Workers.WorkerHelper, queue: "mailer" + use Pleroma.Workers.WorkerHelper, queue: "background" @impl Oban.Worker def perform(_job) do diff --git a/lib/pleroma/workers/mailer_worker.ex b/lib/pleroma/workers/mailer_worker.ex index 940716558..652bf77e0 100644 --- a/lib/pleroma/workers/mailer_worker.ex +++ b/lib/pleroma/workers/mailer_worker.ex @@ -3,7 +3,7 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Workers.MailerWorker do - use Pleroma.Workers.WorkerHelper, queue: "mailer" + use Pleroma.Workers.WorkerHelper, queue: "background" @impl Oban.Worker def perform(%Job{args: %{"op" => "email", "encoded_email" => encoded_email, "config" => config}}) do diff --git a/lib/pleroma/workers/mute_expire_worker.ex b/lib/pleroma/workers/mute_expire_worker.ex index 8ce458d48..8ad287a7f 100644 --- a/lib/pleroma/workers/mute_expire_worker.ex +++ b/lib/pleroma/workers/mute_expire_worker.ex @@ -3,7 +3,7 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Workers.MuteExpireWorker do - use Pleroma.Workers.WorkerHelper, queue: "mute_expire" + use Pleroma.Workers.WorkerHelper, queue: "background" @impl Oban.Worker def perform(%Job{args: %{"op" => "unmute_user", "muter_id" => muter_id, "mutee_id" => mutee_id}}) do diff --git a/lib/pleroma/workers/poll_worker.ex b/lib/pleroma/workers/poll_worker.ex index 022d026f8..70df54193 100644 --- a/lib/pleroma/workers/poll_worker.ex +++ b/lib/pleroma/workers/poll_worker.ex @@ -6,7 +6,7 @@ defmodule Pleroma.Workers.PollWorker do @moduledoc """ Generates notifications when a poll ends. """ - use Pleroma.Workers.WorkerHelper, queue: "poll_notifications" + use Pleroma.Workers.WorkerHelper, queue: "background" alias Pleroma.Activity alias Pleroma.Notification diff --git a/lib/pleroma/workers/purge_expired_activity.ex b/lib/pleroma/workers/purge_expired_activity.ex index e554684fe..a65593b6e 100644 --- a/lib/pleroma/workers/purge_expired_activity.ex +++ b/lib/pleroma/workers/purge_expired_activity.ex @@ -7,7 +7,7 @@ defmodule Pleroma.Workers.PurgeExpiredActivity do Worker which purges expired activity. """ - use Oban.Worker, queue: :activity_expiration, max_attempts: 1, unique: [period: :infinity] + use Oban.Worker, queue: :slow, max_attempts: 1, unique: [period: :infinity] import Ecto.Query @@ -59,7 +59,7 @@ defp find_user(ap_id) do def get_expiration(id) do from(j in Oban.Job, where: j.state == "scheduled", - where: j.queue == "activity_expiration", + where: j.queue == "slow", where: fragment("?->>'activity_id' = ?", j.args, ^id) ) |> Pleroma.Repo.one() diff --git a/lib/pleroma/workers/purge_expired_filter.ex b/lib/pleroma/workers/purge_expired_filter.ex index 9114aeb7f..1f6931e4c 100644 --- a/lib/pleroma/workers/purge_expired_filter.ex +++ b/lib/pleroma/workers/purge_expired_filter.ex @@ -7,7 +7,7 @@ defmodule Pleroma.Workers.PurgeExpiredFilter do Worker which purges expired filters """ - use Oban.Worker, queue: :filter_expiration, max_attempts: 1, unique: [period: :infinity] + use Oban.Worker, queue: :background, max_attempts: 1, unique: [period: :infinity] import Ecto.Query @@ -38,7 +38,7 @@ def timeout(_job), do: :timer.seconds(5) def get_expiration(id) do from(j in Job, where: j.state == "scheduled", - where: j.queue == "filter_expiration", + where: j.queue == "background", where: fragment("?->'filter_id' = ?", j.args, ^id) ) |> Repo.one() diff --git a/lib/pleroma/workers/purge_expired_token.ex b/lib/pleroma/workers/purge_expired_token.ex index 2ccd9e80b..1854bf561 100644 --- a/lib/pleroma/workers/purge_expired_token.ex +++ b/lib/pleroma/workers/purge_expired_token.ex @@ -7,7 +7,7 @@ defmodule Pleroma.Workers.PurgeExpiredToken do Worker which purges expired OAuth tokens """ - use Oban.Worker, queue: :token_expiration, max_attempts: 1 + use Oban.Worker, queue: :background, max_attempts: 1 @spec enqueue(%{token_id: integer(), valid_until: DateTime.t(), mod: module()}) :: {:ok, Oban.Job.t()} | {:error, Ecto.Changeset.t()} diff --git a/lib/pleroma/workers/remote_fetcher_worker.ex b/lib/pleroma/workers/remote_fetcher_worker.ex index c26418483..ed04c54b2 100644 --- a/lib/pleroma/workers/remote_fetcher_worker.ex +++ b/lib/pleroma/workers/remote_fetcher_worker.ex @@ -5,7 +5,7 @@ defmodule Pleroma.Workers.RemoteFetcherWorker do alias Pleroma.Object.Fetcher - use Pleroma.Workers.WorkerHelper, queue: "remote_fetcher" + use Pleroma.Workers.WorkerHelper, queue: "background" @impl Oban.Worker def perform(%Job{args: %{"op" => "fetch_remote", "id" => id} = args}) do diff --git a/lib/pleroma/workers/rich_media_expiration_worker.ex b/lib/pleroma/workers/rich_media_expiration_worker.ex index d7ae497a7..0b74687cf 100644 --- a/lib/pleroma/workers/rich_media_expiration_worker.ex +++ b/lib/pleroma/workers/rich_media_expiration_worker.ex @@ -6,7 +6,7 @@ defmodule Pleroma.Workers.RichMediaExpirationWorker do alias Pleroma.Web.RichMedia.Card use Oban.Worker, - queue: :rich_media_expiration + queue: :background @impl Oban.Worker def perform(%Job{args: %{"url" => url} = _args}) do diff --git a/lib/pleroma/workers/scheduled_activity_worker.ex b/lib/pleroma/workers/scheduled_activity_worker.ex index 4df84d00f..ab62686f4 100644 --- a/lib/pleroma/workers/scheduled_activity_worker.ex +++ b/lib/pleroma/workers/scheduled_activity_worker.ex @@ -7,7 +7,7 @@ defmodule Pleroma.Workers.ScheduledActivityWorker do The worker to post scheduled activity. """ - use Pleroma.Workers.WorkerHelper, queue: "scheduled_activities" + use Pleroma.Workers.WorkerHelper, queue: "federator_outgoing" alias Pleroma.Repo alias Pleroma.ScheduledActivity diff --git a/priv/gettext/config_descriptions.pot b/priv/gettext/config_descriptions.pot index 4f60e1c85..b4792868b 100644 --- a/priv/gettext/config_descriptions.pot +++ b/priv/gettext/config_descriptions.pot @@ -5973,3 +5973,87 @@ msgstr "" msgctxt "config label at :pleroma-:instance > :languages" msgid "Languages" msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-:mrf_emoji" +msgid "Reject or force-unlisted emojis whose URLs or names match a keyword or [Regex](https://hexdocs.pm/elixir/Regex.html)." +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-:mrf_emoji > :federated_timeline_removal_shortcode" +msgid " A list of patterns which result in message with emojis whose shortcodes match being removed from federated timelines (a.k.a unlisted). This will apply only to statuses.\n\n Each pattern can be a string or [Regex](https://hexdocs.pm/elixir/Regex.html) in the format of `~r/PATTERN/`.\n" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-:mrf_emoji > :federated_timeline_removal_url" +msgid " A list of patterns which result in message with emojis whose URLs match being removed from federated timelines (a.k.a unlisted). This will apply only to statuses.\n\n Each pattern can be a string or [Regex](https://hexdocs.pm/elixir/Regex.html) in the format of `~r/PATTERN/`.\n" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-:mrf_emoji > :remove_shortcode" +msgid " A list of patterns which result in emoji whose shortcode matches being removed from the message. This will apply to statuses, emoji reactions, and user profiles.\n\n Each pattern can be a string or [Regex](https://hexdocs.pm/elixir/Regex.html) in the format of `~r/PATTERN/`.\n" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-:mrf_emoji > :remove_url" +msgid " A list of patterns which result in emoji whose URL matches being removed from the message. This will apply to statuses, emoji reactions, and user profiles.\n\n Each pattern can be a string or [Regex](https://hexdocs.pm/elixir/Regex.html) in the format of `~r/PATTERN/`.\n" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-Pleroma.User.Backup > :process_chunk_size" +msgid "The number of activities to fetch in the backup job for each chunk." +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config description at :pleroma-Pleroma.User.Backup > :process_wait_time" +msgid "The amount of time to wait for backup to report progress, in milliseconds. If no progress is received from the backup job for that much time, terminate it and deem it failed." +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-:mrf_emoji" +msgid "MRF Emoji" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-:mrf_emoji > :federated_timeline_removal_shortcode" +msgid "Federated timeline removal shortcode" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-:mrf_emoji > :federated_timeline_removal_url" +msgid "Federated timeline removal url" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-:mrf_emoji > :remove_shortcode" +msgid "Remove shortcode" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-:mrf_emoji > :remove_url" +msgid "Remove url" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-Pleroma.User.Backup > :process_chunk_size" +msgid "Process Chunk Size" +msgstr "" + +#: lib/pleroma/docs/translator.ex:5 +#, elixir-autogen, elixir-format +msgctxt "config label at :pleroma-Pleroma.User.Backup > :process_wait_time" +msgid "Process Wait Time" +msgstr "" diff --git a/priv/gettext/errors.pot b/priv/gettext/errors.pot index d320ee1bd..aca77f8fa 100644 --- a/priv/gettext/errors.pot +++ b/priv/gettext/errors.pot @@ -110,7 +110,7 @@ msgstr "" msgid "Can't display this activity" msgstr "" -#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:334 +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:346 #, elixir-autogen, elixir-format msgid "Can't find user" msgstr "" @@ -198,7 +198,7 @@ msgstr "" msgid "Invalid password." msgstr "" -#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:267 +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:279 #, elixir-autogen, elixir-format msgid "Invalid request" msgstr "" @@ -225,7 +225,7 @@ msgstr "" #: lib/pleroma/web/feed/tag_controller.ex:16 #: lib/pleroma/web/feed/user_controller.ex:69 #: lib/pleroma/web/o_status/o_status_controller.ex:132 -#: lib/pleroma/web/plugs/uploaded_media.ex:104 +#: lib/pleroma/web/plugs/uploaded_media.ex:84 #, elixir-autogen, elixir-format msgid "Not found" msgstr "" @@ -235,7 +235,7 @@ msgstr "" msgid "Poll's author can't vote" msgstr "" -#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:499 +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:511 #: lib/pleroma/web/mastodon_api/controllers/fallback_controller.ex:20 #: lib/pleroma/web/mastodon_api/controllers/poll_controller.ex:39 #: lib/pleroma/web/mastodon_api/controllers/poll_controller.ex:51 @@ -341,7 +341,7 @@ msgstr "" msgid "CAPTCHA expired" msgstr "" -#: lib/pleroma/web/plugs/uploaded_media.ex:77 +#: lib/pleroma/web/plugs/uploaded_media.ex:57 #, elixir-autogen, elixir-format msgid "Failed" msgstr "" @@ -361,7 +361,7 @@ msgstr "" msgid "Insufficient permissions: %{permissions}." msgstr "" -#: lib/pleroma/web/plugs/uploaded_media.ex:131 +#: lib/pleroma/web/plugs/uploaded_media.ex:111 #, elixir-autogen, elixir-format msgid "Internal Error" msgstr "" @@ -557,7 +557,7 @@ msgstr "" msgid "Access denied" msgstr "" -#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:331 +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:343 #, elixir-autogen, elixir-format msgid "This API requires an authenticated user" msgstr "" @@ -567,7 +567,7 @@ msgstr "" msgid "User is not an admin." msgstr "" -#: lib/pleroma/user/backup.ex:73 +#: lib/pleroma/user/backup.ex:78 #, elixir-format msgid "Last export was less than a day ago" msgid_plural "Last export was less than %{days} days ago" @@ -607,3 +607,23 @@ msgstr "" #, elixir-autogen, elixir-format msgid "User isn't privileged." msgstr "" + +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:267 +#, elixir-autogen, elixir-format +msgid "Bio is too long" +msgstr "" + +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:270 +#, elixir-autogen, elixir-format +msgid "Name is too long" +msgstr "" + +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:273 +#, elixir-autogen, elixir-format +msgid "One or more field entries are too long" +msgstr "" + +#: lib/pleroma/web/mastodon_api/controllers/account_controller.ex:276 +#, elixir-autogen, elixir-format +msgid "Too many field entries" +msgstr "" diff --git a/priv/gettext/oauth_scopes.pot b/priv/gettext/oauth_scopes.pot index 50ad0dd9e..83328770e 100644 --- a/priv/gettext/oauth_scopes.pot +++ b/priv/gettext/oauth_scopes.pot @@ -219,3 +219,43 @@ msgstr "" #, elixir-autogen, elixir-format msgid "read:mutes" msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "push" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "read:backups" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "read:chats" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "read:media" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "read:reports" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "write:chats" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "write:follow" +msgstr "" + +#: lib/pleroma/web/api_spec/scopes/translator.ex:5 +#, elixir-autogen, elixir-format +msgid "write:reports" +msgstr "" diff --git a/priv/repo/migrations/20240527144418_oban_queues_refactor.exs b/priv/repo/migrations/20240527144418_oban_queues_refactor.exs new file mode 100644 index 000000000..64ee28dfd --- /dev/null +++ b/priv/repo/migrations/20240527144418_oban_queues_refactor.exs @@ -0,0 +1,32 @@ +defmodule Pleroma.Repo.Migrations.ObanQueuesRefactor do + use Ecto.Migration + + @changed_queues [ + {"attachments_cleanup", "slow"}, + {"mailer", "background"}, + {"mute_expire", "background"}, + {"poll_notifications", "background"}, + {"activity_expiration", "slow"}, + {"filter_expiration", "background"}, + {"token_expiration", "background"}, + {"remote_fetcher", "background"}, + {"rich_media_expiration", "background"} + ] + + def up do + Enum.each(@changed_queues, fn {old, new} -> + execute("UPDATE oban_jobs SET queue = '#{new}' WHERE queue = '#{old}';") + end) + + # Handled special as reverting this would not be ideal and leaving it is harmless + execute( + "UPDATE oban_jobs SET queue = 'federator_outgoing' WHERE queue = 'scheduled_activities';" + ) + end + + def down do + # Just move all slow queue jobs to background queue if we are reverting + # as the slow queue will not be processing jobs + execute("UPDATE oban_jobs SET queue = 'background' WHERE queue = 'slow';") + end +end diff --git a/supplemental/search/fastembed-api/Dockerfile b/supplemental/search/fastembed-api/Dockerfile new file mode 100644 index 000000000..c1e0ef51f --- /dev/null +++ b/supplemental/search/fastembed-api/Dockerfile @@ -0,0 +1,9 @@ +FROM python:3.9 + +WORKDIR /code +COPY fastembed-server.py /workdir/fastembed-server.py +COPY requirements.txt /workdir/requirements.txt + +RUN pip install -r /workdir/requirements.txt + +CMD ["python", "/workdir/fastembed-server.py"] diff --git a/supplemental/search/fastembed-api/README.md b/supplemental/search/fastembed-api/README.md new file mode 100644 index 000000000..63a037207 --- /dev/null +++ b/supplemental/search/fastembed-api/README.md @@ -0,0 +1,6 @@ +# About +This is a minimal implementation of the [OpenAI Embeddings API](https://platform.openai.com/docs/guides/embeddings/what-are-embeddings) meant to be used with the QdrantSearch backend. + +# Usage + +The easiest way to run it is to just use docker compose with `docker compose up`. This starts the server on the default configured port. Different models can be used, for a full list of supported models, check the [fastembed documentation](https://qdrant.github.io/fastembed/examples/Supported_Models/). The first time a model is requested it will be downloaded, which can take a few seconds. diff --git a/supplemental/search/fastembed-api/compose.yml b/supplemental/search/fastembed-api/compose.yml new file mode 100644 index 000000000..d4cb31722 --- /dev/null +++ b/supplemental/search/fastembed-api/compose.yml @@ -0,0 +1,5 @@ +services: + web: + build: . + ports: + - "11345:11345" diff --git a/supplemental/search/fastembed-api/fastembed-server.py b/supplemental/search/fastembed-api/fastembed-server.py new file mode 100644 index 000000000..02da69db2 --- /dev/null +++ b/supplemental/search/fastembed-api/fastembed-server.py @@ -0,0 +1,27 @@ +from fastembed import TextEmbedding +from fastapi import FastAPI +from pydantic import BaseModel + +models = {} + +app = FastAPI() + +class EmbeddingRequest(BaseModel): + model: str + input: str + +@app.post("/v1/embeddings") +def embeddings(request: EmbeddingRequest): + model = models.get(request.model) or TextEmbedding(request.model) + models[request.model] = model + embeddings = next(model.embed(request.input)).tolist() + return {"data": [{"embedding": embeddings}]} + +@app.get("/health") +def health(): + return {"status": "ok"} + +if __name__ == "__main__": + import uvicorn + + uvicorn.run(app, host="0.0.0.0", port=11345) diff --git a/supplemental/search/fastembed-api/requirements.txt b/supplemental/search/fastembed-api/requirements.txt new file mode 100644 index 000000000..db67a8402 --- /dev/null +++ b/supplemental/search/fastembed-api/requirements.txt @@ -0,0 +1,4 @@ +fastapi==0.111.0 +fastembed==0.2.7 +pydantic==1.10.15 +uvicorn==0.29.0 diff --git a/test/pleroma/scheduled_activity_test.exs b/test/pleroma/scheduled_activity_test.exs index 4818e8bcf..aaf643cfc 100644 --- a/test/pleroma/scheduled_activity_test.exs +++ b/test/pleroma/scheduled_activity_test.exs @@ -31,8 +31,7 @@ test "scheduled activities with jobs when ScheduledActivity enabled" do {:ok, sa1} = ScheduledActivity.create(user, attrs) {:ok, sa2} = ScheduledActivity.create(user, attrs) - jobs = - Repo.all(from(j in Oban.Job, where: j.queue == "scheduled_activities", select: j.args)) + jobs = Repo.all(from(j in Oban.Job, where: j.queue == "federator_outgoing", select: j.args)) assert jobs == [%{"activity_id" => sa1.id}, %{"activity_id" => sa2.id}] end diff --git a/test/pleroma/search/qdrant_search_test.exs b/test/pleroma/search/qdrant_search_test.exs new file mode 100644 index 000000000..47a77a391 --- /dev/null +++ b/test/pleroma/search/qdrant_search_test.exs @@ -0,0 +1,199 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2021 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Search.QdrantSearchTest do + use Pleroma.DataCase, async: true + use Oban.Testing, repo: Pleroma.Repo + + import Pleroma.Factory + import Mox + + alias Pleroma.Search.QdrantSearch + alias Pleroma.UnstubbedConfigMock, as: Config + alias Pleroma.Web.CommonAPI + alias Pleroma.Workers.SearchIndexingWorker + + describe "Qdrant search" do + test "returns the correct healthcheck endpoints" do + # No openai healthcheck URL + Config + |> expect(:get, 2, fn + [Pleroma.Search.QdrantSearch, key], nil -> + %{qdrant_url: "https://qdrant.url"}[key] + end) + + [health_endpoint] = QdrantSearch.healthcheck_endpoints() + + assert "https://qdrant.url/healthz" == health_endpoint + + # Set openai healthcheck URL + Config + |> expect(:get, 2, fn + [Pleroma.Search.QdrantSearch, key], nil -> + %{qdrant_url: "https://qdrant.url", openai_healthcheck_url: "https://openai.url/health"}[ + key + ] + end) + + [_, health_endpoint] = QdrantSearch.healthcheck_endpoints() + + assert "https://openai.url/health" == health_endpoint + end + + test "searches for a term by encoding it and sending it to qdrant" do + user = insert(:user) + + {:ok, activity} = + CommonAPI.post(user, %{ + status: "guys i just don't wanna leave the swamp", + visibility: "public" + }) + + Config + |> expect(:get, 3, fn + [Pleroma.Search, :module], nil -> + QdrantSearch + + [Pleroma.Search.QdrantSearch, key], nil -> + %{ + openai_model: "a_model", + openai_url: "https://openai.url", + qdrant_url: "https://qdrant.url" + }[key] + end) + + Tesla.Mock.mock(fn + %{url: "https://openai.url/v1/embeddings", method: :post} -> + Tesla.Mock.json(%{ + data: [%{embedding: [1, 2, 3]}] + }) + + %{url: "https://qdrant.url/collections/posts/points/search", method: :post, body: body} -> + data = Jason.decode!(body) + refute data["filter"] + + Tesla.Mock.json(%{ + result: [%{"id" => activity.id |> FlakeId.from_string() |> Ecto.UUID.cast!()}] + }) + end) + + results = QdrantSearch.search(nil, "guys i just don't wanna leave the swamp", %{}) + + assert results == [activity] + end + + test "for a given actor, ask for only relevant matches" do + user = insert(:user) + + {:ok, activity} = + CommonAPI.post(user, %{ + status: "guys i just don't wanna leave the swamp", + visibility: "public" + }) + + Config + |> expect(:get, 3, fn + [Pleroma.Search, :module], nil -> + QdrantSearch + + [Pleroma.Search.QdrantSearch, key], nil -> + %{ + openai_model: "a_model", + openai_url: "https://openai.url", + qdrant_url: "https://qdrant.url" + }[key] + end) + + Tesla.Mock.mock(fn + %{url: "https://openai.url/v1/embeddings", method: :post} -> + Tesla.Mock.json(%{ + data: [%{embedding: [1, 2, 3]}] + }) + + %{url: "https://qdrant.url/collections/posts/points/search", method: :post, body: body} -> + data = Jason.decode!(body) + + assert data["filter"] == %{ + "must" => [%{"key" => "actor", "match" => %{"value" => user.ap_id}}] + } + + Tesla.Mock.json(%{ + result: [%{"id" => activity.id |> FlakeId.from_string() |> Ecto.UUID.cast!()}] + }) + end) + + results = + QdrantSearch.search(nil, "guys i just don't wanna leave the swamp", %{author: user}) + + assert results == [activity] + end + + test "indexes a public post on creation, deletes from the index on deletion" do + user = insert(:user) + + Tesla.Mock.mock(fn + %{method: :post, url: "https://openai.url/v1/embeddings"} -> + send(self(), "posted_to_openai") + + Tesla.Mock.json(%{ + data: [%{embedding: [1, 2, 3]}] + }) + + %{method: :put, url: "https://qdrant.url/collections/posts/points", body: body} -> + send(self(), "posted_to_qdrant") + + data = Jason.decode!(body) + %{"points" => [%{"vector" => vector, "payload" => payload}]} = data + + assert vector == [1, 2, 3] + assert payload["actor"] + assert payload["published_at"] + + Tesla.Mock.json("ok") + + %{method: :post, url: "https://qdrant.url/collections/posts/points/delete"} -> + send(self(), "deleted_from_qdrant") + Tesla.Mock.json("ok") + end) + + Config + |> expect(:get, 6, fn + [Pleroma.Search, :module], nil -> + QdrantSearch + + [Pleroma.Search.QdrantSearch, key], nil -> + %{ + openai_model: "a_model", + openai_url: "https://openai.url", + qdrant_url: "https://qdrant.url" + }[key] + end) + + {:ok, activity} = + CommonAPI.post(user, %{ + status: "guys i just don't wanna leave the swamp", + visibility: "public" + }) + + args = %{"op" => "add_to_index", "activity" => activity.id} + + assert_enqueued( + worker: SearchIndexingWorker, + args: args + ) + + assert :ok = perform_job(SearchIndexingWorker, args) + assert_received("posted_to_openai") + assert_received("posted_to_qdrant") + + {:ok, _} = CommonAPI.delete(activity.id, user) + + delete_args = %{"op" => "remove_from_index", "object" => activity.object.id} + assert_enqueued(worker: SearchIndexingWorker, args: delete_args) + assert :ok = perform_job(SearchIndexingWorker, delete_args) + + assert_received("deleted_from_qdrant") + end + end +end diff --git a/test/pleroma/web/activity_pub/mrf/anti_mention_spam_policy_test.exs b/test/pleroma/web/activity_pub/mrf/anti_mention_spam_policy_test.exs new file mode 100644 index 000000000..63947858c --- /dev/null +++ b/test/pleroma/web/activity_pub/mrf/anti_mention_spam_policy_test.exs @@ -0,0 +1,65 @@ +# Pleroma: A lightweight social networking server +# Copyright © 2017-2022 Pleroma Authors +# SPDX-License-Identifier: AGPL-3.0-only + +defmodule Pleroma.Web.ActivityPub.MRF.AntiMentionSpamPolicyTest do + use Pleroma.DataCase + import Pleroma.Factory + alias Pleroma.Web.ActivityPub.MRF.AntiMentionSpamPolicy + + test "it allows posts without mentions" do + user = insert(:user, local: false) + assert user.note_count == 0 + + message = %{ + "type" => "Create", + "actor" => user.ap_id + } + + {:ok, _message} = AntiMentionSpamPolicy.filter(message) + end + + test "it allows posts from users with followers, posts, and age" do + user = + insert(:user, + local: false, + follower_count: 1, + note_count: 1, + inserted_at: ~N[1970-01-01 00:00:00] + ) + + message = %{ + "type" => "Create", + "actor" => user.ap_id + } + + {:ok, _message} = AntiMentionSpamPolicy.filter(message) + end + + test "it allows posts from local users" do + user = insert(:user, local: true) + + message = %{ + "type" => "Create", + "actor" => user.ap_id + } + + {:ok, _message} = AntiMentionSpamPolicy.filter(message) + end + + test "it rejects posts with mentions from users without followers" do + user = insert(:user, local: false, follower_count: 0) + + message = %{ + "type" => "Create", + "actor" => user.ap_id, + "object" => %{ + "to" => ["https://pleroma.soykaf.com/users/1"], + "cc" => ["https://pleroma.soykaf.com/users/1"], + "actor" => user.ap_id + } + } + + {:reject, _message} = AntiMentionSpamPolicy.filter(message) + end +end diff --git a/test/pleroma/web/activity_pub/object_validators/attachment_validator_test.exs b/test/pleroma/web/activity_pub/object_validators/attachment_validator_test.exs index a615c1d9a..6627fa6db 100644 --- a/test/pleroma/web/activity_pub/object_validators/attachment_validator_test.exs +++ b/test/pleroma/web/activity_pub/object_validators/attachment_validator_test.exs @@ -27,19 +27,22 @@ test "fails without url" do end test "works with honkerific attachments" do - attachment = %{ + honk = %{ "mediaType" => "", - "name" => "", - "summary" => "298p3RG7j27tfsZ9RQ.jpg", + "summary" => "Select your spirit chonk", + "name" => "298p3RG7j27tfsZ9RQ.jpg", "type" => "Document", "url" => "https://honk.tedunangst.com/d/298p3RG7j27tfsZ9RQ.jpg" } assert {:ok, attachment} = - AttachmentValidator.cast_and_validate(attachment) + honk + |> AttachmentValidator.cast_and_validate() |> Ecto.Changeset.apply_action(:insert) assert attachment.mediaType == "application/octet-stream" + assert attachment.summary == "Select your spirit chonk" + assert attachment.name == "298p3RG7j27tfsZ9RQ.jpg" end test "works with an unknown but valid mime type" do diff --git a/test/pleroma/web/mastodon_api/controllers/scheduled_activity_controller_test.exs b/test/pleroma/web/mastodon_api/controllers/scheduled_activity_controller_test.exs index 632242221..2d6b2aee2 100644 --- a/test/pleroma/web/mastodon_api/controllers/scheduled_activity_controller_test.exs +++ b/test/pleroma/web/mastodon_api/controllers/scheduled_activity_controller_test.exs @@ -3,6 +3,7 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Web.MastodonAPI.ScheduledActivityControllerTest do + use Oban.Testing, repo: Pleroma.Repo use Pleroma.Web.ConnCase, async: true alias Pleroma.Repo @@ -78,7 +79,7 @@ test "updates a scheduled activity" do } ) - job = Repo.one(from(j in Oban.Job, where: j.queue == "scheduled_activities")) + job = Repo.one(from(j in Oban.Job, where: j.queue == "federator_outgoing")) assert job.args == %{"activity_id" => scheduled_activity.id} assert DateTime.truncate(job.scheduled_at, :second) == to_datetime(scheduled_at) @@ -124,9 +125,11 @@ test "deletes a scheduled activity" do } ) - job = Repo.one(from(j in Oban.Job, where: j.queue == "scheduled_activities")) - - assert job.args == %{"activity_id" => scheduled_activity.id} + assert_enqueued( + worker: Pleroma.Workers.ScheduledActivityWorker, + args: %{"activity_id" => scheduled_activity.id}, + queue: :federator_outgoing + ) res_conn = conn @@ -135,7 +138,11 @@ test "deletes a scheduled activity" do assert %{} = json_response_and_validate_schema(res_conn, 200) refute Repo.get(ScheduledActivity, scheduled_activity.id) - refute Repo.get(Oban.Job, job.id) + + refute_enqueued( + worker: Pleroma.Workers.ScheduledActivityWorker, + args: %{"activity_id" => scheduled_activity.id} + ) res_conn = conn diff --git a/test/pleroma/web/mastodon_api/views/status_view_test.exs b/test/pleroma/web/mastodon_api/views/status_view_test.exs index 1c2d7f7fd..167692dfb 100644 --- a/test/pleroma/web/mastodon_api/views/status_view_test.exs +++ b/test/pleroma/web/mastodon_api/views/status_view_test.exs @@ -591,45 +591,78 @@ test "create mentions from the 'tag' field" do assert mention.url == recipient.ap_id end - test "attachments" do - object = %{ - "type" => "Image", - "url" => [ - %{ - "mediaType" => "image/png", - "href" => "someurl", - "width" => 200, - "height" => 100 - } - ], - "blurhash" => "UJJ8X[xYW,%Jtq%NNFbXB5j]IVM|9GV=WHRn", - "uuid" => 6 - } + describe "attachments" do + test "Complete Mastodon style" do + object = %{ + "type" => "Image", + "url" => [ + %{ + "mediaType" => "image/png", + "href" => "someurl", + "width" => 200, + "height" => 100 + } + ], + "blurhash" => "UJJ8X[xYW,%Jtq%NNFbXB5j]IVM|9GV=WHRn", + "uuid" => 6 + } - expected = %{ - id: "1638338801", - type: "image", - url: "someurl", - remote_url: "someurl", - preview_url: "someurl", - text_url: "someurl", - description: nil, - pleroma: %{mime_type: "image/png"}, - meta: %{original: %{width: 200, height: 100, aspect: 2}}, - blurhash: "UJJ8X[xYW,%Jtq%NNFbXB5j]IVM|9GV=WHRn" - } + expected = %{ + id: "1638338801", + type: "image", + url: "someurl", + remote_url: "someurl", + preview_url: "someurl", + text_url: "someurl", + description: nil, + pleroma: %{mime_type: "image/png"}, + meta: %{original: %{width: 200, height: 100, aspect: 2}}, + blurhash: "UJJ8X[xYW,%Jtq%NNFbXB5j]IVM|9GV=WHRn" + } - api_spec = Pleroma.Web.ApiSpec.spec() + api_spec = Pleroma.Web.ApiSpec.spec() - assert expected == StatusView.render("attachment.json", %{attachment: object}) - assert_schema(expected, "Attachment", api_spec) + assert expected == StatusView.render("attachment.json", %{attachment: object}) + assert_schema(expected, "Attachment", api_spec) - # If theres a "id", use that instead of the generated one - object = Map.put(object, "id", 2) - result = StatusView.render("attachment.json", %{attachment: object}) + # If theres a "id", use that instead of the generated one + object = Map.put(object, "id", 2) + result = StatusView.render("attachment.json", %{attachment: object}) - assert %{id: "2"} = result - assert_schema(result, "Attachment", api_spec) + assert %{id: "2"} = result + assert_schema(result, "Attachment", api_spec) + end + + test "Honkerific" do + object = %{ + "type" => "Image", + "url" => [ + %{ + "mediaType" => "image/png", + "href" => "someurl" + } + ], + "name" => "fool.jpeg", + "summary" => "they have played us for absolute fools." + } + + expected = %{ + blurhash: nil, + description: "they have played us for absolute fools.", + id: "1638338801", + pleroma: %{mime_type: "image/png", name: "fool.jpeg"}, + preview_url: "someurl", + remote_url: "someurl", + text_url: "someurl", + type: "image", + url: "someurl" + } + + api_spec = Pleroma.Web.ApiSpec.spec() + + assert expected == StatusView.render("attachment.json", %{attachment: object}) + assert_schema(expected, "Attachment", api_spec) + end end test "put the url advertised in the Activity in to the url attribute" do diff --git a/test/pleroma/web/plugs/http_security_plug_test.exs b/test/pleroma/web/plugs/http_security_plug_test.exs index c79170382..11a351a41 100644 --- a/test/pleroma/web/plugs/http_security_plug_test.exs +++ b/test/pleroma/web/plugs/http_security_plug_test.exs @@ -3,14 +3,52 @@ # SPDX-License-Identifier: AGPL-3.0-only defmodule Pleroma.Web.Plugs.HTTPSecurityPlugTest do - use Pleroma.Web.ConnCase + use Pleroma.Web.ConnCase, async: true alias Plug.Conn - describe "http security enabled" do - setup do: clear_config([:http_security, :enabled], true) + import Mox - test "it sends CSP headers when enabled", %{conn: conn} do + setup do + base_config = Pleroma.Config.get([:http_security]) + %{base_config: base_config} + end + + defp mock_config(config, additional \\ %{}) do + Pleroma.StaticStubbedConfigMock + |> stub(:get, fn + [:http_security, key] -> config[key] + key -> additional[key] + end) + end + + describe "http security enabled" do + setup %{base_config: base_config} do + %{base_config: Keyword.put(base_config, :enabled, true)} + end + + test "it does not contain unsafe-eval", %{conn: conn, base_config: base_config} do + mock_config(base_config) + + conn = get(conn, "/api/v1/instance") + [header] = Conn.get_resp_header(conn, "content-security-policy") + refute header =~ ~r/unsafe-eval/ + end + + test "with allow_unsafe_eval set, it does contain it", %{conn: conn, base_config: base_config} do + base_config = + base_config + |> Keyword.put(:allow_unsafe_eval, true) + + mock_config(base_config) + + conn = get(conn, "/api/v1/instance") + [header] = Conn.get_resp_header(conn, "content-security-policy") + assert header =~ ~r/unsafe-eval/ + end + + test "it sends CSP headers when enabled", %{conn: conn, base_config: base_config} do + mock_config(base_config) conn = get(conn, "/api/v1/instance") refute Conn.get_resp_header(conn, "x-xss-protection") == [] @@ -22,8 +60,10 @@ test "it sends CSP headers when enabled", %{conn: conn} do refute Conn.get_resp_header(conn, "content-security-policy") == [] end - test "it sends STS headers when enabled", %{conn: conn} do - clear_config([:http_security, :sts], true) + test "it sends STS headers when enabled", %{conn: conn, base_config: base_config} do + base_config + |> Keyword.put(:sts, true) + |> mock_config() conn = get(conn, "/api/v1/instance") @@ -31,8 +71,10 @@ test "it sends STS headers when enabled", %{conn: conn} do refute Conn.get_resp_header(conn, "expect-ct") == [] end - test "it does not send STS headers when disabled", %{conn: conn} do - clear_config([:http_security, :sts], false) + test "it does not send STS headers when disabled", %{conn: conn, base_config: base_config} do + base_config + |> Keyword.put(:sts, false) + |> mock_config() conn = get(conn, "/api/v1/instance") @@ -40,19 +82,30 @@ test "it does not send STS headers when disabled", %{conn: conn} do assert Conn.get_resp_header(conn, "expect-ct") == [] end - test "referrer-policy header reflects configured value", %{conn: conn} do - resp = get(conn, "/api/v1/instance") + test "referrer-policy header reflects configured value", %{ + conn: conn, + base_config: base_config + } do + mock_config(base_config) + resp = get(conn, "/api/v1/instance") assert Conn.get_resp_header(resp, "referrer-policy") == ["same-origin"] - clear_config([:http_security, :referrer_policy], "no-referrer") + base_config + |> Keyword.put(:referrer_policy, "no-referrer") + |> mock_config resp = get(conn, "/api/v1/instance") assert Conn.get_resp_header(resp, "referrer-policy") == ["no-referrer"] end - test "it sends `report-to` & `report-uri` CSP response headers", %{conn: conn} do + test "it sends `report-to` & `report-uri` CSP response headers", %{ + conn: conn, + base_config: base_config + } do + mock_config(base_config) + conn = get(conn, "/api/v1/instance") [csp] = Conn.get_resp_header(conn, "content-security-policy") @@ -65,7 +118,11 @@ test "it sends `report-to` & `report-uri` CSP response headers", %{conn: conn} d "{\"endpoints\":[{\"url\":\"https://endpoint.com\"}],\"group\":\"csp-endpoint\",\"max-age\":10886400}" end - test "default values for img-src and media-src with disabled media proxy", %{conn: conn} do + test "default values for img-src and media-src with disabled media proxy", %{ + conn: conn, + base_config: base_config + } do + mock_config(base_config) conn = get(conn, "/api/v1/instance") [csp] = Conn.get_resp_header(conn, "content-security-policy") @@ -73,60 +130,129 @@ test "default values for img-src and media-src with disabled media proxy", %{con assert csp =~ "img-src 'self' data: blob: https:;" end - test "it sets the Service-Worker-Allowed header", %{conn: conn} do - clear_config([:http_security, :enabled], true) - clear_config([:frontends, :primary], %{"name" => "fedi-fe", "ref" => "develop"}) + test "it sets the Service-Worker-Allowed header", %{conn: conn, base_config: base_config} do + base_config + |> Keyword.put(:enabled, true) - clear_config([:frontends, :available], %{ - "fedi-fe" => %{ - "name" => "fedi-fe", - "custom-http-headers" => [{"service-worker-allowed", "/"}] - } - }) + additional_config = + %{} + |> Map.put([:frontends, :primary], %{"name" => "fedi-fe", "ref" => "develop"}) + |> Map.put( + [:frontends, :available], + %{ + "fedi-fe" => %{ + "name" => "fedi-fe", + "custom-http-headers" => [{"service-worker-allowed", "/"}] + } + } + ) + mock_config(base_config, additional_config) conn = get(conn, "/api/v1/instance") assert Conn.get_resp_header(conn, "service-worker-allowed") == ["/"] end end describe "img-src and media-src" do - setup do - clear_config([:http_security, :enabled], true) - clear_config([:media_proxy, :enabled], true) - clear_config([:media_proxy, :proxy_opts, :redirect_on_failure], false) + setup %{base_config: base_config} do + base_config = + base_config + |> Keyword.put(:enabled, true) + + additional_config = + %{} + |> Map.put([:media_proxy, :enabled], true) + |> Map.put([:media_proxy, :proxy_opts, :redirect_on_failure], false) + |> Map.put([:media_proxy, :whitelist], []) + + %{base_config: base_config, additional_config: additional_config} end - test "media_proxy with base_url", %{conn: conn} do + test "media_proxy with base_url", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do url = "https://example.com" - clear_config([:media_proxy, :base_url], url) + + additional_config = + additional_config + |> Map.put([:media_proxy, :base_url], url) + + mock_config(base_config, additional_config) + assert_media_img_src(conn, url) end - test "upload with base url", %{conn: conn} do + test "upload with base url", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do url = "https://example2.com" - clear_config([Pleroma.Upload, :base_url], url) + + additional_config = + additional_config + |> Map.put([Pleroma.Upload, :base_url], url) + + mock_config(base_config, additional_config) + assert_media_img_src(conn, url) end - test "with S3 public endpoint", %{conn: conn} do + test "with S3 public endpoint", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do url = "https://example3.com" - clear_config([Pleroma.Uploaders.S3, :public_endpoint], url) + + additional_config = + additional_config + |> Map.put([Pleroma.Uploaders.S3, :public_endpoint], url) + + mock_config(base_config, additional_config) assert_media_img_src(conn, url) end - test "with captcha endpoint", %{conn: conn} do - clear_config([Pleroma.Captcha.Mock, :endpoint], "https://captcha.com") + test "with captcha endpoint", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do + additional_config = + additional_config + |> Map.put([Pleroma.Captcha.Mock, :endpoint], "https://captcha.com") + |> Map.put([Pleroma.Captcha, :method], Pleroma.Captcha.Mock) + + mock_config(base_config, additional_config) assert_media_img_src(conn, "https://captcha.com") end - test "with media_proxy whitelist", %{conn: conn} do - clear_config([:media_proxy, :whitelist], ["https://example6.com", "https://example7.com"]) + test "with media_proxy whitelist", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do + additional_config = + additional_config + |> Map.put([:media_proxy, :whitelist], ["https://example6.com", "https://example7.com"]) + + mock_config(base_config, additional_config) assert_media_img_src(conn, "https://example7.com https://example6.com") end # TODO: delete after removing support bare domains for media proxy whitelist - test "with media_proxy bare domains whitelist (deprecated)", %{conn: conn} do - clear_config([:media_proxy, :whitelist], ["example4.com", "example5.com"]) + test "with media_proxy bare domains whitelist (deprecated)", %{ + conn: conn, + base_config: base_config, + additional_config: additional_config + } do + additional_config = + additional_config + |> Map.put([:media_proxy, :whitelist], ["example4.com", "example5.com"]) + + mock_config(base_config, additional_config) assert_media_img_src(conn, "example5.com example4.com") end end @@ -138,8 +264,10 @@ defp assert_media_img_src(conn, url) do assert csp =~ "img-src 'self' data: blob: #{url};" end - test "it does not send CSP headers when disabled", %{conn: conn} do - clear_config([:http_security, :enabled], false) + test "it does not send CSP headers when disabled", %{conn: conn, base_config: base_config} do + base_config + |> Keyword.put(:enabled, false) + |> mock_config conn = get(conn, "/api/v1/instance")