Pleroma.MIME: use gen_magic

.gitlab-ci.yml

@@ -24,6 +24,8 @@ stages:
 before_script:
   - mix local.hex --force
   - mix local.rebar --force
+  - apt-get -qq update
+  - apt-get install -y libmagic-dev
 
 build:
   stage: build

CHANGELOG.md

@@ -8,6 +8,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 ### Changed
 - MFR policy to set global expiration for all local Create activities
 - OGP rich media parser merged with TwitterCard
+- **Breaking** Requires `libmagic` (or `file`) to guess file types.
 <details>
   <summary>API Changes</summary>
 - **Breaking:** Emoji API: changed methods and renamed routes.

config/config.exs

@@ -678,6 +678,8 @@
     timeout: 300_000
   ]
 
+config :pleroma, :gen_magic_pool, size: 2
+
 config :pleroma, :restrict_unauthenticated,
   timelines: %{local: false, federated: false},
   profiles: %{local: false, remote: false},

config/description.exs

@@ -3325,5 +3325,19 @@
         suggestions: [false]
       }
     ]
+  },
+  %{
+    group: :pleroma,
+    key: :gen_magic_pool,
+    type: :group,
+    description: "GenMagic/libmagic configuration",
+    children: [
+      %{
+        key: :size,
+        type: :integer,
+        description: "Number of gen_magic workers to start.",
+        suggestions: [2]
+      }
+    ]
   }
 ]

docs/installation/alpine_linux_en.md

@@ -13,6 +13,7 @@ It assumes that you have administrative rights, either as root or a user with [s
 * `erlang-parsetools`
 * `erlang-xmerl`
 * `git`
+* `file-dev`
 * Development Tools
 
 #### Optional packages used in this guide

docs/installation/arch_linux_en.md

@@ -9,6 +9,7 @@ This guide will assume that you have administrative rights, either as root or a
 * `elixir`
 * `git`
 * `base-devel`
+* `file`
 
 #### Optional packages used in this guide
 

docs/installation/debian_based_en.md

@@ -10,6 +10,7 @@ This guide will assume you are on Debian Stretch. This guide should also work wi
 * `elixir` (1.8+, Follow the guide to install from the Erlang Solutions repo or use [asdf](https://github.com/asdf-vm/asdf) as the pleroma user)
 * `erlang-dev`
 * `erlang-nox`
+* `libmagic-dev`
 * `git`
 * `build-essential`
 

docs/installation/gentoo_en.md

@@ -28,6 +28,7 @@ Gentoo quite pointedly does not come with a cron daemon installed, and as such i
 * `dev-db/postgresql`
 * `dev-lang/elixir`
 * `dev-vcs/git`
+* `sys-apps/file`
 
 #### Optional ebuilds used in this guide
 

docs/installation/otp_en.md

@@ -27,15 +27,16 @@ Other than things bundled in the OTP release Pleroma depends on:
 * PostgreSQL (also utilizes extensions in postgresql-contrib)
 * nginx (could be swapped with another reverse proxy but this guide covers only it)
 * certbot (for Let's Encrypt certificates, could be swapped with another ACME client, but this guide covers only it)
+* libmagic/file
 
 ```sh tab="Alpine"
 echo "http://nl.alpinelinux.org/alpine/latest-stable/community" >> /etc/apk/repositories
 apk update
-apk add curl unzip ncurses postgresql postgresql-contrib nginx certbot
+apk add curl unzip ncurses postgresql postgresql-contrib nginx certbot libmagic
 ```
 
 ```sh tab="Debian/Ubuntu"
-apt install curl unzip libncurses5 postgresql postgresql-contrib nginx certbot
+apt install curl unzip libncurses5 postgresql postgresql-contrib nginx certbot libmagic
 ```
 
 ## Setup

lib/pleroma/application.ex

@@ -80,6 +80,7 @@ def start(_type, _args) do
         [
           Pleroma.Stats,
           Pleroma.JobQueueMonitor,
+          Pleroma.MIME,
           {Oban, Config.get(Oban)}
         ] ++
         task_children(@env) ++

lib/pleroma/mime.ex

@@ -6,8 +6,21 @@ defmodule Pleroma.MIME do
   @moduledoc """
   Returns the mime-type of a binary and optionally a normalized file-name.
   """
-  @default "application/octet-stream"
   @read_bytes 35
+  @pool __MODULE__.GenMagicPool
+
+  def child_spec(_) do
+    pool_size = Pleroma.Config.get!([:gen_magic_pool, :size])
+    name = @pool
+
+    %{
+      id: __MODULE__,
+      start: {GenMagic.Pool, :start_link, [[name: name, pool_size: pool_size]]},
+      type: :worker,
+      restart: :permanent,
+      shutdown: 500
+    }
+  end
 
   @spec file_mime_type(String.t(), String.t()) ::
           {:ok, content_type :: String.t(), filename :: String.t()} | {:error, any()} | :error
@@ -20,9 +33,10 @@ def file_mime_type(path, filename) do
 
   @spec file_mime_type(String.t()) :: {:ok, String.t()} | {:error, any()} | :error
   def file_mime_type(filename) do
-    File.open(filename, [:read], fn f ->
+    case GenMagic.Pool.perform(@pool, filename) do
-      check_mime_type(IO.binread(f, @read_bytes))
+      {:ok, %GenMagic.Result{mime_type: content_type}} -> {:ok, content_type}
-    end)
+      error -> error
+    end
   end
 
   def bin_mime_type(binary, filename) do
@@ -34,13 +48,14 @@ def bin_mime_type(binary, filename) do
 
   @spec bin_mime_type(binary()) :: {:ok, String.t()} | :error
   def bin_mime_type(<<head::binary-size(@read_bytes), _::binary>>) do
-    {:ok, check_mime_type(head)}
+    case GenMagic.Pool.perform(@pool, {:bytes, head}) do
+      {:ok, %GenMagic.Result{mime_type: content_type}} -> {:ok, content_type}
+      error -> error
+    end
   end
 
   def bin_mime_type(_), do: :error
 
-  def mime_type(<<_::binary>>), do: {:ok, @default}
-
   defp fix_extension(filename, content_type) do
     parts = String.split(filename, ".")
 
@@ -62,59 +77,4 @@ defp fix_extension(filename, content_type) do
         Enum.join([new_filename, String.split(content_type, "/") |> List.last()], ".")
     end
   end
-
-  defp check_mime_type(<<0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A, _::binary>>) do
-    "image/png"
-  end
-
-  defp check_mime_type(<<0x47, 0x49, 0x46, 0x38, _, 0x61, _::binary>>) do
-    "image/gif"
-  end
-
-  defp check_mime_type(<<0xFF, 0xD8, 0xFF, _::binary>>) do
-    "image/jpeg"
-  end
-
-  defp check_mime_type(<<0x1A, 0x45, 0xDF, 0xA3, _::binary>>) do
-    "video/webm"
-  end
-
-  defp check_mime_type(<<0x00, 0x00, 0x00, _, 0x66, 0x74, 0x79, 0x70, _::binary>>) do
-    "video/mp4"
-  end
-
-  defp check_mime_type(<<0x49, 0x44, 0x33, _::binary>>) do
-    "audio/mpeg"
-  end
-
-  defp check_mime_type(<<255, 251, _, 68, 0, 0, 0, 0, _::binary>>) do
-    "audio/mpeg"
-  end
-
-  defp check_mime_type(
-         <<0x4F, 0x67, 0x67, 0x53, 0x00, 0x02, 0x00, 0x00, _::size(160), 0x80, 0x74, 0x68, 0x65,
-           0x6F, 0x72, 0x61, _::binary>>
-       ) do
-    "video/ogg"
-  end
-
-  defp check_mime_type(<<0x4F, 0x67, 0x67, 0x53, 0x00, 0x02, 0x00, 0x00, _::binary>>) do
-    "audio/ogg"
-  end
-
-  defp check_mime_type(<<"RIFF", _::binary-size(4), "WAVE", _::binary>>) do
-    "audio/wav"
-  end
-
-  defp check_mime_type(<<"RIFF", _::binary-size(4), "WEBP", _::binary>>) do
-    "image/webp"
-  end
-
-  defp check_mime_type(<<"RIFF", _::binary-size(4), "AVI.", _::binary>>) do
-    "video/avi"
-  end
-
-  defp check_mime_type(_) do
-    @default
-  end
 end

mix.exs

@@ -197,6 +197,7 @@ defp deps do
        ref: "e0f16822d578866e186a0974d65ad58cddc1e2ab"},
       {:mox, "~> 0.5", only: :test},
       {:restarter, path: "./restarter"},
+      {:gen_magic, git: "https://github.com/hrefhref/gen_magic", branch: "develop"},
       {:open_api_spex,
        git: "https://git.pleroma.social/pleroma/elixir-libraries/open_api_spex.git",
        ref: "f296ac0924ba3cf79c7a588c4c252889df4c2edd"}

mix.lock

@@ -45,6 +45,7 @@
   "fast_sanitize": {:hex, :fast_sanitize, "0.1.7", "2a7cd8734c88a2de6de55022104f8a3b87f1fdbe8bbf131d9049764b53d50d0d", [:mix], [{:fast_html, "~> 1.0", [hex: :fast_html, repo: "hexpm", optional: false]}, {:plug, "~> 1.8", [hex: :plug, repo: "hexpm", optional: false]}], "hexpm", "f39fe8ea08fbac17487c30bf09b7d9f3e12472e51fb07a88ffeb8fd17da8ab67"},
   "flake_id": {:hex, :flake_id, "0.1.0", "7716b086d2e405d09b647121a166498a0d93d1a623bead243e1f74216079ccb3", [:mix], [{:base62, "~> 1.2", [hex: :base62, repo: "hexpm", optional: false]}, {:ecto, ">= 2.0.0", [hex: :ecto, repo: "hexpm", optional: true]}], "hexpm", "31fc8090fde1acd267c07c36ea7365b8604055f897d3a53dd967658c691bd827"},
   "floki": {:hex, :floki, "0.25.0", "b1c9ddf5f32a3a90b43b76f3386ca054325dc2478af020e87b5111c19f2284ac", [:mix], [{:html_entities, "~> 0.5.0", [hex: :html_entities, repo: "hexpm", optional: false]}], "hexpm", "631f4e627c46d5ecd347df5a2accdaf0621c77c3693c5b75a8ad58e84c61f242"},
+  "gen_magic": {:git, "https://github.com/hrefhref/gen_magic", "eafdc2ea156433ccd87d1d99eaf1be758064a1db", [branch: "develop"]},
   "gen_smtp": {:hex, :gen_smtp, "0.15.0", "9f51960c17769b26833b50df0b96123605a8024738b62db747fece14eb2fbfcc", [:rebar3], [], "hexpm", "29bd14a88030980849c7ed2447b8db6d6c9278a28b11a44cafe41b791205440f"},
   "gen_stage": {:hex, :gen_stage, "0.14.3", "d0c66f1c87faa301c1a85a809a3ee9097a4264b2edf7644bf5c123237ef732bf", [:mix], [], "hexpm"},
   "gen_state_machine": {:hex, :gen_state_machine, "2.0.5", "9ac15ec6e66acac994cc442dcc2c6f9796cf380ec4b08267223014be1c728a95", [:mix], [], "hexpm"},
@@ -74,6 +75,7 @@
   "mox": {:hex, :mox, "0.5.1", "f86bb36026aac1e6f924a4b6d024b05e9adbed5c63e8daa069bd66fb3292165b", [:mix], [], "hexpm", "052346cf322311c49a0f22789f3698eea030eec09b8c47367f0686ef2634ae14"},
   "myhtmlex": {:git, "https://git.pleroma.social/pleroma/myhtmlex.git", "ad0097e2f61d4953bfef20fb6abddf23b87111e6", [ref: "ad0097e2f61d4953bfef20fb6abddf23b87111e6", submodules: true]},
   "nimble_parsec": {:hex, :nimble_parsec, "0.5.3", "def21c10a9ed70ce22754fdeea0810dafd53c2db3219a0cd54cf5526377af1c6", [:mix], [], "hexpm", "589b5af56f4afca65217a1f3eb3fee7e79b09c40c742fddc1c312b3ac0b3399f"},
+  "nimble_pool": {:hex, :nimble_pool, "0.1.0", "ffa9d5be27eee2b00b0c634eb649aa27f97b39186fec3c493716c2a33e784ec6", [:mix], [], "hexpm", "343a1eaa620ddcf3430a83f39f2af499fe2370390d4f785cd475b4df5acaf3f9"},
   "nodex": {:git, "https://git.pleroma.social/pleroma/nodex", "cb6730f943cfc6aad674c92161be23a8411f15d1", [ref: "cb6730f943cfc6aad674c92161be23a8411f15d1"]},
   "oban": {:hex, :oban, "1.2.0", "7cca94d341be43d220571e28f69131c4afc21095b25257397f50973d3fc59b07", [:mix], [{:ecto_sql, "~> 3.1", [hex: :ecto_sql, repo: "hexpm", optional: false]}, {:jason, "~> 1.1", [hex: :jason, repo: "hexpm", optional: false]}, {:postgrex, "~> 0.14", [hex: :postgrex, repo: "hexpm", optional: false]}, {:telemetry, "~> 0.4", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "ba5f8b3f7d76967b3e23cf8014f6a13e4ccb33431e4808f036709a7f822362ee"},
   "open_api_spex": {:git, "https://git.pleroma.social/pleroma/elixir-libraries/open_api_spex.git", "f296ac0924ba3cf79c7a588c4c252889df4c2edd", [ref: "f296ac0924ba3cf79c7a588c4c252889df4c2edd"]},