Bumped version

* Updated downloader and runner to handle sponsorblock failures more gracefully

* stopped download worker from running if a media item is preventing download

.github/workflows/build_and_push_docker.yml

@@ -1,46 +0,0 @@
-name: Build and Push Docker Image
-
-on:
-  workflow_dispatch:
-    inputs:
-      image_tag:
-        description: 'Docker Image Tag'
-        required: true
-        default: 'dev'
-      platforms:
-        description: 'Build Platforms'
-        required: true
-        default: 'linux/amd64,linux/arm64'
-
-jobs:
-  docker:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_TOKEN }}
-
-      - name: Login to GHCR
-        uses: docker/login-action@v3
-        with:
-          registry: ghcr.io
-          username: ${{ github.repository_owner }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Build and push
-        uses: docker/build-push-action@v5
-        with:
-          platforms: ${{ github.event.inputs.platforms }}
-          push: true
-          file: ./selfhosted.Dockerfile
-          tags: |
-            keglin/pinchflat:${{ github.event.inputs.image_tag }}
-            ghcr.io/${{ github.repository_owner }}/pinchflat:${{ github.event.inputs.image_tag }}

.github/workflows/docker_release.yml

@@ -0,0 +1,81 @@
+name: Docker Release
+
+on:
+  workflow_dispatch:
+    inputs:
+      platforms:
+        type: choice
+        description: 'Build Platforms'
+        required: true
+        default: 'linux/amd64'
+        options:
+          - 'linux/amd64'
+          - 'linux/amd64,linux/arm64'
+  push:
+    branches:
+      - master
+  release:
+    types:
+      - published
+
+jobs:
+  build_and_push_docker:
+    runs-on: ubuntu-latest
+    env:
+      dev_arch: 'linux/amd64'
+      release_arch: 'linux/amd64,linux/arm64'
+    permissions:
+      contents: read
+      packages: write
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            docker.io/keglin/pinchflat
+            ghcr.io/${{ github.repository }}
+          # All non-release actions will be tagged as `dev` (ie: push, workflow_dispatch)
+          tags: |
+            type=semver,pattern={{version}},prefix=v
+            type=semver,pattern={{major}}.{{minor}},prefix=v
+            type=semver,pattern={{major}},prefix=v
+            type=raw,value=dev,enable=${{ github.event_name != 'release' }}
+          flavor: |
+            latest=auto
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Login to GitHub Container Registry
+        uses: docker/login-action@v3
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Build and Push
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          file: ./selfhosted.Dockerfile
+          # If the event is a release, use the release_arch, otherwise use the
+          # platforms input if present, falling back to dev_arch
+          platforms: ${{ github.event_name == 'release' && env.release_arch || (github.event.inputs.platforms || env.dev_arch) }}
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+          cache-from: type=gha
+          cache-to: type=gha,mode=max

.sobelow-conf

@@ -9,7 +9,15 @@
   threshold: :low,
   # All of these are ignorable because this app is intended to be single-user and self-hosted.
   # There is an expectation that the user won't intentionally run a FS Traversal on themselves
-  ignore: ["CI.System", "Traversal.FileModule", "Config.HTTPS", "Config.CSP"],
+  ignore:
+    [
+      "CI.System",
+      "Traversal.FileModule",
+      "Config.HTTPS",
+      "Config.CSP",
+      "XSS.ContentType",
+      "Traversal.SendFile"
+    ],
   ignore_files: [],
   version: false
 ]

README.md

@@ -22,12 +22,12 @@
 - [Installation](#installation)
   - [Unraid](#unraid)
   - [Docker](#docker)
-- [Authentication](#authentication)
+- [Username and Password (authentication)](https://github.com/kieraneglin/pinchflat/wiki/Username-and-Password)
 - [Frequently asked questions](https://github.com/kieraneglin/pinchflat/wiki/Frequently-Asked-Questions)
 - [Documentation](https://github.com/kieraneglin/pinchflat/wiki)
 - [EFF donations](#eff-donations)
 - [Pre-release disclaimer](#pre-release-disclaimer)
-- [Development](#development)
+- [Development and Contributing](https://github.com/kieraneglin/pinchflat/wiki/Development-and-Contributing)
 
 ## What it does
 
@@ -43,13 +43,16 @@ If it doesn't work for your use case, please make a feature request! You can als
 - Powerful naming system so content is stored where and how you want it
 - Easy-to-use web interface with presets to get you started right away
 - First-class support for media center apps like Plex, Jellyfin, and Kodi
+- Supports serving RSS feeds to your favourite podcast app (beta - [docs](<https://github.com/kieraneglin/pinchflat/wiki/Podcast-RSS-Feeds-(beta)>))
 - Automatically downloads new content from channels and playlists
   - Uses a novel approach to download new content more quickly than other apps
 - Supports downloading audio content
 - Custom rules for handling YouTube Shorts and livestreams
+- Optionally automatically delete old content ([docs](https://github.com/kieraneglin/pinchflat/wiki/Automatically-Delete-Media))
 - Advanced options like setting cutoff dates and filtering by title
 - Reliable hands-off operation
 - Can pass cookies to YouTube to download your private playlists ([docs](https://github.com/kieraneglin/pinchflat/wiki/YouTube-Cookies))
+- Sponsorblock integration
 
 ## Screenshots
 
@@ -64,6 +67,8 @@ Simply search for Pinchflat in the Community Apps store!
 
 ### Portainer
 
+Important: See the note below about storing config on a network file share. It's preferred to store the config on a local disk if at all possible.
+
 Docker Compose file:
 
 ```yaml
@@ -103,9 +108,13 @@ You _must_ ensure the host directories you've mounted are writable by the user r
 
 It's recommended to not run the container as root. Doing so can create permission issues if other apps need to work with the downloaded media. If you need to run any command as root, you can run `su` from the container's shell as there is no password set for the root user.
 
-## Username and Password
+### Advanced: storing Pinchflat config directory on a network share
 
-HTTP basic authentication is optionally supported. To use it, set the `BASIC_AUTH_USERNAME` and `BASIC_AUTH_PASSWORD` environment variables when starting the container. No authentication will be required unless you set _both_ of these.
+README: This is currently in the testing phase and not a recommended option (yet). The implications of changing this setting isn't clear and this could, conceivably, result in data loss. Only change this setting if you know what you're doing, why this is important, and are okay with possible data loss or DB corruption. This may become the default in the future once it's been tested more thoroughly.
+
+As pointed out in [#137](https://github.com/kieraneglin/pinchflat/issues/137), SQLite doesn't like being run in WAL mode on network shares. If you're running Pinchflat on a network share, you can disable WAL mode by setting the `JOURNAL_MODE` environment variable to `delete`. This will make Pinchflat run in rollback journal mode which is less performant but should work on network shares.
+
+If you change this setting and it works well for you, please leave a comment on [#137](https://github.com/kieraneglin/pinchflat/issues/137)! Doubly so if it does _not_ work well.
 
 ## EFF donations
 
@@ -115,30 +124,6 @@ A portion of all donations to Pinchflat will be donated to the [Electronic Front
 
 This is pre-release software and anything can break at any time. I make not guarantees about the stability of this software, forward-compatibility of updates, or integrity (both related to and independent of Pinchflat). Essentially, use at your own risk and expect there will be rough edges for now.
 
-## Development
-
-Pinchflat is written in Elixir - a functional programming language that runs on the Erlang VM. It uses the Phoenix web framework and SQLite for the database. The frontend is mostly normal server-rendered HTML with a little Alpine.js as-needed.
-
-Elixir is a personal favourite of mine and is ideal for building fault-tolerant systems. It's also a joy to work with and has a great community. If you're interested in contributing, I'd be happy to help you get started with Elixir - just open an issue with some questions and we can chat!
-
-### Local setup
-
-- `docker compose build --no-cache`
-- `docker compose up -d && docker attach pinchflat-phx-1`
-- After a few minutes the app should be accessible at `localhost:4008`
-  - Please let me know if you run into any hiccups here - I haven't had to bootstrap the app from scratch in a long time and I might have forgotten something
-- Media downloads and config will be stored in the `tmp` directory. Not the OS's `/tmp` directory, but the one in the root of the project
-
-### Running tests and linting
-
-- Open a shell with `docker compose exec phx bash`
-- Run `mix test` to run the tests
-- Run `mix check` to do a full testing, linting, and static analysis pass
-
-### Top tips
-
-- Look for any module that ends in `*_worker.ex` - these are where the interesting stuff happens and you can trace back from there to see how the app works. `lib/pinchflat/slow_indexing/media_collection_indexing_worker.ex` is a good place to start
-
 ## License
 
 See `LICENSE` file

assets/js/app.js

@@ -22,6 +22,7 @@ import { Socket } from 'phoenix'
 import { LiveSocket } from 'phoenix_live_view'
 import topbar from '../vendor/topbar'
 import Alpine from 'alpinejs'
+import './tabs'
 
 window.Alpine = Alpine
 Alpine.start()
@@ -38,6 +39,30 @@ let liveSocket = new LiveSocket('/live', Socket, {
   }
 })
 
+window.copyTextToClipboard = async (text) => {
+  // Navigator clipboard api needs a secure context (https)
+  if (navigator.clipboard && window.isSecureContext) {
+    await navigator.clipboard.writeText(text)
+  } else {
+    const textArea = document.createElement('textarea')
+    textArea.value = text
+    // Move textarea out of the viewport so it's not visible
+    textArea.style.position = 'absolute'
+    textArea.style.left = '-999999px'
+
+    document.body.prepend(textArea)
+    textArea.select()
+
+    try {
+      document.execCommand('copy')
+    } catch (error) {
+      console.error(error)
+    } finally {
+      textArea.remove()
+    }
+  }
+}
+
 // Show progress bar on live navigation and form submits
 topbar.config({ barColors: { 0: '#29d' }, shadowColor: 'rgba(0, 0, 0, .3)' })
 window.addEventListener('phx:page-loading-start', (_info) => topbar.show(300))

assets/js/tabs.js

@@ -0,0 +1,20 @@
+window.setTabIndex = (index) => {
+  window.location.hash = `tab-${index}`
+
+  return index
+}
+
+// The conditionals and currIndex stuff ensures that
+// the tab index is always set to 0 if the hash is empty
+// AND other hash values are ignored
+window.getTabIndex = (currIndex) => {
+  if (window.location.hash === '' || window.location.hash === '#') {
+    return 0
+  }
+
+  if (window.location.hash.startsWith('#tab-')) {
+    return parseInt(window.location.hash.replace('#tab-', ''))
+  }
+
+  return currIndex
+}

config/config.exs

@@ -20,10 +20,15 @@ config :pinchflat,
   tmpfile_directory: Path.join([System.tmp_dir!(), "pinchflat", "data"]),
   # Setting BASIC_AUTH_USERNAME and BASIC_AUTH_PASSWORD implies you want to use basic auth.
   # If either is unset, basic auth will not be used.
-  basic_auth_username: System.get_env("BASIC_AUTH_USERNAME"),
-  basic_auth_password: System.get_env("BASIC_AUTH_PASSWORD"),
+  basic_auth_username: "",
+  basic_auth_password: "",
+  expose_feed_endpoints: false,
   file_watcher_poll_interval: 1000
 
+config :pinchflat, Pinchflat.Repo,
+  journal_mode: :wal,
+  pool_size: 5
+
 # Configures the endpoint
 config :pinchflat, PinchflatWeb.Endpoint,
   url: [host: "localhost", port: 8945],
@@ -41,7 +46,13 @@ config :pinchflat, Oban,
   engine: Oban.Engines.Lite,
   repo: Pinchflat.Repo,
   # Keep old jobs for 30 days for display in the UI
-  plugins: [{Oban.Plugins.Pruner, max_age: 30 * 24 * 60 * 60}],
+  plugins: [
+    {Oban.Plugins.Pruner, max_age: 30 * 24 * 60 * 60},
+    {Oban.Plugins.Cron,
+     crontab: [
+       {"@daily", Pinchflat.Downloading.MediaRetentionWorker}
+     ]}
+  ],
   # TODO: consider making this an env var or something?
   queues: [
     default: 10,
@@ -74,7 +85,7 @@ config :esbuild,
 
 # Configure tailwind (the version is required)
 config :tailwind,
-  version: "3.3.2",
+  version: "3.4.3",
   default: [
     args: ~w(
       --config=tailwind.config.js

config/runtime.exs

@@ -47,16 +47,13 @@ if config_env() == :prod do
   metadata_path = System.get_env("METADATA_PATH", Path.join([config_path, "metadata"]))
   extras_path = System.get_env("EXTRAS_PATH", Path.join([config_path, "extras"]))
 
-  # We want to force _some_ level of useful logging in production
-  acceptable_log_levels = ~w(debug info)a
-  log_level = String.to_existing_atom(System.get_env("LOG_LEVEL", "info"))
+  # For running PF as a podcast host on self-hosted environments
+  expose_feed_endpoints = String.length(System.get_env("EXPOSE_FEED_ENDPOINTS", "")) > 0
 
-  if log_level in acceptable_log_levels do
-    config :logger, level: log_level
-  else
-    Logger.error("Invalid log level: #{log_level}. Defaulting to info.")
-    config :logger, level: :info
-  end
+  # For testing alternate journal modes (see issue #137)
+  journal_mode = String.to_existing_atom(System.get_env("JOURNAL_MODE", "wal"))
+
+  config :logger, level: String.to_existing_atom(System.get_env("LOG_LEVEL", "info"))
 
   config :pinchflat,
     yt_dlp_executable: System.find_executable("yt-dlp"),
@@ -64,11 +61,12 @@ if config_env() == :prod do
     metadata_directory: metadata_path,
     extras_directory: extras_path,
     tmpfile_directory: Path.join([System.tmp_dir!(), "pinchflat", "data"]),
-    dns_cluster_query: System.get_env("DNS_CLUSTER_QUERY")
+    dns_cluster_query: System.get_env("DNS_CLUSTER_QUERY"),
+    expose_feed_endpoints: expose_feed_endpoints
 
   config :pinchflat, Pinchflat.Repo,
     database: db_path,
-    pool_size: String.to_integer(System.get_env("POOL_SIZE") || "5")
+    journal_mode: journal_mode
 
   # The secret key base is used to sign/encrypt cookies and other secrets.
   # A default value is used in config/dev.exs and config/test.exs but you
@@ -121,54 +119,4 @@ if config_env() == :prod do
        formatter: Logger.Formatter.new()
      }}
   ]
-
-  # ## SSL Support
-  #
-  # To get SSL working, you will need to add the `https` key
-  # to your endpoint configuration:
-  #
-  #     config :pinchflat, PinchflatWeb.Endpoint,
-  #       https: [
-  #         ...,
-  #         port: 443,
-  #         cipher_suite: :strong,
-  #         keyfile: System.get_env("SOME_APP_SSL_KEY_PATH"),
-  #         certfile: System.get_env("SOME_APP_SSL_CERT_PATH")
-  #       ]
-  #
-  # The `cipher_suite` is set to `:strong` to support only the
-  # latest and more secure SSL ciphers. This means old browsers
-  # and clients may not be supported. You can set it to
-  # `:compatible` for wider support.
-  #
-  # `:keyfile` and `:certfile` expect an absolute path to the key
-  # and cert in disk or a relative path inside priv, for example
-  # "priv/ssl/server.key". For all supported SSL configuration
-  # options, see https://hexdocs.pm/plug/Plug.SSL.html#configure/1
-  #
-  # We also recommend setting `force_ssl` in your endpoint, ensuring
-  # no data is ever sent via http, always redirecting to https:
-  #
-  #     config :pinchflat, PinchflatWeb.Endpoint,
-  #       force_ssl: [hsts: true]
-  #
-  # Check `Plug.SSL` for all available options in `force_ssl`.
-
-  # ## Configuring the mailer
-  #
-  # In production you need to configure the mailer to use a different adapter.
-  # Also, you may need to configure the Swoosh API client of your choice if you
-  # are not using SMTP. Here is an example of the configuration:
-  #
-  #     config :pinchflat, Pinchflat.Mailer,
-  #       adapter: Swoosh.Adapters.Mailgun,
-  #       api_key: System.get_env("MAILGUN_API_KEY"),
-  #       domain: System.get_env("MAILGUN_DOMAIN")
-  #
-  # For this example you need include a HTTP client required by Swoosh API client.
-  # Swoosh supports Hackney and Finch out of the box:
-  #
-  #     config :swoosh, :api_client, Swoosh.ApiClient.Hackney
-  #
-  # See https://hexdocs.pm/swoosh/Swoosh.html#module-installation for details.
 end

config/test.exs

@@ -17,7 +17,6 @@ config :pinchflat, Oban, testing: :manual
 # Run `mix help test` for more information.
 config :pinchflat, Pinchflat.Repo,
   database: Path.expand("../priv/repo/pinchflat_test.db", Path.dirname(__ENV__.file)),
-  pool_size: 5,
   pool: Ecto.Adapters.SQL.Sandbox
 
 # We don't run a server during test. If one is required,
@@ -33,8 +32,7 @@ config :pinchflat, Pinchflat.Mailer, adapter: Swoosh.Adapters.Test
 # Disable swoosh api client as it is only required for production adapters.
 config :swoosh, :api_client, false
 
-# Print only warnings and errors during test
-config :logger, level: :warning
+config :logger, level: :critical
 
 # Initialize plugs at runtime for faster test compilation
 config :phoenix, :plug_init_mode, :runtime

lib/pinchflat/boot/data_backfill_worker.ex

@@ -1,62 +0,0 @@
-defmodule Pinchflat.Boot.DataBackfillWorker do
-  @moduledoc false
-
-  use Oban.Worker,
-    queue: :local_metadata,
-    unique: [period: :infinity, states: [:available, :scheduled, :retryable]],
-    tags: ["media_item", "media_metadata", "local_metadata", "data_backfill"]
-
-  # This one is going to be a little more self-contained
-  # instead of relying on outside modules for the methods.
-  # That's because, for now, these methods are not intended
-  # to be used elsewhere.
-  #
-  # I'm just trying out that pattern and seeing if I like it better
-  # so this may change.
-  import Ecto.Query, warn: false
-  require Logger
-
-  alias __MODULE__
-  alias Pinchflat.Repo
-
-  @doc """
-  Cancels all pending backfill jobs. Useful for ensuring worker runs immediately
-  on app boot.
-
-  Returns {:ok, integer()}
-  """
-  def cancel_pending_backfill_jobs do
-    Oban.Job
-    |> where(worker: "Pinchflat.Boot.DataBackfillWorker")
-    |> Oban.cancel_all_jobs()
-  end
-
-  @impl Oban.Worker
-  @doc """
-  Performs one-off tasks to get data in the right shape.
-  This can be needed when we add new features or change the way
-  we store data. Must be idempotent. All new data should already
-  conform to the expected schema so this should only be needed
-  for existing data. Still runs periodically to be safe.
-
-  Returns :ok
-  """
-  def perform(%Oban.Job{}) do
-    Logger.info("Running data backfill worker")
-    # Nothing to do for now - just reschedule
-    # Keeping in-place because we _will_ need it in the future
-
-    reschedule_backfill()
-
-    :ok
-  end
-
-  defp reschedule_backfill do
-    # Run hourly
-    next_run_in = 60 * 60
-
-    %{}
-    |> DataBackfillWorker.new(schedule_in: next_run_in)
-    |> Repo.insert_unique_job()
-  end
-end

lib/pinchflat/boot/nfo_backfill_worker.ex

@@ -0,0 +1,70 @@
+defmodule Pinchflat.Boot.NfoBackfillWorker do
+  @moduledoc false
+
+  use Oban.Worker,
+    queue: :local_metadata,
+    # This should have it running once _ever_ (until the job is pruned, anyway)
+    # NOTE: remove within the next month
+    unique: [period: :infinity, states: Oban.Job.states()],
+    tags: ["media_item", "media_metadata", "local_metadata", "data_backfill"]
+
+  import Ecto.Query, warn: false
+  require Logger
+
+  alias Pinchflat.Repo
+  alias Pinchflat.Media
+  alias Pinchflat.Media.MediaItem
+  alias Pinchflat.Metadata.NfoBuilder
+  alias Pinchflat.Metadata.MetadataFileHelpers
+
+  @doc """
+  Runs a one-off backfill job to regenerate NFO files for media items that have
+  both an NFO file and a metadata file. This is needed because NFO files weren't
+  escaping characters properly so we need to regenerate them.
+
+  This job will only run once as long as I remove it before the jobs are pruned in a month.
+
+  Returns :ok
+  """
+  @impl Oban.Worker
+  def perform(%Oban.Job{}) do
+    Logger.info("Running NFO backfill worker")
+
+    media_items = get_media_items_to_backfill()
+
+    Enum.each(media_items, fn media_item ->
+      nfo_exists = File.exists?(media_item.nfo_filepath)
+      metadata_exists = File.exists?(media_item.metadata.metadata_filepath)
+
+      if nfo_exists && metadata_exists do
+        Logger.info("NFO and metadata exist for media item #{media_item.id} - proceeding")
+
+        regenerate_nfo_for_media_item(media_item)
+      end
+    end)
+
+    :ok
+  end
+
+  defp get_media_items_to_backfill do
+    from(m in MediaItem, where: not is_nil(m.nfo_filepath))
+    |> Repo.all()
+    |> Repo.preload([:metadata, source: :media_profile])
+  end
+
+  defp regenerate_nfo_for_media_item(media_item) do
+    try do
+      case MetadataFileHelpers.read_compressed_metadata(media_item.metadata.metadata_filepath) do
+        {:ok, metadata} ->
+          Media.update_media_item(media_item, %{
+            nfo_filepath: NfoBuilder.build_and_store_for_media_item(media_item.nfo_filepath, metadata)
+          })
+
+        _err ->
+          Logger.error("Failed to read metadata for media item #{media_item.id}")
+      end
+    rescue
+      e -> Logger.error("Unknown error regenerating NFO file for MI ##{media_item.id}: #{inspect(e)}")
+    end
+  end
+end

lib/pinchflat/boot/post_job_startup_tasks.ex

@@ -7,13 +7,13 @@ defmodule Pinchflat.Boot.PostJobStartupTasks do
   Phoenix supervision tree.
   """
 
+  alias Pinchflat.Repo
+  alias Pinchflat.Boot.NfoBackfillWorker
+
   # restart: :temporary means that this process will never be restarted (ie: will run once and then die)
   use GenServer, restart: :temporary
   import Ecto.Query, warn: false
 
-  alias Pinchflat.Repo
-  alias Pinchflat.Boot.DataBackfillWorker
-
   def start_link(opts \\ []) do
     GenServer.start_link(__MODULE__, %{}, opts)
   end
@@ -29,16 +29,8 @@ defmodule Pinchflat.Boot.PostJobStartupTasks do
   """
   @impl true
   def init(state) do
-    enqueue_backfill_worker()
+    Repo.insert_unique_job(NfoBackfillWorker.new(%{}))
 
     {:ok, state}
   end
-
-  defp enqueue_backfill_worker do
-    DataBackfillWorker.cancel_pending_backfill_jobs()
-
-    %{}
-    |> DataBackfillWorker.new()
-    |> Repo.insert_unique_job()
-  end
 end

lib/pinchflat/boot/pre_job_startup_tasks.ex

@@ -14,6 +14,7 @@ defmodule Pinchflat.Boot.PreJobStartupTasks do
 
   alias Pinchflat.Repo
   alias Pinchflat.Settings
+  alias Pinchflat.YtDlp.CommandRunner
   alias Pinchflat.Filesystem.FilesystemHelpers
 
   def start_link(opts \\ []) do
@@ -62,7 +63,8 @@ defmodule Pinchflat.Boot.PreJobStartupTasks do
   end
 
   defp apply_default_settings do
-    Settings.fetch!(:onboarding, true)
-    Settings.fetch!(:pro_enabled, false)
+    {:ok, yt_dlp_version} = CommandRunner.version()
+
+    Settings.set(yt_dlp_version: yt_dlp_version)
   end
 end

lib/pinchflat/downloading/download_option_builder.ex

@@ -22,6 +22,7 @@ defmodule Pinchflat.Downloading.DownloadOptionBuilder do
         thumbnail_options(media_item_with_preloads) ++
         metadata_options(media_profile) ++
         quality_options(media_profile) ++
+        sponsorblock_options(media_profile) ++
         output_options(media_item_with_preloads)
 
     {:ok, built_options}
@@ -116,6 +117,17 @@ defmodule Pinchflat.Downloading.DownloadOptionBuilder do
     end
   end
 
+  defp sponsorblock_options(media_profile) do
+    categories = media_profile.sponsorblock_categories
+    behaviour = media_profile.sponsorblock_behaviour
+
+    case {behaviour, categories} do
+      {_, []} -> []
+      {:remove, _} -> [sponsorblock_remove: Enum.join(categories, ",")]
+      {:disabled, _} -> []
+    end
+  end
+
   defp output_options(media_item_with_preloads) do
     [
       output: build_output_path_for(media_item_with_preloads.source)

lib/pinchflat/downloading/media_download_worker.ex

@@ -19,27 +19,29 @@ defmodule Pinchflat.Downloading.MediaDownloadWorker do
 
   Returns {:ok, %Task{}} | {:error, :duplicate_job} | {:error, %Ecto.Changeset{}}
   """
-  def kickoff_with_task(media_item, opts \\ []) do
+  def kickoff_with_task(media_item, job_args \\ %{}, job_opts \\ []) do
     %{id: media_item.id}
-    |> MediaDownloadWorker.new(opts)
+    |> Map.merge(job_args)
+    |> MediaDownloadWorker.new(job_opts)
     |> Tasks.create_job_with_task(media_item)
   end
 
   @doc """
   For a given media item, download the media alongside any options.
-  Does not download media if its source is set to not download media.
+  Does not download media if its source is set to not download media
+  (unless forced).
 
   Returns :ok | {:ok, %MediaItem{}} | {:error, any, ...any}
   """
   @impl Oban.Worker
-  def perform(%Oban.Job{args: %{"id" => media_item_id}}) do
+  def perform(%Oban.Job{args: %{"id" => media_item_id} = args}) do
     media_item =
       media_item_id
       |> Media.get_media_item!()
       |> Repo.preload(:source)
 
-    # If the source is set to not download media, perform a no-op
-    if media_item.source.download_media do
+    # If the source or media item is set to not download media, perform a no-op unless forced
+    if (media_item.source.download_media && !media_item.prevent_download) || args["force"] do
       download_media_and_schedule_jobs(media_item)
     else
       :ok
@@ -56,8 +58,11 @@ defmodule Pinchflat.Downloading.MediaDownloadWorker do
 
         {:ok, updated_media_item}
 
-      err ->
-        err
+      {:recovered, _} ->
+        {:error, :retry}
+
+      {:error, _message} ->
+        {:error, :download_failed}
     end
   end
 

lib/pinchflat/downloading/media_downloader.ex

@@ -5,12 +5,15 @@ defmodule Pinchflat.Downloading.MediaDownloader do
   to download the media with the desired options.
   """
 
+  require Logger
+
   alias Pinchflat.Repo
   alias Pinchflat.Media
   alias Pinchflat.Media.MediaItem
   alias Pinchflat.Metadata.NfoBuilder
   alias Pinchflat.Metadata.MetadataParser
   alias Pinchflat.Metadata.MetadataFileHelpers
+  alias Pinchflat.Filesystem.FilesystemHelpers
   alias Pinchflat.Downloading.DownloadOptionBuilder
 
   alias Pinchflat.YtDlp.Media, as: YtDlpMedia
@@ -27,31 +30,67 @@ defmodule Pinchflat.Downloading.MediaDownloader do
   Returns {:ok, %MediaItem{}} | {:error, any, ...any}
   """
   def download_for_media_item(%MediaItem{} = media_item) do
-    item_with_preloads = Repo.preload(media_item, [:metadata, source: :media_profile])
+    output_filepath = FilesystemHelpers.generate_metadata_tmpfile(:json)
+    media_with_preloads = Repo.preload(media_item, [:metadata, source: :media_profile])
 
-    case download_with_options(media_item.original_url, item_with_preloads) do
+    case download_with_options(media_item.original_url, media_with_preloads, output_filepath) do
       {:ok, parsed_json} ->
+        update_media_item_from_parsed_json(media_with_preloads, parsed_json)
+
+      {:error, message, _exit_code} ->
+        Logger.error("yt-dlp download error for media item ##{media_with_preloads.id}: #{inspect(message)}")
+
+        if String.contains?(to_string(message), recoverable_errors()) do
+          attempt_update_media_item(media_with_preloads, output_filepath)
+
+          {:recovered, message}
+        else
+          {:error, message}
+        end
+
+      err ->
+        Logger.error("Unknown error downloading media item ##{media_with_preloads.id}: #{inspect(err)}")
+
+        {:error, "Unknown error: #{inspect(err)}"}
+    end
+  end
+
+  defp attempt_update_media_item(media_with_preloads, output_filepath) do
+    with {:ok, contents} <- File.read(output_filepath),
+         {:ok, parsed_json} <- Phoenix.json_library().decode(contents) do
+      Logger.info("""
+      Recovery from yt-dlp error seems possible. Updating media item ##{media_with_preloads.id}
+      with parsed JSON from partial download attempt. Full download will be re-attemted in future
+      anyway
+      """)
+
+      update_media_item_from_parsed_json(media_with_preloads, parsed_json)
+    else
+      err ->
+        Logger.error("Unable to recover error for media item ##{media_with_preloads.id}: #{inspect(err)}")
+
+        {:error, :retry_failed}
+    end
+  end
+
+  defp update_media_item_from_parsed_json(media_with_preloads, parsed_json) do
     parsed_attrs =
       parsed_json
       |> MetadataParser.parse_for_media_item()
       |> Map.merge(%{
         media_downloaded_at: DateTime.utc_now(),
-            nfo_filepath: determine_nfo_filepath(item_with_preloads, parsed_json),
+        nfo_filepath: determine_nfo_filepath(media_with_preloads, parsed_json),
         metadata: %{
           # IDEA: might be worth kicking off a job for this since thumbnail fetching
           # could fail and I want to handle that in isolation
-              metadata_filepath: MetadataFileHelpers.compress_and_store_metadata_for(media_item, parsed_json),
-              thumbnail_filepath: MetadataFileHelpers.download_and_store_thumbnail_for(media_item, parsed_json)
+          metadata_filepath: MetadataFileHelpers.compress_and_store_metadata_for(media_with_preloads, parsed_json),
+          thumbnail_filepath: MetadataFileHelpers.download_and_store_thumbnail_for(media_with_preloads, parsed_json)
         }
       })
 
     # Don't forgor to use preloaded associations or updates to
     # associations won't work!
-        Media.update_media_item(item_with_preloads, parsed_attrs)
-
-      err ->
-        err
-    end
+    Media.update_media_item(media_with_preloads, parsed_attrs)
   end
 
   defp determine_nfo_filepath(media_item, parsed_json) do
@@ -64,9 +103,15 @@ defmodule Pinchflat.Downloading.MediaDownloader do
     end
   end
 
-  defp download_with_options(url, item_with_preloads) do
+  defp download_with_options(url, item_with_preloads, output_filepath) do
     {:ok, options} = DownloadOptionBuilder.build(item_with_preloads)
 
-    YtDlpMedia.download(url, options)
+    YtDlpMedia.download(url, options, output_filepath: output_filepath)
+  end
+
+  defp recoverable_errors do
+    [
+      "Unable to communicate with SponsorBlock"
+    ]
   end
 end

lib/pinchflat/downloading/media_retention_worker.ex

@@ -0,0 +1,33 @@
+defmodule Pinchflat.Downloading.MediaRetentionWorker do
+  @moduledoc false
+
+  use Oban.Worker,
+    queue: :local_metadata,
+    unique: [period: :infinity, states: [:available, :scheduled, :retryable, :executing]],
+    tags: ["media_item", "local_metadata"]
+
+  require Logger
+
+  alias Pinchflat.Media
+
+  @doc """
+  Deletes media items that are past their retention date and prevents
+  them from being re-downloaded.
+
+  This worker is scheduled to run daily via the Oban Cron plugin.
+
+  Returns :ok
+  """
+  @impl Oban.Worker
+  def perform(%Oban.Job{}) do
+    cullable_media = Media.list_cullable_media_items()
+    Logger.info("Culling #{length(cullable_media)} media items past their retention date")
+
+    Enum.each(cullable_media, fn media_item ->
+      Media.delete_media_files(media_item, %{
+        prevent_download: true,
+        culled_at: DateTime.utc_now()
+      })
+    end)
+  end
+end

lib/pinchflat/fast_indexing/fast_indexing_helpers.ex

@@ -5,8 +5,10 @@ defmodule Pinchflat.FastIndexing.FastIndexingHelpers do
   Many of these methods are made to be kickoff or be consumed by workers.
   """
 
+  alias Pinchflat.Repo
   alias Pinchflat.Media
   alias Pinchflat.Sources.Source
+  alias Pinchflat.Media.MediaQuery
   alias Pinchflat.FastIndexing.YoutubeRss
   alias Pinchflat.Downloading.MediaDownloadWorker
   alias Pinchflat.FastIndexing.MediaIndexingWorker
@@ -27,7 +29,7 @@ defmodule Pinchflat.FastIndexing.FastIndexingHelpers do
   """
   def kickoff_indexing_tasks_from_youtube_rss_feed(%Source{} = source) do
     {:ok, media_ids} = YoutubeRss.get_recent_media_ids_from_rss(source)
-    existing_media_items = Media.list_media_items_by_media_id_for(source, media_ids)
+    existing_media_items = list_media_items_by_media_id_for(source, media_ids)
     new_media_ids = media_ids -- Enum.map(existing_media_items, & &1.media_id)
 
     Enum.each(new_media_ids, fn media_id ->
@@ -61,6 +63,13 @@ defmodule Pinchflat.FastIndexing.FastIndexingHelpers do
     end
   end
 
+  defp list_media_items_by_media_id_for(source, media_ids) do
+    MediaQuery.new()
+    |> MediaQuery.for_source(source)
+    |> MediaQuery.with_media_ids(media_ids)
+    |> Repo.all()
+  end
+
   defp create_media_item_from_url(source, url) do
     {:ok, media_attrs} = YtDlpMedia.get_media_attributes(url)
 

lib/pinchflat/media/media.ex

@@ -9,6 +9,7 @@ defmodule Pinchflat.Media do
   alias Pinchflat.Tasks
   alias Pinchflat.Sources.Source
   alias Pinchflat.Media.MediaItem
+  alias Pinchflat.Media.MediaQuery
   alias Pinchflat.Metadata.MediaMetadata
   alias Pinchflat.Filesystem.FilesystemHelpers
 
@@ -22,66 +23,32 @@ defmodule Pinchflat.Media do
   end
 
   @doc """
-  Returns a list of media_items for a given source.
+  Returns a list of media_items that are cullable based on the retention period
+  of the source they belong to.
 
-  Returns [%MediaItem{}, ...].
+  Returns [%MediaItem{}, ...]
   """
-  def list_media_items_for(%Source{} = source) do
-    MediaItem
-    |> where([mi], mi.source_id == ^source.id)
-    |> Repo.all()
-  end
-
-  @doc """
-  Fetches all media items belonging to a given source that have a media_id in the given list.
-  Useful for determining the what media items we DON'T already have for fast indexing.
-
-  NOTE: These queries are getting a little tedious. When I have the time, I should see about
-  implementing a query pattern and having these compose queries from a common base. This would
-  also let me compose simple queries in the module using them for one-off methods
-
-  Returns [%MediaItem{}, ...].
-  """
-  def list_media_items_by_media_id_for(%Source{} = source, media_ids) do
-    MediaItem
-    |> where([mi], mi.source_id == ^source.id and mi.media_id in ^media_ids)
+  def list_cullable_media_items do
+    MediaQuery.new()
+    |> MediaQuery.with_media_filepath()
+    |> MediaQuery.with_passed_retention_period()
+    |> MediaQuery.with_no_culling_prevention()
     |> Repo.all()
   end
 
   @doc """
   Returns a list of pending media_items for a given source, where
   pending means the `media_filepath` is `nil` AND the media_item
-  matches the format selection rules of the parent media_profile.
-
-  See `build_format_clauses` but tl;dr is it _may_ filter based
-  on shorts or livestreams depending on the media_profile settings.
+  matches satisfies `MediaQuery.with_media_pending_download`. You
+  should really check out that function if you need to know more
+  because it has a lot going on.
 
   Returns [%MediaItem{}, ...].
   """
-  def list_pending_media_items_for(%Source{} = source, opts \\ []) do
-    limit = Keyword.get(opts, :limit, nil)
-    media_profile = Repo.preload(source, :media_profile).media_profile
-
-    MediaItem
-    |> where([mi], mi.source_id == ^source.id and is_nil(mi.media_filepath))
-    |> where(^build_format_clauses(media_profile))
-    |> where(^maybe_apply_cutoff_date(source))
-    |> where(^maybe_apply_title_regex(source))
-    |> Repo.maybe_limit(limit)
-    |> Repo.all()
-  end
-
-  @doc """
-  Returns a list of downloaded media_items for a given source.
-
-  Returns [%MediaItem{}, ...].
-  """
-  def list_downloaded_media_items_for(%Source{} = source, opts \\ []) do
-    limit = Keyword.get(opts, :limit, nil)
-
-    MediaItem
-    |> where([mi], mi.source_id == ^source.id and not is_nil(mi.media_filepath))
-    |> Repo.maybe_limit(limit)
+  def list_pending_media_items_for(%Source{} = source) do
+    MediaQuery.new()
+    |> MediaQuery.for_source(source)
+    |> MediaQuery.with_media_pending_download()
     |> Repo.all()
   end
 
@@ -97,10 +64,9 @@ defmodule Pinchflat.Media do
   def pending_download?(%MediaItem{} = media_item) do
     media_item = Repo.preload(media_item, source: :media_profile)
 
-    MediaItem
-    |> where([mi], mi.id == ^media_item.id and is_nil(mi.media_filepath))
-    |> where(^build_format_clauses(media_item.source.media_profile))
-    |> where(^maybe_apply_cutoff_date(media_item.source))
+    MediaQuery.new()
+    |> MediaQuery.with_id(media_item.id)
+    |> MediaQuery.with_media_pending_download()
     |> Repo.exists?()
   end
 
@@ -119,20 +85,9 @@ defmodule Pinchflat.Media do
   def search(search_term, opts) do
     limit = Keyword.get(opts, :limit, 50)
 
-    from(mi in MediaItem,
-      join: mi_search_index in assoc(mi, :media_items_search_index),
-      where: fragment("media_items_search_index MATCH ?", ^search_term),
-      select_merge: %{
-        matching_search_term:
-          fragment("""
-            coalesce(snippet(media_items_search_index, 0, '[PF_HIGHLIGHT]', '[/PF_HIGHLIGHT]', '...', 20), '') ||
-            ' ' ||
-            coalesce(snippet(media_items_search_index, 1, '[PF_HIGHLIGHT]', '[/PF_HIGHLIGHT]', '...', 20), '')
-          """)
-      },
-      order_by: [desc: fragment("rank")],
-      limit: ^limit
-    )
+    MediaQuery.new()
+    |> MediaQuery.matching_search_term(search_term)
+    |> Repo.maybe_limit(limit)
     |> Repo.all()
   end
 
@@ -200,7 +155,7 @@ defmodule Pinchflat.Media do
     Tasks.delete_tasks_for(media_item)
 
     if delete_files do
-      {:ok, _} = delete_media_files(media_item)
+      {:ok, _} = do_delete_media_files(media_item)
     end
 
     # Should delete these no matter what
@@ -208,6 +163,25 @@ defmodule Pinchflat.Media do
     Repo.delete(media_item)
   end
 
+  @doc """
+  Deletes the tasks and media files associated with a media_item but leaves the
+  media_item in the database. Does not delete anything to do with associated metadata.
+
+  Optionally accepts a second argument `addl_attrs` which will be merged into the
+  media_item before it is updated. Useful for setting things like `prevent_download`
+  and `culled_at`, if wanted
+
+  Returns {:ok, %MediaItem{}} | {:error, %Ecto.Changeset{}}
+  """
+  def delete_media_files(%MediaItem{} = media_item, addl_attrs \\ %{}) do
+    filepath_attrs = MediaItem.filepath_attribute_defaults()
+
+    Tasks.delete_tasks_for(media_item)
+    {:ok, _} = do_delete_media_files(media_item)
+
+    update_media_item(media_item, Map.merge(filepath_attrs, addl_attrs))
+  end
+
   @doc """
   Returns an `%Ecto.Changeset{}` for tracking media_item changes.
   """
@@ -215,7 +189,7 @@ defmodule Pinchflat.Media do
     MediaItem.changeset(media_item, attrs)
   end
 
-  defp delete_media_files(media_item) do
+  defp do_delete_media_files(media_item) do
     mapped_struct = Map.from_struct(media_item)
 
     MediaItem.filepath_attributes()
@@ -239,55 +213,4 @@ defmodule Pinchflat.Media do
     |> Enum.filter(&is_binary/1)
     |> Enum.each(&FilesystemHelpers.delete_file_and_remove_empty_directories/1)
   end
-
-  defp maybe_apply_cutoff_date(source) do
-    if source.download_cutoff_date do
-      dynamic([mi], mi.upload_date >= ^source.download_cutoff_date)
-    else
-      dynamic(true)
-    end
-  end
-
-  defp maybe_apply_title_regex(source) do
-    if source.title_filter_regex do
-      dynamic([mi], fragment("regexp_like(?, ?)", mi.title, ^source.title_filter_regex))
-    else
-      dynamic(true)
-    end
-  end
-
-  defp build_format_clauses(media_profile) do
-    mapped_struct = Map.from_struct(media_profile)
-
-    Enum.reduce(mapped_struct, dynamic(true), fn attr, dynamic ->
-      case {attr, media_profile} do
-        {{:shorts_behaviour, :only}, %{livestream_behaviour: :only}} ->
-          dynamic(
-            [mi],
-            ^dynamic and (mi.livestream == true or mi.short_form_content == true)
-          )
-
-        # Technically redundant, but makes the other clauses easier to parse
-        # (redundant because this condition is the same as the condition above, just flipped)
-        {{:livestream_behaviour, :only}, %{shorts_behaviour: :only}} ->
-          dynamic
-
-        {{:shorts_behaviour, :only}, _} ->
-          dynamic([mi], ^dynamic and mi.short_form_content == true)
-
-        {{:livestream_behaviour, :only}, _} ->
-          dynamic([mi], ^dynamic and mi.livestream == true)
-
-        {{:shorts_behaviour, :exclude}, %{livestream_behaviour: lb}} when lb != :only ->
-          dynamic([mi], ^dynamic and mi.short_form_content == false)
-
-        {{:livestream_behaviour, :exclude}, %{shorts_behaviour: sb}} when sb != :only ->
-          # return records with livestream: false
-          dynamic([mi], ^dynamic and mi.livestream == false)
-
-        _ ->
-          dynamic
-      end
-    end)
-  end
 end

lib/pinchflat/media/media_item.ex

@@ -5,6 +5,7 @@ defmodule Pinchflat.Media.MediaItem do
 
   use Ecto.Schema
   import Ecto.Changeset
+  import Pinchflat.Utils.ChangesetUtils
 
   alias Pinchflat.Tasks.Task
   alias Pinchflat.Sources.Source
@@ -21,6 +22,7 @@ defmodule Pinchflat.Media.MediaItem do
     :source_id,
     :short_form_content,
     :upload_date,
+    :duration_seconds,
     # these fields are captured only on download
     :media_downloaded_at,
     :media_filepath,
@@ -28,10 +30,15 @@ defmodule Pinchflat.Media.MediaItem do
     :subtitle_filepaths,
     :thumbnail_filepath,
     :metadata_filepath,
-    :nfo_filepath
+    :nfo_filepath,
+    # These are user or system controlled fields
+    :prevent_download,
+    :prevent_culling,
+    :culled_at
   ]
   # Pretty much all the fields captured at index are required.
   @required_fields ~w(
+    uuid
     title
     original_url
     livestream
@@ -42,6 +49,11 @@ defmodule Pinchflat.Media.MediaItem do
   )a
 
   schema "media_items" do
+    # This is _not_ used as the primary key or internally in the database
+    # relations. This is only used to prevent an enumeration attack on the streaming
+    # and RSS feed endpoints since those _must_ be public (ie: no basic auth)
+    field :uuid, Ecto.UUID
+
     field :title, :string
     field :media_id, :string
     field :description, :string
@@ -50,6 +62,7 @@ defmodule Pinchflat.Media.MediaItem do
     field :short_form_content, :boolean, default: false
     field :media_downloaded_at, :utc_datetime
     field :upload_date, :date
+    field :duration_seconds, :integer
 
     field :media_filepath, :string
     field :media_size_bytes, :integer
@@ -61,6 +74,10 @@ defmodule Pinchflat.Media.MediaItem do
     # Will very likely revisit because I can't leave well-enough alone.
     field :subtitle_filepaths, {:array, {:array, :string}}, default: []
 
+    field :prevent_download, :boolean, default: false
+    field :prevent_culling, :boolean, default: false
+    field :culled_at, :utc_datetime
+
     field :matching_search_term, :string, virtual: true
 
     belongs_to :source, Source
@@ -78,6 +95,7 @@ defmodule Pinchflat.Media.MediaItem do
     media_item
     |> cast(attrs, @allowed_fields)
     |> cast_assoc(:metadata, with: &MediaMetadata.changeset/2, required: false)
+    |> dynamic_default(:uuid, fn _ -> Ecto.UUID.generate() end)
     |> validate_required(@required_fields)
     |> unique_constraint([:media_id, :source_id])
   end
@@ -86,4 +104,14 @@ defmodule Pinchflat.Media.MediaItem do
   def filepath_attributes do
     ~w(media_filepath thumbnail_filepath metadata_filepath subtitle_filepaths nfo_filepath)a
   end
+
+  @doc false
+  def filepath_attribute_defaults do
+    filepath_attributes()
+    |> Enum.map(fn
+      :subtitle_filepaths -> {:subtitle_filepaths, []}
+      field -> {field, nil}
+    end)
+    |> Enum.into(%{})
+  end
 end

lib/pinchflat/media/media_query.ex

@@ -0,0 +1,158 @@
+defmodule Pinchflat.Media.MediaQuery do
+  @moduledoc """
+  Query helpers for the Media context.
+
+  These methods are made to be one-ish liners used
+  to compose queries. Each method should strive to do
+  _one_ thing. These don't need to be tested as
+  they are just building blocks for other functionality
+  which, itself, will be tested.
+  """
+  import Ecto.Query, warn: false
+
+  alias Pinchflat.Media.MediaItem
+
+  # Prefixes:
+  # - for_* - belonging to a certain record
+  # - join_* - for joining on a certain record
+  # - with_* - for filtering based on full, concrete attributes
+  # - matching_* - for filtering based on partial attributes (e.g. LIKE, regex, full-text search)
+  #
+  # Suffixes:
+  # - _for - the arg passed is an association record
+
+  def new do
+    MediaItem
+  end
+
+  def for_source(query, source) do
+    where(query, [mi], mi.source_id == ^source.id)
+  end
+
+  def join_sources(query) do
+    from(mi in query, join: s in assoc(mi, :source), as: :sources)
+  end
+
+  def with_passed_retention_period(query) do
+    query
+    |> require_assoc(:source)
+    |> where(
+      [mi, source],
+      fragment(
+        "IFNULL(?, 0) > 0 AND DATETIME('now', '-' || ? || ' day') > ?",
+        source.retention_period_days,
+        source.retention_period_days,
+        mi.media_downloaded_at
+      )
+    )
+  end
+
+  def with_no_culling_prevention(query) do
+    where(query, [mi], mi.prevent_culling == false)
+  end
+
+  def with_id(query, id) do
+    where(query, [mi], mi.id == ^id)
+  end
+
+  def with_media_ids(query, media_ids) do
+    where(query, [mi], mi.media_id in ^media_ids)
+  end
+
+  def with_media_filepath(query) do
+    where(query, [mi], not is_nil(mi.media_filepath))
+  end
+
+  def with_no_media_filepath(query) do
+    where(query, [mi], is_nil(mi.media_filepath))
+  end
+
+  def with_upload_date_after_source_cutoff(query) do
+    query
+    |> require_assoc(:source)
+    |> where([mi, source], is_nil(source.download_cutoff_date) or mi.upload_date >= source.download_cutoff_date)
+  end
+
+  def with_no_prevented_download(query) do
+    where(query, [mi], mi.prevent_download == false)
+  end
+
+  def matching_source_title_regex(query) do
+    query
+    |> require_assoc(:source)
+    |> where(
+      [mi, source],
+      is_nil(source.title_filter_regex) or fragment("regexp_like(?, ?)", mi.title, source.title_filter_regex)
+    )
+  end
+
+  def matching_search_term(query, nil), do: query
+
+  def matching_search_term(query, term) do
+    from(mi in query,
+      join: mi_search_index in assoc(mi, :media_items_search_index),
+      where: fragment("media_items_search_index MATCH ?", ^term),
+      select_merge: %{
+        matching_search_term:
+          fragment("""
+            coalesce(snippet(media_items_search_index, 0, '[PF_HIGHLIGHT]', '[/PF_HIGHLIGHT]', '...', 20), '') ||
+            ' ' ||
+            coalesce(snippet(media_items_search_index, 1, '[PF_HIGHLIGHT]', '[/PF_HIGHLIGHT]', '...', 20), '')
+          """)
+      },
+      order_by: [desc: fragment("rank")]
+    )
+  end
+
+  def with_format_matching_profile_preference(query) do
+    query
+    |> require_assoc(:media_profile)
+    |> where(
+      fragment("""
+        CASE
+          WHEN shorts_behaviour = 'only' AND livestream_behaviour = 'only' THEN
+            livestream = true OR short_form_content = true
+          WHEN shorts_behaviour = 'only' THEN
+            short_form_content = true
+          WHEN livestream_behaviour = 'only' THEN
+            livestream = true
+          WHEN shorts_behaviour = 'exclude' AND livestream_behaviour = 'exclude' THEN
+            short_form_content = false AND livestream = false
+          WHEN shorts_behaviour = 'exclude' THEN
+            short_form_content = false
+          WHEN livestream_behaviour = 'exclude' THEN
+            livestream = false
+          ELSE
+            true
+        END
+      """)
+    )
+  end
+
+  def with_media_pending_download(query) do
+    query
+    |> with_no_prevented_download()
+    |> with_no_media_filepath()
+    |> with_upload_date_after_source_cutoff()
+    |> with_format_matching_profile_preference()
+    |> matching_source_title_regex()
+  end
+
+  defp require_assoc(query, identifier) do
+    if has_named_binding?(query, identifier) do
+      query
+    else
+      do_require_assoc(query, identifier)
+    end
+  end
+
+  defp do_require_assoc(query, :source) do
+    from(mi in query, join: s in assoc(mi, :source), as: :source)
+  end
+
+  defp do_require_assoc(query, :media_profile) do
+    query
+    |> require_assoc(:source)
+    |> join(:inner, [mi, source], mp in assoc(source, :media_profile), as: :media_profile)
+  end
+end

lib/pinchflat/metadata/metadata_parser.ex

@@ -30,7 +30,8 @@ defmodule Pinchflat.Metadata.MetadataParser do
       original_url: metadata["original_url"],
       description: metadata["description"],
       media_filepath: metadata["filepath"],
-      livestream: metadata["was_live"]
+      livestream: metadata["was_live"],
+      duration_seconds: metadata["duration"] && round(metadata["duration"])
     }
   end
 

lib/pinchflat/metadata/nfo_builder.ex

@@ -4,6 +4,8 @@ defmodule Pinchflat.Metadata.NfoBuilder do
   use by Kodi/Jellyfin and other media center software.
   """
 
+  import Pinchflat.Utils.XmlUtils, only: [safe: 1]
+
   alias Pinchflat.Metadata.MetadataFileHelpers
   alias Pinchflat.Filesystem.FilesystemHelpers
 
@@ -42,12 +44,12 @@ defmodule Pinchflat.Metadata.NfoBuilder do
     """
     <?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
     <episodedetails>
-      <title>#{metadata["title"]}</title>
-      <showtitle>#{metadata["uploader"]}</showtitle>
-      <uniqueid type="youtube" default="true">#{metadata["id"]}</uniqueid>
-      <plot>#{metadata["description"]}</plot>
-      <aired>#{upload_date}</aired>
-      <season>#{upload_date.year}</season>
+      <title>#{safe(metadata["title"])}</title>
+      <showtitle>#{safe(metadata["uploader"])}</showtitle>
+      <uniqueid type="youtube" default="true">#{safe(metadata["id"])}</uniqueid>
+      <plot>#{safe(metadata["description"])}</plot>
+      <aired>#{safe(upload_date)}</aired>
+      <season>#{safe(upload_date.year)}</season>
       <episode>#{Calendar.strftime(upload_date, "%m%d")}</episode>
       <genre>YouTube</genre>
     </episodedetails>
@@ -58,9 +60,9 @@ defmodule Pinchflat.Metadata.NfoBuilder do
     """
     <?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
     <tvshow>
-      <title>#{metadata["title"]}</title>
-      <plot>#{metadata["description"]}</plot>
-      <uniqueid type="youtube" default="true">#{metadata["id"]}</uniqueid>
+      <title>#{safe(metadata["title"])}</title>
+      <plot>#{safe(metadata["description"])}</plot>
+      <uniqueid type="youtube" default="true">#{safe(metadata["id"])}</uniqueid>
       <genre>YouTube</genre>
     </tvshow>
     """

lib/pinchflat/podcasts/podcast_helpers.ex

@@ -0,0 +1,78 @@
+defmodule Pinchflat.Podcasts.PodcastHelpers do
+  @moduledoc """
+  Methods for fetching postcast-related data from a source
+  or its media items
+  """
+
+  alias Pinchflat.Repo
+  alias Pinchflat.Media.MediaQuery
+  alias Pinchflat.Metadata.MediaMetadata
+  alias Pinchflat.Metadata.SourceMetadata
+
+  @doc """
+  Returns a list of media items that have been downloaded to disk
+  and have been proven to still exist there.
+
+  Useful for podcasts since we don't want to serve media that
+  has been deleted or moved, but it's also fairly generally useful
+  so I could see this being moved in the future.
+
+  Options:
+    - limit: integer - the maximum number of media items to return
+
+  Returns: [%MediaItem{}]
+  """
+  def persisted_media_items_for(source, opts \\ []) do
+    limit = Keyword.get(opts, :limit, 500)
+
+    MediaQuery.new()
+    |> MediaQuery.for_source(source)
+    |> MediaQuery.with_media_filepath()
+    |> Repo.maybe_limit(limit)
+    |> Repo.all()
+    |> Enum.filter(fn media_item -> File.exists?(media_item.media_filepath) end)
+  end
+
+  @doc """
+  Selects a cover image for a source based on the source's metadata
+  and the metadata of the media items associated with the source. Also
+  ensures images exist on disk.
+
+  Only one media item should need to be returned since this is using the
+  internal metadata which, so long as the media_item was _downloaded_, should
+  be guaranteed to exist.
+
+  Prefers the source's poster, then fanart, then the media item's thumbnail.
+
+  Returns: {:ok, filepath} | {:error, :no_suitable_image}
+  """
+  def select_cover_image(source, media_items) do
+    source_with_preloads = Repo.preload(source, :metadata)
+
+    source_with_preloads
+    |> get_images_by_preference(media_items)
+    |> Enum.reject(&is_nil(&1))
+    |> Enum.find(&File.exists?/1)
+    |> case do
+      nil -> {:error, :no_suitable_image}
+      filepath -> {:ok, filepath}
+    end
+  end
+
+  defp get_images_by_preference(source_with_preloads, []) do
+    source_metadata = source_with_preloads.metadata || %SourceMetadata{}
+
+    [
+      source_metadata.poster_filepath,
+      source_metadata.fanart_filepath
+    ]
+  end
+
+  defp get_images_by_preference(source_with_preloads, [media_item | _]) do
+    media_item_with_preloads = Repo.preload(media_item, :metadata)
+    media_item_metadata = media_item_with_preloads.metadata || %MediaMetadata{}
+    source_images = get_images_by_preference(source_with_preloads, [])
+
+    source_images ++ [media_item_metadata.thumbnail_filepath]
+  end
+end

lib/pinchflat/podcasts/rss_feed_builder.ex

@@ -0,0 +1,133 @@
+defmodule Pinchflat.Podcasts.RssFeedBuilder do
+  @moduledoc """
+  Methods for building an RSS feed for a source and its media items.
+  """
+
+  @datetime_format "%a, %d %b %Y %H:%M:%S %z"
+
+  import Pinchflat.Utils.XmlUtils, only: [safe: 1]
+
+  alias Pinchflat.Utils.DatetimeUtils
+  alias Pinchflat.Podcasts.PodcastHelpers
+  alias PinchflatWeb.Router.Helpers, as: Routes
+
+  @doc """
+  Builds an RSS feed for a given source and its media items.
+  Only MediaItems that have been persisted will be included in the feed.
+
+  ## Options:
+    - `:limit` - The maximum number of media items to include in the feed. Defaults to 300.
+
+  Returns an XML document as a string.
+  """
+  def build(source, opts \\ []) do
+    limit = Keyword.get(opts, :limit, 300)
+    url_base = Keyword.get(opts, :url_base, PinchflatWeb.Endpoint.url())
+
+    media_items = PodcastHelpers.persisted_media_items_for(source, limit: limit)
+    build_source_xml(source, media_items, url_base)
+  end
+
+  defp build_source_xml(source, media_items, url_base) do
+    media_item_xml = Enum.map(media_items, &build_media_item_xml(source, &1, url_base))
+    # "caching" the image path since it requires some DB calls and is used twice
+    feed_image_path = feed_image_path(url_base, source, media_items)
+
+    # Useful: resources:
+    #   - https://validator.w3.org/feed/#validate_by_input
+    #   - https://github.com/Podcastindex-org/podcast-namespace/blob/main/docs/1.0.md
+    #   - https://podba.se/validate
+    """
+    <?xml version="1.0" encoding="UTF-8"?>
+    <rss version="2.0"
+      xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd"
+      xmlns:podcast="https://podcastindex.org/namespace/1.0"
+      xmlns:atom="http://www.w3.org/2005/Atom">
+      <channel>
+        <title>#{safe(source.custom_name)}</title>
+        <link>#{safe(source.original_url)}</link>
+        <description>#{safe(source.description)}</description>
+        <category>TV &amp; Film</category>
+        <generator>Generated by Pinchflat</generator>
+        <language>en-us</language>
+        <lastBuildDate>#{Calendar.strftime(source.updated_at, @datetime_format)}</lastBuildDate>
+        <pubDate>#{Calendar.strftime(source.inserted_at, @datetime_format)}</pubDate>
+        <atom:link href="#{safe(generate_self_link(url_base, source))}" rel="self" type="application/rss+xml" />
+        <podcast:locked>yes</podcast:locked>
+        <podcast:guid>#{source.uuid}</podcast:guid>
+        <image>
+          <url>#{safe(feed_image_path)}</url>
+          <title>#{safe(source.custom_name)}</title>
+          <link>#{safe(source.original_url)}</link>
+        </image>
+        <itunes:author>#{safe(source.custom_name)}</itunes:author>
+        <itunes:subtitle>#{safe(source.custom_name)}</itunes:subtitle>
+        <itunes:block>yes</itunes:block>
+        <itunes:image href="#{safe(feed_image_path)}"></itunes:image>
+        <itunes:explicit>false</itunes:explicit>
+        <itunes:category text="TV &amp; Film"></itunes:category>
+
+        #{Enum.join(media_item_xml, "\n")}
+
+      </channel>
+    </rss>
+    """
+  end
+
+  defp build_media_item_xml(source, media_item, url_base) do
+    """
+    <item>
+      <guid isPermaLink="false">#{media_item.uuid}</guid>
+      <title>#{safe(media_item.title)}</title>
+      <link>#{safe(media_item.original_url)}</link>
+      <description>#{safe(media_item.description)}</description>
+      <pubDate>#{generate_upload_date(media_item)}</pubDate>
+      <itunes:duration>#{media_item.duration_seconds}</itunes:duration>
+      <enclosure
+        url="#{media_stream_path(url_base, media_item)}"
+        length="#{media_item.media_size_bytes}"
+        type="#{MIME.from_path(media_item.media_filepath)}"
+      />
+      <itunes:author>#{safe(source.custom_name)}</itunes:author>
+      <itunes:subtitle>#{safe(media_item.title)}</itunes:subtitle>
+      <itunes:summary><![CDATA[#{media_item.description}]]></itunes:summary>
+      <itunes:explicit>false</itunes:explicit>
+    </item>
+    """
+  end
+
+  defp generate_self_link(url_base, source) do
+    Path.join(url_base, "#{podcast_route(:rss_feed, source.uuid)}.xml")
+  end
+
+  defp media_stream_path(url_base, media_item) do
+    extension = Path.extname(media_item.media_filepath)
+
+    Path.join(url_base, "#{media_route(:stream, media_item.uuid)}#{extension}")
+  end
+
+  defp feed_image_path(url_base, source, media_items) do
+    case PodcastHelpers.select_cover_image(source, media_items) do
+      {:error, _} ->
+        ""
+
+      {:ok, filepath} ->
+        extension = Path.extname(filepath)
+        Path.join(url_base, "#{podcast_route(:feed_image, source.uuid)}#{extension}")
+    end
+  end
+
+  defp generate_upload_date(media_item) do
+    media_item.upload_date
+    |> DatetimeUtils.date_to_datetime()
+    |> Calendar.strftime(@datetime_format)
+  end
+
+  defp podcast_route(action, params) do
+    Routes.podcast_path(PinchflatWeb.Endpoint, action, params)
+  end
+
+  defp media_route(action, params) do
+    Routes.media_item_path(PinchflatWeb.Endpoint, action, params)
+  end
+end

lib/pinchflat/profiles/media_profile.ex

@@ -21,6 +21,8 @@ defmodule Pinchflat.Profiles.MediaProfile do
     download_metadata
     embed_metadata
     download_nfo
+    sponsorblock_behaviour
+    sponsorblock_categories
     shorts_behaviour
     livestream_behaviour
     preferred_resolution
@@ -47,6 +49,8 @@ defmodule Pinchflat.Profiles.MediaProfile do
     field :embed_metadata, :boolean, default: false
 
     field :download_nfo, :boolean, default: false
+    field :sponsorblock_behaviour, Ecto.Enum, values: [:disabled, :remove], default: :disabled
+    field :sponsorblock_categories, {:array, :string}, default: []
     # NOTE: these do NOT speed up indexing - the indexer still has to go
     # through the entire collection to determine if a media is a short or
     # a livestream.

lib/pinchflat/release.ex

@@ -25,14 +25,15 @@ defmodule Pinchflat.Release do
   def check_file_permissions do
     load_app()
 
-    directories = [
+    directories =
+      Enum.uniq([
         "/config",
         "/downloads",
         Application.get_env(:pinchflat, :media_directory),
         Application.get_env(:pinchflat, :tmpfile_directory),
         Application.get_env(:pinchflat, :extras_directory),
         Application.get_env(:pinchflat, :metadata_directory)
-    ]
+      ])
 
     Enum.each(directories, fn dir ->
       Logger.info("Checking permissions for #{dir}")

lib/pinchflat/settings/setting.ex

@@ -1,24 +1,32 @@
 defmodule Pinchflat.Settings.Setting do
   @moduledoc """
-  A Setting is a key-value pair with a datatype used to track user-level settings.
+  The Setting schema.
   """
 
   use Ecto.Schema
   import Ecto.Changeset
 
-  schema "settings" do
-    field :name, :string
-    field :value, :string
-    field :datatype, Ecto.Enum, values: ~w(boolean string integer float)a
+  @allowed_fields [
+    :onboarding,
+    :pro_enabled,
+    :yt_dlp_version
+  ]
 
-    timestamps(type: :utc_datetime)
+  @required_fields ~w(
+    onboarding
+    pro_enabled
+  )a
+
+  schema "settings" do
+    field :onboarding, :boolean, default: true
+    field :pro_enabled, :boolean, default: false
+    field :yt_dlp_version, :string
   end
 
   @doc false
   def changeset(setting, attrs) do
     setting
-    |> cast(attrs, [:name, :value, :datatype])
-    |> validate_required([:name, :value, :datatype])
-    |> unique_constraint([:name])
+    |> cast(attrs, @allowed_fields)
+    |> validate_required(@required_fields)
   end
 end

lib/pinchflat/settings/settings.ex

@@ -2,94 +2,63 @@ defmodule Pinchflat.Settings do
   @moduledoc """
   The Settings context.
   """
-
   import Ecto.Query, warn: false
-  alias Pinchflat.Repo
 
+  alias Pinchflat.Repo
   alias Pinchflat.Settings.Setting
 
   @doc """
-  Returns the list of settings.
+  Returns the only setting record. It _should_ be impossible
+  to create or delete this record, so it's assertive about
+  assuming it's the only one.
 
-  Returns [%Setting{}, ...]
+  Returns %Setting{}
   """
-  def list_settings do
-    Repo.all(Setting)
+  def record do
+    Setting
+    |> limit(1)
+    |> Repo.one()
   end
 
   @doc """
-  Creates or updates a setting, returning the parsed value.
-  Raises if an unsupported datatype is used. Optionally allows
-  specifying the datatype.
+  Updates a setting, returning the new value.
+  Is setup to take a keyword list argument so you
+  can call it like `Settings.set(onboarding: true)`
 
-  Returns value in type of `Ecto.Enum.mappings(Setting, :datatype)`
+  Returns {:ok, value} | {:error, :invalid_key} | {:error, %Ecto.Changeset{}}
   """
-  def set!(name, value) do
-    set!(name, value, infer_datatype(value))
-  end
-
-  def set!(name, value, datatype) do
-    # Only create if doesn't exist
-    case Repo.get_by(Setting, name: to_string(name)) do
-      nil -> create_setting!(name, value, datatype)
-      setting -> update_setting!(setting, value, datatype)
+  def set([{attr, value}]) do
+    record()
+    |> Setting.changeset(%{attr => value})
+    |> Repo.update()
+    |> case do
+      {:ok, %{^attr => _}} -> {:ok, value}
+      {:ok, _} -> {:error, :invalid_key}
+      {:error, changeset} -> {:error, changeset}
     end
   end
 
   @doc """
-  Gets the parsed value of a setting. Raises if the setting does not exist.
+  Gets the value of a setting.
 
-  Returns value in type of `Ecto.Enum.mappings(Setting, :datatype)`
+  Returns {:ok, value} | {:error, :invalid_key}
+  """
+  def get(name) do
+    case Map.fetch(record(), name) do
+      {:ok, value} -> {:ok, value}
+      :error -> {:error, :invalid_key}
+    end
+  end
+
+  @doc """
+  Gets the value of a setting, raising if it doesn't exist.
+
+  Returns value
   """
   def get!(name) do
-    Setting
-    |> Repo.get_by!(name: to_string(name))
-    |> read_setting()
-  end
-
-  @doc """
-  Attempts to find a setting by name or creates a setting with value
-  if one doesn't exist, returning the parsed value. Optionally allows
-  specifying the datatype.
-
-  Returns value in type of `Ecto.Enum.mappings(Setting, :datatype)`
-  """
-  def fetch!(name, value) do
-    fetch!(name, value, infer_datatype(value))
-  end
-
-  def fetch!(name, value, datatype) do
-    case Repo.get_by(Setting, name: to_string(name)) do
-      nil -> create_setting!(name, value, datatype)
-      setting -> read_setting(setting)
+    case get(name) do
+      {:ok, value} -> value
+      {:error, _} -> raise "Setting `#{name}` not found"
     end
   end
-
-  defp change_setting(setting, attrs) do
-    Setting.changeset(setting, attrs)
-  end
-
-  defp create_setting!(name, value, datatype) do
-    %Setting{}
-    |> change_setting(%{name: to_string(name), value: to_string(value), datatype: datatype})
-    |> Repo.insert!()
-    |> read_setting()
-  end
-
-  defp update_setting!(setting, value, datatype) do
-    setting
-    |> change_setting(%{value: to_string(value), datatype: datatype})
-    |> Repo.update!()
-    |> read_setting()
-  end
-
-  defp read_setting(%{value: value, datatype: :string}), do: value
-  defp read_setting(%{value: value, datatype: :boolean}), do: value in ["true", "t", "1"]
-  defp read_setting(%{value: value, datatype: :integer}), do: String.to_integer(value)
-  defp read_setting(%{value: value, datatype: :float}), do: String.to_float(value)
-
-  defp infer_datatype(value) when is_boolean(value), do: :boolean
-  defp infer_datatype(value) when is_integer(value), do: :integer
-  defp infer_datatype(value) when is_float(value), do: :float
-  defp infer_datatype(value) when is_binary(value), do: :string
 end

lib/pinchflat/slow_indexing/media_collection_indexing_worker.ex

@@ -20,9 +20,10 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorker do
 
   Returns {:ok, %Task{}} | {:error, :duplicate_job} | {:error, %Ecto.Changeset{}}
   """
-  def kickoff_with_task(source, opts \\ []) do
+  def kickoff_with_task(source, job_args \\ %{}, job_opts \\ []) do
     %{id: source.id}
-    |> MediaCollectionIndexingWorker.new(opts)
+    |> Map.merge(job_args)
+    |> MediaCollectionIndexingWorker.new(job_opts)
     |> Tasks.create_job_with_task(source)
   end
 
@@ -30,8 +31,8 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorker do
   The ID is that of a source _record_, not a YouTube channel/playlist ID. Indexes
   the provided source, kicks off downloads for each new MediaItem, and
   reschedules the job to run again in the future. It will ALWAYS index a source
-  if it's never been indexed before, but rescheduling is determined by the
-  `index_frequency_minutes` field.
+  if it's never been indexed before or if `force` is set to `true`, but rescheduling
+  is determined by the `index_frequency_minutes` field.
 
   README: Re-scheduling here works a little different than you may expect.
   The reschedule time is relative to the time the job has actually _completed_.
@@ -71,7 +72,7 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorker do
   Returns :ok | {:ok, %Task{}}
   """
   @impl Oban.Worker
-  def perform(%Oban.Job{args: %{"id" => source_id}}) do
+  def perform(%Oban.Job{args: %{"id" => source_id} = args}) do
     source = Sources.get_source!(source_id)
 
     case {source.index_frequency_minutes, source.last_indexed_at} do
@@ -89,7 +90,11 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorker do
 
       _ ->
         # If the source HAS been indexed and is not meant to reschedule,
-        # perform a no-op
+        # perform a no-op (unless forced)
+        if args["force"] do
+          SlowIndexingHelpers.index_and_enqueue_download_for_media_items(source)
+        end
+
         :ok
     end
   rescue

lib/pinchflat/slow_indexing/slow_indexing_helpers.ex

@@ -27,12 +27,12 @@ defmodule Pinchflat.SlowIndexing.SlowIndexingHelpers do
 
   Returns {:ok, %Task{}}.
   """
-  def kickoff_indexing_task(%Source{} = source) do
+  def kickoff_indexing_task(%Source{} = source, job_args \\ %{}, job_opts \\ []) do
     Tasks.delete_pending_tasks_for(source, "FastIndexingWorker")
     Tasks.delete_pending_tasks_for(source, "MediaIndexingWorker")
     Tasks.delete_pending_tasks_for(source, "MediaCollectionIndexingWorker")
 
-    MediaCollectionIndexingWorker.kickoff_with_task(source)
+    MediaCollectionIndexingWorker.kickoff_with_task(source, job_args, job_opts)
   end
 
   @doc """

lib/pinchflat/sources/source.ex

@@ -29,6 +29,7 @@ defmodule Pinchflat.Sources.Source do
     last_indexed_at
     original_url
     download_cutoff_date
+    retention_period_days
     title_filter_regex
     media_profile_id
   )a
@@ -48,6 +49,7 @@ defmodule Pinchflat.Sources.Source do
 
   @pre_insert_required_fields @initially_required_fields ++
                                 ~w(
+                                  uuid
                                   custom_name
                                   collection_name
                                   collection_id
@@ -55,6 +57,11 @@ defmodule Pinchflat.Sources.Source do
                                 )a
 
   schema "sources" do
+    # This is _not_ used as the primary key or internally in the database
+    # relations. This is only used to prevent an enumeration attack on the streaming
+    # and RSS feed endpoints since those _must_ be public (ie: no basic auth)
+    field :uuid, Ecto.UUID
+
     field :custom_name, :string
     field :description, :string
     field :collection_name, :string
@@ -66,6 +73,7 @@ defmodule Pinchflat.Sources.Source do
     field :last_indexed_at, :utc_datetime
     # Only download media items that were published after this date
     field :download_cutoff_date, :date
+    field :retention_period_days, :integer
     field :original_url, :string
     field :title_filter_regex, :string
 
@@ -98,8 +106,11 @@ defmodule Pinchflat.Sources.Source do
     source
     |> cast(attrs, @allowed_fields)
     |> dynamic_default(:custom_name, fn cs -> get_field(cs, :collection_name) end)
+    |> dynamic_default(:uuid, fn _ -> Ecto.UUID.generate() end)
     |> validate_required(required_fields)
+    |> validate_number(:retention_period_days, greater_than_or_equal_to: 0)
     |> cast_assoc(:metadata, with: &SourceMetadata.changeset/2, required: false)
+    |> unique_constraint([:collection_id, :media_profile_id, :title_filter_regex], error_key: :original_url)
   end
 
   @doc false

lib/pinchflat/sources/sources.ex

@@ -9,6 +9,7 @@ defmodule Pinchflat.Sources do
   alias Pinchflat.Media
   alias Pinchflat.Tasks
   alias Pinchflat.Sources.Source
+  alias Pinchflat.Media.MediaQuery
   alias Pinchflat.Profiles.MediaProfile
   alias Pinchflat.YtDlp.MediaCollection
   alias Pinchflat.Metadata.SourceMetadata
@@ -112,8 +113,9 @@ defmodule Pinchflat.Sources do
     delete_files = Keyword.get(opts, :delete_files, false)
     Tasks.delete_tasks_for(source)
 
-    source
-    |> Media.list_media_items_for()
+    MediaQuery.new()
+    |> MediaQuery.for_source(source)
+    |> Repo.all()
     |> Enum.each(fn media_item ->
       Media.delete_media_item(media_item, delete_files: delete_files)
     end)

lib/pinchflat/sources/sources_query.ex

@@ -0,0 +1,31 @@
+defmodule Pinchflat.Sources.SourcesQuery do
+  @moduledoc """
+  Query helpers for the Sources context.
+
+  These methods are made to be one-ish liners used
+  to compose queries. Each method should strive to do
+  _one_ thing. These don't need to be tested as
+  they are just building blocks for other functionality
+  which, itself, will be tested.
+  """
+  import Ecto.Query, warn: false
+
+  alias Pinchflat.Sources.Source
+
+  # Prefixes:
+  # - for_* - belonging to a certain record
+  # - join_* - for joining on a certain record
+  # - with_* - for filtering based on full, concrete attributes
+  # - matching_* - for filtering based on partial attributes (e.g. LIKE, regex, full-text search)
+  #
+  # Suffixes:
+  # - _for - the arg passed is an association record
+
+  def new do
+    Source
+  end
+
+  def for_media_profile(query, media_profile) do
+    where(query, [s], s.media_profile_id == ^media_profile.id)
+  end
+end

lib/pinchflat/utils/datetime_utils.ex

@@ -0,0 +1,17 @@
+defmodule Pinchflat.Utils.DatetimeUtils do
+  @moduledoc """
+  Utility methods for working with dates and datetimes
+  """
+
+  @doc """
+  Converts a Date to a DateTime
+
+  Returns %DateTime{}
+  """
+  def date_to_datetime(date) do
+    date
+    |> Date.to_gregorian_days()
+    |> Kernel.*(86_400)
+    |> DateTime.from_gregorian_seconds()
+  end
+end

lib/pinchflat/utils/xml_utils.ex

@@ -0,0 +1,17 @@
+defmodule Pinchflat.Utils.XmlUtils do
+  @moduledoc """
+  Utility methods for working with XML documents
+  """
+
+  @doc """
+  Escapes invalid XML characters in a string
+
+  Returns binary()
+  """
+  def safe(value) do
+    value
+    |> to_string()
+    |> Phoenix.HTML.html_escape()
+    |> Phoenix.HTML.safe_to_string()
+  end
+end

lib/pinchflat/yt_dlp/backend_command_runner.ex

@@ -8,4 +8,5 @@ defmodule Pinchflat.YtDlp.BackendCommandRunner do
 
   @callback run(binary(), keyword(), binary()) :: {:ok, binary()} | {:error, binary(), integer()}
   @callback run(binary(), keyword(), binary(), keyword()) :: {:ok, binary()} | {:error, binary(), integer()}
+  @callback version() :: {:ok, binary()} | {:error, binary()}
 end

lib/pinchflat/yt_dlp/command_runner.ex

@@ -29,7 +29,7 @@ defmodule Pinchflat.YtDlp.CommandRunner do
     command = backend_executable()
     # These must stay in exactly this order, hence why I'm giving it its own variable.
     # Also, can't use RAM file since yt-dlp needs a concrete filepath.
-    output_filepath = Keyword.get(addl_opts, :output_filepath, FSUtils.generate_metadata_tmpfile(:json))
+    output_filepath = generate_output_filepath(addl_opts)
     print_to_file_opts = [{:print_to_file, output_template}, output_filepath]
     cookie_opts = build_cookie_options()
     formatted_command_opts = [url] ++ parse_options(command_opts ++ print_to_file_opts ++ cookie_opts)
@@ -48,6 +48,26 @@ defmodule Pinchflat.YtDlp.CommandRunner do
     end
   end
 
+  @impl BackendCommandRunner
+  def version do
+    command = backend_executable()
+
+    case System.cmd(command, ["--version"]) do
+      {output, 0} ->
+        {:ok, String.trim(output)}
+
+      {output, _} ->
+        {:error, output}
+    end
+  end
+
+  defp generate_output_filepath(addl_opts) do
+    case Keyword.get(addl_opts, :output_filepath) do
+      nil -> FSUtils.generate_metadata_tmpfile(:json)
+      path -> path
+    end
+  end
+
   defp build_cookie_options do
     base_dir = Application.get_env(:pinchflat, :extras_directory)
     cookie_file = Path.join(base_dir, "cookies.txt")

lib/pinchflat/yt_dlp/media.ex

@@ -10,7 +10,8 @@ defmodule Pinchflat.YtDlp.Media do
     :original_url,
     :livestream,
     :short_form_content,
-    :upload_date
+    :upload_date,
+    :duration_seconds
   ]
 
   defstruct [
@@ -20,7 +21,8 @@ defmodule Pinchflat.YtDlp.Media do
     :original_url,
     :livestream,
     :short_form_content,
-    :upload_date
+    :upload_date,
+    :duration_seconds
   ]
 
   alias __MODULE__
@@ -33,10 +35,10 @@ defmodule Pinchflat.YtDlp.Media do
 
   Returns {:ok, map()} | {:error, any, ...}.
   """
-  def download(url, command_opts \\ []) do
+  def download(url, command_opts \\ [], addl_opts \\ []) do
     opts = [:no_simulate] ++ command_opts
 
-    with {:ok, output} <- backend_runner().run(url, opts, "after_move:%()j"),
+    with {:ok, output} <- backend_runner().run(url, opts, "after_move:%()j", addl_opts),
          {:ok, parsed_json} <- Phoenix.json_library().decode(output) do
       {:ok, parsed_json}
     else
@@ -86,6 +88,7 @@ defmodule Pinchflat.YtDlp.Media do
       description: response["description"],
       original_url: response["webpage_url"],
       livestream: response["was_live"],
+      duration_seconds: response["duration"] && round(response["duration"]),
       short_form_content: response["webpage_url"] && short_form_content?(response),
       upload_date: response["upload_date"] && MetadataFileHelpers.parse_upload_date(response["upload_date"])
     }

lib/pinchflat_web.ex

@@ -21,7 +21,7 @@ defmodule PinchflatWeb do
 
   def router do
     quote do
-      use Phoenix.Router, helpers: false
+      use Phoenix.Router, helpers: true
 
       # Import common connection and controller functions to use in pipelines
       import Plug.Conn

lib/pinchflat_web/components/core_components.ex

@@ -14,7 +14,7 @@ defmodule PinchflatWeb.CoreComponents do
 
   Icons are provided by [heroicons](https://heroicons.com). See `icon/1` for usage.
   """
-  use Phoenix.Component
+  use Phoenix.Component, global_prefixes: ~w(x-)
 
   import PinchflatWeb.Gettext
 
@@ -251,7 +251,7 @@ defmodule PinchflatWeb.CoreComponents do
   attr :type, :string,
     default: "text",
     values: ~w(checkbox color date datetime-local email file hidden month number password
-               toggle range radio search select tel text textarea time url week)
+            checkbox_group toggle range radio search select tel text textarea time url week)
 
   attr :field, Phoenix.HTML.FormField, doc: "a form field struct retrieved from the form, for example: @form[:email]"
 
@@ -304,6 +304,33 @@ defmodule PinchflatWeb.CoreComponents do
     """
   end
 
+  def input(%{type: "checkbox_group"} = assigns) do
+    ~H"""
+    <div phx-feedback-for={@name}>
+      <.label for={@id}>
+        <%= @label %><span :if={@label_suffix} class="text-xs text-bodydark"><%= @label_suffix %></span>
+      </.label>
+      <section class="grid grid-cols-1 gap-2 md:grid-cols-2 max-w-prose mb-4 ml-1">
+        <div :for={{option_name, option_value} <- @options} class="flex items-center">
+          <input
+            type="checkbox"
+            id={"#{@id}-#{option_value}"}
+            name={"#{@name}[]"}
+            value={option_value}
+            checked={option_value in @value}
+            class={["rounded focus:ring-offset-0 ring-offset-0 focus:ring-0 h-5 w-5 ", @inputclass]}
+          />
+          <label for={"#{@id}-#{option_value}"} class="ml-2 cursor-pointer select-none">
+            <%= option_name %>
+          </label>
+        </div>
+      </section>
+      <.help :if={@help}><%= @help %></.help>
+      <.error :for={msg <- @errors}><%= msg %></.error>
+    </div>
+    """
+  end
+
   def input(%{type: "toggle"} = assigns) do
     assigns =
       assign_new(assigns, :checked, fn ->
@@ -347,8 +374,9 @@ defmodule PinchflatWeb.CoreComponents do
           id={@id}
           name={@name}
           class={[
-            "relative z-20 w-full appearance-none rounded border border-stroke bg-transparent py-3 pl-5 pr-12 outline-none transition",
-            "focus:border-primary active:border-primary dark:border-form-strokedark dark:bg-form-input text-black dark:text-white",
+            "relative z-20 w-full appearance-none rounded border border-form-strokedark py-3 pl-5 pr-12 outline-none transition",
+            "focus:border-primary active:border-primary bg-form-input text-black text-white",
+            "disabled:text-opacity-50 disabled:cursor-not-allowed disabled:border-black",
             @inputclass
           ]}
           multiple={@multiple}
@@ -600,6 +628,10 @@ defmodule PinchflatWeb.CoreComponents do
         _ ->
           true
       end)
+      |> Enum.map(fn
+        {k, v} when is_list(v) -> {k, Enum.join(v, ", ")}
+        rest -> rest
+      end)
 
     assigns = assign(assigns, iterable_attributes: attrs)
 
@@ -654,10 +686,11 @@ defmodule PinchflatWeb.CoreComponents do
   """
   attr :name, :string, required: true
   attr :class, :string, default: nil
+  attr :rest, :global
 
   def icon(%{name: "hero-" <> _} = assigns) do
     ~H"""
-    <span class={[@name, @class]} />
+    <span class={[@name, @class]} {@rest} />
     """
   end
 

lib/pinchflat_web/components/custom_components/button_components.ex

@@ -1,6 +1,8 @@
 defmodule PinchflatWeb.CustomComponents.ButtonComponents do
   @moduledoc false
-  use Phoenix.Component
+  use Phoenix.Component, global_prefixes: ~w(x-)
+
+  alias PinchflatWeb.CoreComponents
 
   @doc """
   Render a button
@@ -39,4 +41,52 @@ defmodule PinchflatWeb.CustomComponents.ButtonComponents do
     </button>
     """
   end
+
+  @doc """
+  Render a dropdown based off a button
+
+  ## Examples
+
+      <.button_dropdown text="Actions">
+        <:option>TEST</:option>
+      </.button_dropdown>
+  """
+  attr :text, :string, required: true
+  attr :class, :string, default: ""
+
+  slot :option, required: true
+
+  def button_dropdown(assigns) do
+    ~H"""
+    <div x-data="{ dropdownOpen: false }" class={["relative flex", @class]}>
+      <span
+        x-on:click.prevent="dropdownOpen = !dropdownOpen"
+        class={[
+          "cursor-pointer inline-flex gap-2.5 rounded-md bg-primary px-5.5 py-3",
+          "font-medium text-white hover:bg-opacity-95"
+        ]}
+      >
+        <%= @text %>
+        <CoreComponents.icon
+          name="hero-chevron-down"
+          class="fill-current duration-200 ease-linear mt-1"
+          x-bind:class="dropdownOpen && 'rotate-180'"
+        />
+      </span>
+      <div
+        x-show="dropdownOpen"
+        x-on:click.outside="dropdownOpen = false"
+        class="absolute left-0 top-full z-40 mt-2 w-full rounded-md bg-black py-3 shadow-card"
+      >
+        <ul class="flex flex-col">
+          <li :for={option <- @option}>
+            <span class="flex px-5 py-2 font-medium text-bodydark2 hover:text-white cursor-pointer">
+              <%= render_slot(option) %>
+            </span>
+          </li>
+        </ul>
+      </div>
+    </div>
+    """
+  end
 end

lib/pinchflat_web/components/custom_components/tab_components.ex

@@ -9,24 +9,32 @@ defmodule PinchflatWeb.CustomComponents.TabComponents do
     attr :title, :string, required: true
   end
 
+  slot :tab_append, required: false
+
   def tabbed_layout(assigns) do
     ~H"""
     <div
-      x-data="{ openTab: 0, activeClasses: 'text-meta-5 border-meta-5', inactiveClasses: 'border-transparent' }"
+      x-data="{ openTab: getTabIndex(0), activeClasses: 'text-meta-5 border-meta-5', inactiveClasses: 'border-transparent' }"
+      @hashchange.window="openTab = getTabIndex(openTab)"
       class="w-full"
     >
-      <div class="mb-6 flex flex-wrap gap-5 border-b border-strokedark sm:gap-10">
+      <header class="flex flex-col md:flex-row md:justify-between border-b border-strokedark">
+        <div class="flex flex-wrap gap-5 sm:gap-10">
           <a
             :for={{tab, idx} <- Enum.with_index(@tab)}
             href="#"
-          @click.prevent={"openTab = #{idx}"}
+            @click.prevent={"openTab = setTabIndex(#{idx})"}
             x-bind:class={"openTab === #{idx} ? activeClasses : inactiveClasses"}
-          class="border-b-2 py-4 text-sm font-medium hover:text-meta-5 md:text-base"
+            class="border-b-2 py-4 w-full sm:w-fit text-sm font-medium hover:text-meta-5 md:text-base"
           >
             <span class="text-xl"><%= tab.title %></span>
           </a>
         </div>
-      <div>
+        <div class="mx-4 my-4 lg:my-0 flex gap-5 sm:gap-10 items-center">
+          <%= render_slot(@tab_append) %>
+        </div>
+      </header>
+      <div class="mt-4 min-h-60">
         <div :for={{tab, idx} <- Enum.with_index(@tab)} x-show={"openTab === #{idx}"} class="font-medium leading-relaxed">
           <%= render_slot(tab) %>
         </div>

lib/pinchflat_web/components/layouts/partials/sidebar.html.heex

@@ -58,8 +58,11 @@
           </span>
         </li>
         <li>
-          <span class="group relative flex items-center gap-2.5 px-4 py-2 text-sm">
-            v<%= Application.spec(:pinchflat)[:vsn] %>
+          <span class="group relative flex items-center gap-2.5 px-4 pt-2 text-sm">
+            Pinchflat v<%= Application.spec(:pinchflat)[:vsn] %>
+          </span>
+          <span class="group relative flex items-center gap-2.5 px-4 text-sm">
+            yt-dlp <%= Settings.get!(:yt_dlp_version) %>
           </span>
         </li>
       </ul>

lib/pinchflat_web/components/layouts/partials/upgrade_button_live.ex

@@ -30,7 +30,7 @@ defmodule Pinchflat.UpgradeButtonLive do
       |> String.downcase()
 
     if normalized_text == "got it!" do
-      Settings.set!(:pro_enabled, true)
+      Settings.set(pro_enabled: true)
 
       {:noreply, update(socket, :button_disabled, fn _ -> false end)}
     else

lib/pinchflat_web/controllers/health_controller.ex

@@ -0,0 +1,9 @@
+defmodule PinchflatWeb.HealthController do
+  use PinchflatWeb, :controller
+
+  def check(conn, _params) do
+    conn
+    |> put_status(:ok)
+    |> json(%{status: "ok"})
+  end
+end

lib/pinchflat_web/controllers/media_items/media_item_controller.ex

@@ -1,8 +1,12 @@
 defmodule PinchflatWeb.MediaItems.MediaItemController do
   use PinchflatWeb, :controller
 
+  require Logger
+
   alias Pinchflat.Repo
   alias Pinchflat.Media
+  alias Pinchflat.Media.MediaItem
+  alias Pinchflat.Downloading.MediaDownloadWorker
 
   def show(conn, %{"id" => id}) do
     media_item =
@@ -13,20 +17,108 @@ defmodule PinchflatWeb.MediaItems.MediaItemController do
     render(conn, :show, media_item: media_item)
   end
 
-  def delete(conn, %{"id" => id} = params) do
-    delete_files = Map.get(params, "delete_files", false)
+  def edit(conn, %{"id" => id}) do
     media_item = Media.get_media_item!(id)
-    {:ok, _} = Media.delete_media_item(media_item, delete_files: delete_files)
+    changeset = Media.change_media_item(media_item)
 
-    flash_message =
-      if delete_files do
-        "Record and files deleted successfully."
-      else
-        "Record deleted successfully. Files were not deleted."
+    render(conn, :edit, media_item: media_item, changeset: changeset)
   end
 
+  def update(conn, %{"id" => id, "media_item" => params}) do
+    media_item = Media.get_media_item!(id)
+
+    case Media.update_media_item(media_item, params) do
+      {:ok, media_item} ->
+        conn
+        |> put_flash(:info, "Media Item updated successfully.")
+        |> redirect(to: ~p"/sources/#{media_item.source_id}/media/#{media_item}")
+
+      {:error, %Ecto.Changeset{} = changeset} ->
+        render(conn, :edit, media_item: media_item, changeset: changeset)
+    end
+  end
+
+  def delete(conn, %{"id" => id} = params) do
+    prevent_download = Map.get(params, "prevent_download", false)
+    media_item = Media.get_media_item!(id)
+    {:ok, _} = Media.delete_media_files(media_item, %{prevent_download: prevent_download})
+
     conn
-    |> put_flash(:info, flash_message)
+    |> put_flash(:info, "Files deleted successfully.")
     |> redirect(to: ~p"/sources/#{media_item.source_id}")
   end
+
+  def force_download(conn, %{"media_item_id" => id}) do
+    media_item = Media.get_media_item!(id)
+    {:ok, _} = MediaDownloadWorker.kickoff_with_task(media_item, %{force: true})
+
+    conn
+    |> put_flash(:info, "Download task enqueued.")
+    |> redirect(to: ~p"/sources/#{media_item.source_id}/media/#{media_item}")
+  end
+
+  # See here for details on streaming files and range requests:
+  # https://www.zeng.dev/post/2023-http-range-and-play-mp4-in-browser/
+  #
+  # Uses the UUID instead of the ID to avoid enumeration attacks
+  # since streaming is a public endpoint (ie: no auth required)
+  def stream(conn, %{"uuid" => uuid}) do
+    media_item = Repo.get_by!(MediaItem, uuid: uuid)
+
+    if File.exists?(media_item.media_filepath) do
+      file_size = File.stat!(media_item.media_filepath).size
+      mime_type = MIME.from_path(media_item.media_filepath)
+
+      case parse_range(conn, file_size) do
+        {:ok, {start_pos, end_pos}} ->
+          Logger.debug("Streaming media item: #{media_item.uuid} from #{start_pos} to #{end_pos}")
+          length = end_pos - start_pos + 1
+
+          conn
+          |> put_resp_content_type(mime_type)
+          |> put_resp_header("accept-ranges", "bytes")
+          |> put_resp_header("content-range", "bytes #{start_pos}-#{end_pos}/#{file_size}")
+          |> put_resp_header("content-length", to_string(length))
+          |> send_file(206, media_item.media_filepath, start_pos, length)
+
+        {:error, :invalid_range} ->
+          Logger.debug("Invalid range request for media item: #{media_item.uuid} - serving full file")
+
+          conn
+          |> put_resp_content_type(mime_type)
+          |> put_resp_header("content-length", to_string(file_size))
+          |> put_resp_header("accept-ranges", "bytes")
+          |> send_file(200, media_item.media_filepath)
+      end
+    else
+      send_resp(conn, 404, "File not found")
+    end
+  end
+
+  defp parse_range(conn, file_size) do
+    with [range_header | _] <- get_req_header(conn, "range"),
+         ["bytes", range] <- String.split(range_header, "="),
+         [start_pos, end_pos] <- String.split(range, "-") do
+      validate_range(start_pos, end_pos, file_size)
+    else
+      _ -> {:error, :invalid_range}
+    end
+  end
+
+  defp validate_range(start_pos, end_pos, file_size) do
+    case {Integer.parse(start_pos), Integer.parse(end_pos)} do
+      {:error, :error} ->
+        {:error, :invalid_range}
+
+      {{start_pos, _}, :error} ->
+        {:ok, {start_pos, file_size - 1}}
+
+      # See RFC7233
+      {{start_pos, _}, {end_pos, _}} when end_pos >= file_size ->
+        {:ok, {start_pos, file_size - 1}}
+
+      {{start_pos, _}, {end_pos, _}} ->
+        {:ok, {start_pos, end_pos}}
+    end
+  end
 end

lib/pinchflat_web/controllers/media_items/media_item_html.ex

@@ -2,4 +2,24 @@ defmodule PinchflatWeb.MediaItems.MediaItemHTML do
   use PinchflatWeb, :html
 
   embed_templates "media_item_html/*"
+
+  @doc """
+  Renders a media item form.
+  """
+  attr :changeset, Ecto.Changeset, required: true
+  attr :action, :string, required: true
+
+  def media_item_form(assigns)
+
+  def media_file_exists?(media_item) do
+    !!media_item.media_filepath and File.exists?(media_item.media_filepath)
+  end
+
+  def media_type(media_item) do
+    case Path.extname(media_item.media_filepath) do
+      ext when ext in [".mp4", ".webm", ".mkv"] -> :video
+      ext when ext in [".mp3", ".m4a"] -> :audio
+      _ -> :unknown
+    end
+  end
 end

lib/pinchflat_web/controllers/media_items/media_item_html/actions_dropdown.html.heex

@@ -0,0 +1,32 @@
+<.button_dropdown text="Actions" class="justify-center w-full sm:w-50">
+  <:option>
+    <.link
+      href={~p"/sources/#{@media_item.source_id}/media/#{@media_item}/force_download"}
+      method="post"
+      data-confirm="Are you sure you force a download of this media?"
+    >
+      Force Download
+    </.link>
+  </:option>
+  <:option>
+    <div class="h-px w-full bg-bodydark2"></div>
+  </:option>
+  <:option>
+    <.link
+      href={~p"/sources/#{@media_item.source_id}/media/#{@media_item}"}
+      method="delete"
+      data-confirm="Are you sure you want to delete all files for this media item? This cannot be undone."
+    >
+      Delete Files
+    </.link>
+  </:option>
+  <:option>
+    <.link
+      href={~p"/sources/#{@media_item.source_id}/media/#{@media_item}?prevent_download=true"}
+      method="delete"
+      data-confirm="Are you sure you want to delete all files for this media item and prevent it from re-downloading in the future? This cannot be undone."
+    >
+      Delete and Ignore
+    </.link>
+  </:option>
+</.button_dropdown>

lib/pinchflat_web/controllers/media_items/media_item_html/edit.html.heex

@@ -0,0 +1,13 @@
+<div class="mb-6 flex gap-3 flex-row items-center">
+  <h2 class="text-title-md2 font-bold text-black dark:text-white ml-4">
+    Editing "<%= StringUtils.truncate(@media_item.title, 35) %>"
+  </h2>
+</div>
+
+<div class="rounded-sm border border-stroke bg-white px-5 pb-2.5 pt-6 shadow-default dark:border-strokedark dark:bg-boxdark sm:px-7.5 xl:pb-1">
+  <div class="max-w-full overflow-x-auto">
+    <div class="flex flex-col gap-10">
+      <.media_item_form changeset={@changeset} action={~p"/sources/#{@media_item.source_id}/media/#{@media_item}"} />
+    </div>
+  </div>
+</div>

lib/pinchflat_web/controllers/media_items/media_item_html/media_item_form.html.heex

@@ -0,0 +1,31 @@
+<.simple_form
+  :let={f}
+  for={@changeset}
+  action={@action}
+  x-data="{ advancedMode: !!JSON.parse(localStorage.getItem('advancedMode')) }"
+  x-init="$watch('advancedMode', value => localStorage.setItem('advancedMode', JSON.stringify(value)))"
+>
+  <.error :if={@changeset.action}>
+    Oops, something went wrong! Please check the errors below.
+  </.error>
+
+  <h3 class=" text-2xl text-black dark:text-white">
+    General Options
+  </h3>
+
+  <.input
+    field={f[:prevent_download]}
+    type="toggle"
+    label="Prevent Download"
+    help="Checking excludes this media item from automatic download. Download can still be manually forced"
+  />
+
+  <.input
+    field={f[:prevent_culling]}
+    type="toggle"
+    label="Prevent Automatic Deletion"
+    help="Checking excludes media from being automatically deleted based on media retention rules"
+  />
+
+  <.button class="my-10 sm:mb-7.5 w-full sm:w-auto" rounding="rounded-lg">Save Media Item</.button>
+</.simple_form>

lib/pinchflat_web/controllers/media_items/media_item_html/media_preview.heex

@@ -0,0 +1,13 @@
+<%= if media_type(@media_item) == :video do %>
+  <video controls class="max-h-96 w-full lg:w-2/3 xl:w-1/2">
+    <source src={~p"/media/#{@media_item.uuid}/stream"} type="video/mp4" />
+    Your browser does not support the video element.
+  </video>
+<% end %>
+
+<%= if media_type(@media_item) == :audio do %>
+  <audio controls class="w-full lg:w-2/3 xl:w-1/2">
+    <source src={~p"/media/#{@media_item.uuid}/stream"} type="audio/mpeg" />
+    Your browser does not support the audio element.
+  </audio>
+<% end %>

lib/pinchflat_web/controllers/media_items/media_item_html/show.html.heex

@@ -4,37 +4,43 @@
       <.icon name="hero-arrow-left" class="w-10 h-10 hover:dark:text-white" />
     </.link>
     <h2 class="text-title-md2 font-bold text-black dark:text-white ml-4">
-      Media Item #<%= @media_item.id %>
+      <%= StringUtils.truncate(@media_item.title, 35) %>
     </h2>
   </div>
+
+  <nav>
+    <.link href={~p"/sources/#{@media_item.source_id}/media/#{@media_item}/edit"}>
+      <.button color="bg-primary" rounding="rounded-lg">
+        <.icon name="hero-pencil-square" class="mr-2" /> Edit
+      </.button>
+    </.link>
+  </nav>
 </div>
-<div class="rounded-sm border border-stroke bg-white py-5 pt-6 shadow-default dark:border-strokedark dark:bg-boxdark sm:px-7.5">
+<div class="rounded-sm border border-stroke bg-white py-5 pt-6 shadow-default dark:border-strokedark dark:bg-boxdark px-7.5">
   <div class="max-w-full overflow-x-auto">
     <.tabbed_layout>
+      <:tab_append>
+        <.actions_dropdown media_item={@media_item} />
+      </:tab_append>
+
       <:tab title="Attributes">
         <div class="flex flex-col gap-10 dark:text-white">
+          <%= if media_file_exists?(@media_item) do %>
+            <h3 class="font-bold text-xl">Preview</h3>
+            <.media_preview media_item={@media_item} />
+          <% end %>
+
+          <h2 class="font-bold text-2xl"><%= @media_item.title %></h2>
           <h3 class="font-bold text-xl">Attributes</h3>
           <section>
             <strong>Source:</strong>
-            <.inline_link href={~p"/sources/#{@media_item.source_id}"}>
+            <.subtle_link href={~p"/sources/#{@media_item.source_id}"}>
               <%= @media_item.source.custom_name %>
-            </.inline_link>
+            </.subtle_link>
           </section>
 
           <.list_items_from_map map={Map.from_struct(@media_item)} />
         </div>
-
-        <section class="flex justify-center my-10">
-          <.link
-            href={~p"/sources/#{@media_item.source_id}/media/#{@media_item}?delete_files=true"}
-            method="delete"
-            data-confirm="Are you sure you want to delete this record and all associated files on disk? This cannot be undone."
-          >
-            <.button color="bg-meta-1" rounding="rounded-lg">
-              Delete Files
-            </.button>
-          </.link>
-        </section>
       </:tab>
       <:tab title="Tasks">
         <%= if match?([_|_], @media_item.tasks) do %>

lib/pinchflat_web/controllers/media_profiles/media_profile_controller.ex

@@ -1,12 +1,19 @@
 defmodule PinchflatWeb.MediaProfiles.MediaProfileController do
   use PinchflatWeb, :controller
 
+  import Ecto.Query, warn: false
+
   alias Pinchflat.Repo
   alias Pinchflat.Profiles
+  alias Pinchflat.Sources.SourcesQuery
   alias Pinchflat.Profiles.MediaProfile
 
   def index(conn, _params) do
-    media_profiles = Profiles.list_media_profiles()
+    media_profiles =
+      MediaProfile
+      |> order_by(asc: :name)
+      |> Repo.all()
+
     render(conn, :index, media_profiles: media_profiles)
   end
 
@@ -32,12 +39,15 @@ defmodule PinchflatWeb.MediaProfiles.MediaProfileController do
   end
 
   def show(conn, %{"id" => id}) do
-    media_profile =
-      id
-      |> Profiles.get_media_profile!()
-      |> Repo.preload(:sources)
+    media_profile = Profiles.get_media_profile!(id)
 
-    render(conn, :show, media_profile: media_profile)
+    sources =
+      SourcesQuery.new()
+      |> SourcesQuery.for_media_profile(media_profile)
+      |> order_by(asc: :custom_name)
+      |> Repo.all()
+
+    render(conn, :show, media_profile: media_profile, sources: sources)
   end
 
   def edit(conn, %{"id" => id}) do

lib/pinchflat_web/controllers/media_profiles/media_profile_html.ex

@@ -32,6 +32,26 @@ defmodule PinchflatWeb.MediaProfiles.MediaProfileHTML do
     ]
   end
 
+  def friendly_sponsorblock_options do
+    [
+      {"Disabled (default)", "disabled"},
+      {"Remove Segments", "remove"}
+    ]
+  end
+
+  def frieldly_sponsorblock_categories do
+    [
+      {"Sponsor", "sponsor"},
+      {"Intro/Intermission", "intro"},
+      {"Outro/Credits", "outro"},
+      {"Self Promotion", "selfpromo"},
+      {"Preview/Recap", "preview"},
+      {"Filler Tangent", "filler"},
+      {"Interaction Reminder", "interaction"},
+      {"Non-music Section", "music_offtopic"}
+    ]
+  end
+
   def custom_output_template_options do
     %{
       upload_day: nil,

lib/pinchflat_web/controllers/media_profiles/media_profile_html/media_profile_form.html.heex

@@ -38,7 +38,7 @@
         presets: { 
           default: 'Default',
           media_center: 'TV Shows',
-          audio: 'Audio',
+          audio: 'Music',
           archiving: 'Archiving'
         }
     }">
@@ -238,6 +238,36 @@
       />
     </section>
 
+    <h3 class="mt-10 text-2xl text-black dark:text-white">
+      SponsorBlock Options
+    </h3>
+
+    <section x-data="{ sponsorblockBehaviour: null }">
+      <section x-data="{ presets: { default: 'disabled', media_center: 'disabled', audio: 'disabled', archiving: 'disabled' } }">
+        <.input
+          field={f[:sponsorblock_behaviour]}
+          options={friendly_sponsorblock_options()}
+          type="select"
+          label="SponsorBlock Behaviour"
+          help="Action to take when SponsorBlock segments are found. 'Disabled' won't take any action"
+          x-model="sponsorblockBehaviour"
+          x-init="
+            sponsorblockBehaviour = $el.value
+            $watch('selectedPreset', p => p && ($el.value = presets[p]))
+          "
+        />
+      </section>
+
+      <section x-show="sponsorblockBehaviour !== 'disabled'" x-transition>
+        <.input
+          field={f[:sponsorblock_categories]}
+          options={frieldly_sponsorblock_categories()}
+          type="checkbox_group"
+          label="SponsorBlock Categories"
+        />
+      </section>
+    </section>
+
     <.button class="my-10 sm:mb-7.5 w-full sm:w-auto" rounding="rounded-lg">Save Media profile</.button>
   </section>
 

lib/pinchflat_web/controllers/media_profiles/media_profile_html/show.html.heex

@@ -19,38 +19,42 @@
 <div class="rounded-sm border border-stroke bg-white px-5 pb-2.5 pt-6 shadow-default dark:border-strokedark dark:bg-boxdark sm:px-7.5 xl:pb-1">
   <div class="max-w-full overflow-x-auto">
     <.tabbed_layout>
-      <:tab title="Attributes">
-        <div class="flex flex-col gap-10 text-white">
-          <h3 class="font-bold text-xl">Attributes for "<%= @media_profile.name %>"</h3>
-          <.list_items_from_map map={Map.from_struct(@media_profile)} />
-        </div>
-
-        <section class="flex flex-col md:flex-row items-center md:justify-around my-10">
+      <:tab_append>
+        <.button_dropdown text="Actions" class="justify-center w-full sm:w-50">
+          <:option>
             <.link
               href={~p"/media_profiles/#{@media_profile}"}
               method="delete"
               data-confirm="Are you sure you want to delete this profile and all its sources (leaving files in place)? This cannot be undone."
             >
-            <.button color="bg-meta-1" rounding="rounded-lg">
-              Delete Profile and its Sources
-            </.button>
+              Delete Profile
             </.link>
+          </:option>
+          <:option>
             <.link
               href={~p"/media_profiles/#{@media_profile}?delete_files=true"}
               method="delete"
               data-confirm="Are you sure you want to delete this profile, all its sources, and its files on disk? This cannot be undone."
               class="mt-5 md:mt-0"
             >
-            <.button color="bg-meta-1" rounding="rounded-lg">
-              Delete Profile, Sources, and Files
-            </.button>
+              Delete Profile + Files
             </.link>
-        </section>
+          </:option>
+        </.button_dropdown>
+      </:tab_append>
+
+      <:tab title="Attributes">
+        <div class="flex flex-col gap-10 text-white">
+          <h3 class="font-bold text-xl">Attributes for "<%= @media_profile.name %>"</h3>
+          <.list_items_from_map map={Map.from_struct(@media_profile)} />
+        </div>
       </:tab>
       <:tab title="Sources">
-        <.table rows={@media_profile.sources} table_class="text-black dark:text-white">
+        <.table rows={@sources} table_class="text-black dark:text-white">
           <:col :let={source} label="Name">
+            <.subtle_link href={~p"/sources/#{source.id}"}>
               <%= source.custom_name || source.collection_name %>
+            </.subtle_link>
           </:col>
           <:col :let={source} label="Type"><%= source.collection_type %></:col>
           <:col :let={source} label="Should Download?">

lib/pinchflat_web/controllers/pages/page_controller.ex

@@ -11,7 +11,7 @@ defmodule PinchflatWeb.Pages.PageController do
     done_onboarding = params["onboarding"] == "0"
     force_onboarding = params["onboarding"] == "1"
 
-    if done_onboarding, do: Settings.set!(:onboarding, false)
+    if done_onboarding, do: Settings.set(onboarding: false)
 
     if force_onboarding || Settings.get!(:onboarding) do
       render_onboarding_page(conn)
@@ -30,7 +30,7 @@ defmodule PinchflatWeb.Pages.PageController do
   end
 
   defp render_onboarding_page(conn) do
-    Settings.set!(:onboarding, true)
+    Settings.set(onboarding: true)
 
     conn
     |> render(:onboarding_checklist,

lib/pinchflat_web/controllers/pages/page_html/home.html.heex

@@ -1,27 +1,27 @@
 <div class="grid grid-cols-1 gap-4 md:grid-cols-3">
   <div class="rounded-sm border px-7.5 py-6 shadow-default border-strokedark bg-boxdark">
-    <div class="mt-4 flex flex-col items-center justify-center">
+    <a href={~p"/media_profiles"} class="mt-4 flex flex-col items-center justify-center">
       <span class="text-md font-medium">Media Profile(s)</span>
       <h4 class="text-title-md font-bold text-white">
         <%= @media_profile_count %>
       </h4>
-    </div>
+    </a>
   </div>
   <div class="rounded-sm border px-7.5 py-6 shadow-default border-strokedark bg-boxdark">
-    <div class="mt-4 flex flex-col items-center justify-center">
+    <a href={~p"/sources"} class="mt-4 flex flex-col items-center justify-center">
       <span class="text-md font-medium">Source(s)</span>
       <h4 class="text-title-md font-bold text-white">
         <%= @source_count %>
       </h4>
-    </div>
+    </a>
   </div>
   <div class="rounded-sm border px-7.5 py-6 shadow-default border-strokedark bg-boxdark">
-    <div class="mt-4 flex flex-col items-center justify-center">
+    <a href="#" class="mt-4 flex flex-col items-center justify-center">
       <span class="text-md font-medium">Media Item(s)</span>
       <h4 class="text-title-md font-bold text-white">
         <%= @media_item_count %>
       </h4>
-    </div>
+    </a>
   </div>
   <span class="text-strokedark">I know this page isn't super useful yet, but give it time :&rpar;</span>
 </div>

lib/pinchflat_web/controllers/podcasts/podcast_controller.ex

@@ -0,0 +1,43 @@
+defmodule PinchflatWeb.Podcasts.PodcastController do
+  use PinchflatWeb, :controller
+
+  alias Pinchflat.Repo
+  alias Pinchflat.Media.MediaQuery
+  alias Pinchflat.Sources.Source
+  alias Pinchflat.Podcasts.RssFeedBuilder
+  alias Pinchflat.Podcasts.PodcastHelpers
+
+  def rss_feed(conn, %{"uuid" => uuid}) do
+    source = Repo.get_by!(Source, uuid: uuid)
+    url_base = url(conn, ~p"/")
+    xml = RssFeedBuilder.build(source, limit: 300, url_base: url_base)
+
+    conn
+    |> put_resp_content_type("application/rss+xml")
+    |> put_resp_header("content-disposition", "inline")
+    |> send_resp(200, xml)
+  end
+
+  def feed_image(conn, %{"uuid" => uuid}) do
+    source = Repo.get_by!(Source, uuid: uuid)
+
+    # This is used to fetch a fallback cover image
+    # if the source doesn't have any usable images
+    media_items =
+      MediaQuery.new()
+      |> MediaQuery.for_source(source)
+      |> MediaQuery.with_media_filepath()
+      |> Repo.maybe_limit(1)
+      |> Repo.all()
+
+    case PodcastHelpers.select_cover_image(source, media_items) do
+      {:error, _} ->
+        send_resp(conn, 404, "Image not found")
+
+      {:ok, filepath} ->
+        conn
+        |> put_resp_content_type(MIME.from_path(filepath))
+        |> send_file(200, filepath)
+    end
+  end
+end

lib/pinchflat_web/controllers/sources/source_controller.ex

@@ -1,15 +1,24 @@
 defmodule PinchflatWeb.Sources.SourceController do
   use PinchflatWeb, :controller
 
+  import Ecto.Query, warn: false
+
   alias Pinchflat.Repo
-  alias Pinchflat.Media
   alias Pinchflat.Tasks
   alias Pinchflat.Sources
-  alias Pinchflat.Profiles
+  alias Pinchflat.MediaQuery
   alias Pinchflat.Sources.Source
+  alias Pinchflat.Media.MediaQuery
+  alias Pinchflat.Profiles.MediaProfile
+  alias Pinchflat.Downloading.DownloadingHelpers
+  alias Pinchflat.SlowIndexing.SlowIndexingHelpers
 
   def index(conn, _params) do
-    sources = Repo.preload(Sources.list_sources(), :media_profile)
+    sources =
+      Source
+      |> order_by(asc: :custom_name)
+      |> Repo.all()
+      |> Repo.preload(:media_profile)
 
     render(conn, :index, sources: sources)
   end
@@ -46,15 +55,33 @@ defmodule PinchflatWeb.Sources.SourceController do
   def show(conn, %{"id" => id}) do
     source = Repo.preload(Sources.get_source!(id), :media_profile)
 
-    pending_tasks = Repo.preload(Tasks.list_pending_tasks_for(source), :job)
-    pending_media = Media.list_pending_media_items_for(source, limit: 100)
-    downloaded_media = Media.list_downloaded_media_items_for(source, limit: 100)
+    pending_tasks =
+      source
+      |> Tasks.list_tasks_for(nil, [:executing, :available, :scheduled, :retryable])
+      |> Repo.preload(:job)
+
+    pending_media =
+      MediaQuery.new()
+      |> MediaQuery.for_source(source)
+      |> MediaQuery.with_media_pending_download()
+      |> order_by(desc: :id)
+      |> limit(100)
+      |> Repo.all()
+
+    downloaded_media =
+      MediaQuery.new()
+      |> MediaQuery.for_source(source)
+      |> MediaQuery.with_media_filepath()
+      |> order_by(desc: :id)
+      |> limit(100)
+      |> Repo.all()
 
     render(conn, :show,
       source: source,
       pending_tasks: pending_tasks,
       pending_media: pending_media,
-      downloaded_media: downloaded_media
+      downloaded_media: downloaded_media,
+      total_downloaded: total_downloaded_for(source)
     )
   end
 
@@ -100,8 +127,35 @@ defmodule PinchflatWeb.Sources.SourceController do
     |> redirect(to: ~p"/sources")
   end
 
+  def force_download(conn, %{"source_id" => id}) do
+    source = Sources.get_source!(id)
+    DownloadingHelpers.enqueue_pending_download_tasks(source)
+
+    conn
+    |> put_flash(:info, "Forced download of pending media items.")
+    |> redirect(to: ~p"/sources/#{source}")
+  end
+
+  def force_index(conn, %{"source_id" => id}) do
+    source = Sources.get_source!(id)
+    SlowIndexingHelpers.kickoff_indexing_task(source, %{force: true})
+
+    conn
+    |> put_flash(:info, "Index enqueued.")
+    |> redirect(to: ~p"/sources/#{source}")
+  end
+
   defp media_profiles do
-    Profiles.list_media_profiles()
+    MediaProfile
+    |> order_by(asc: :name)
+    |> Repo.all()
+  end
+
+  defp total_downloaded_for(source) do
+    MediaQuery.new()
+    |> MediaQuery.for_source(source)
+    |> MediaQuery.with_media_filepath()
+    |> Repo.aggregate(:count, :id)
   end
 
   defp get_onboarding_layout do

lib/pinchflat_web/controllers/sources/source_html.ex

@@ -24,4 +24,8 @@ defmodule PinchflatWeb.Sources.SourceHTML do
       {"Monthly", 30 * 24 * 60}
     ]
   end
+
+  def rss_feed_url(conn, source) do
+    url(conn, ~p"/sources/#{source.uuid}/feed") <> ".xml"
+  end
 end

lib/pinchflat_web/controllers/sources/source_html/actions_dropdown.html.heex

@@ -0,0 +1,55 @@
+<.button_dropdown text="Actions" class="justify-center w-full sm:w-50">
+  <:option>
+    <span
+      x-data="{ copied: false }"
+      x-on:click={"
+                window.copyTextToClipboard('#{rss_feed_url(@conn, @source)}')
+                copied = true
+                setTimeout(() => copied = false, 4000)
+              "}
+    >
+      Copy RSS Feed
+      <span x-show="copied" x-transition.duration.150ms><.icon name="hero-check" class="ml-2 h-4 w-4" /></span>
+    </span>
+  </:option>
+  <:option :if={@source.download_media}>
+    <.link
+      href={~p"/sources/#{@source}/force_download"}
+      method="post"
+      data-confirm="Are you sure you want to force a download of all *pending* media items? This isn't normally needed."
+    >
+      Force Download
+    </.link>
+  </:option>
+  <:option>
+    <.link
+      href={~p"/sources/#{@source}/force_index"}
+      method="post"
+      data-confirm="Are you sure you want to force an index of this source? This isn't normally needed."
+    >
+      Force Index
+    </.link>
+  </:option>
+  <:option>
+    <div class="h-px w-full bg-bodydark2"></div>
+  </:option>
+  <:option>
+    <.link
+      href={~p"/sources/#{@source}"}
+      method="delete"
+      data-confirm="Are you sure you want to delete this source (leaving files in place)? This cannot be undone."
+    >
+      Delete Source
+    </.link>
+  </:option>
+  <:option>
+    <.link
+      href={~p"/sources/#{@source}?delete_files=true"}
+      method="delete"
+      data-confirm="Are you sure you want to delete this source and it's files on disk? This cannot be undone."
+      class="mt-5 md:mt-0"
+    >
+      Delete Source + Files
+    </.link>
+  </:option>
+</.button_dropdown>

lib/pinchflat_web/controllers/sources/source_html/index.html.heex

@@ -22,6 +22,13 @@
         <:col :let={source} label="Should Download?">
           <.icon name={if source.download_media, do: "hero-check", else: "hero-x-mark"} />
         </:col>
+        <:col :let={source} label="Retention">
+          <%= if source.retention_period_days && source.retention_period_days > 0 do %>
+            <%= source.retention_period_days %> day(s)
+          <% else %>
+            <span class="text-lg">∞</span>
+          <% end %>
+        </:col>
         <:col :let={source} label="Media Profile">
           <.subtle_link href={~p"/media_profiles/#{source.media_profile_id}"}>
             <%= source.media_profile.name %>

lib/pinchflat_web/controllers/sources/source_html/show.html.heex

@@ -19,50 +19,39 @@
 <div class="rounded-sm border border-stroke bg-white px-5 py-5 shadow-default dark:border-strokedark dark:bg-boxdark sm:px-7.5">
   <div class="max-w-full overflow-x-auto">
     <.tabbed_layout>
+      <:tab_append>
+        <.actions_dropdown source={@source} conn={@conn} />
+      </:tab_append>
+
       <:tab title="Attributes">
         <div class="flex flex-col gap-10 text-white">
           <h3 class="font-bold text-lg">Attributes</h3>
           <section>
             <strong>Media Profile:</strong>
-            <.inline_link href={~p"/media_profiles/#{@source.media_profile_id}"}>
+            <.subtle_link href={~p"/media_profiles/#{@source.media_profile_id}"}>
               <%= @source.media_profile.name %>
-            </.inline_link>
+            </.subtle_link>
           </section>
 
           <.list_items_from_map map={Map.from_struct(@source)} />
         </div>
-
-        <section class="flex flex-col md:flex-row items-center md:justify-around mt-10">
-          <.link
-            href={~p"/sources/#{@source}"}
-            method="delete"
-            data-confirm="Are you sure you want to delete this source (leaving files in place)? This cannot be undone."
-          >
-            <.button color="bg-meta-1" rounding="rounded-lg">
-              Delete Source
-            </.button>
-          </.link>
-          <.link
-            href={~p"/sources/#{@source}?delete_files=true"}
-            method="delete"
-            data-confirm="Are you sure you want to delete this source and it's files on disk? This cannot be undone."
-            class="mt-5 md:mt-0"
-          >
-            <.button color="bg-meta-1" rounding="rounded-lg">
-              Delete Source and Files
-            </.button>
-          </.link>
-        </section>
       </:tab>
       <:tab title="Pending Media">
         <%= if match?([_|_], @pending_media) do %>
           <h4 class="text-white text-lg mb-6">Shows a maximum of 100 media items</h4>
           <.table rows={@pending_media} table_class="text-black dark:text-white">
             <:col :let={media_item} label="Title">
+              <.subtle_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"}>
                 <%= StringUtils.truncate(media_item.title, 50) %>
+              </.subtle_link>
             </:col>
             <:col :let={media_item} label="" class="flex place-content-evenly">
-              <.icon_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"} icon="hero-eye" />
+              <.icon_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"} icon="hero-eye" class="mx-1" />
+              <.icon_link
+                href={~p"/sources/#{@source.id}/media/#{media_item.id}/edit"}
+                icon="hero-pencil-square"
+                class="mx-1"
+              />
             </:col>
           </.table>
         <% else %>
@@ -71,13 +60,20 @@
       </:tab>
       <:tab title="Downloaded Media">
         <%= if match?([_|_], @downloaded_media) do %>
-          <h4 class="text-white text-lg mb-6">Shows a maximum of 100 media items</h4>
+          <h4 class="text-white text-lg mb-6">Shows a maximum of 100 media items (<%= @total_downloaded %> total)</h4>
           <.table rows={@downloaded_media} table_class="text-black dark:text-white">
             <:col :let={media_item} label="Title">
+              <.subtle_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"}>
                 <%= StringUtils.truncate(media_item.title, 50) %>
+              </.subtle_link>
             </:col>
             <:col :let={media_item} label="" class="flex place-content-evenly">
-              <.icon_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"} icon="hero-eye" />
+              <.icon_link href={~p"/sources/#{@source.id}/media/#{media_item.id}"} icon="hero-eye" class="mx-1" />
+              <.icon_link
+                href={~p"/sources/#{@source.id}/media/#{media_item.id}/edit"}
+                icon="hero-pencil-square"
+                class="mx-1"
+              />
             </:col>
           </.table>
         <% else %>

lib/pinchflat_web/controllers/sources/source_html/source_form.html.heex

@@ -39,24 +39,32 @@
     Indexing Options
   </h3>
 
+  <section x-data="{ fastIndexingEnabled: null }">
     <.input
       field={f[:index_frequency_minutes]}
       options={friendly_index_frequencies()}
       type="select"
       label="Index Frequency"
+      x-bind:disabled="fastIndexingEnabled == true"
+      x-init="$watch('fastIndexingEnabled', v => v && ($el.value = 30 * 24 * 60))"
       help="Indexing is the process of checking for media to download. Sets the time between one index of this source finishing and the next one starting"
     />
 
-  <%!-- TODO: use Alpine to disable the index frequency when fast indexing is enabled --%>
     <div phx-click={show_modal("upgrade-modal")}>
       <.input
         field={f[:fast_index]}
         type="toggle"
         label="Use Fast Indexing"
         label_suffix="(pro)"
-      help="Experimental. Ignores 'Index Frequency'. Recommended for large channels that upload frequently. See below for more info"
+        help="Experimental. Overrides 'Index Frequency'. Recommended for large channels that upload frequently. See below for more info"
+        x-init="
+          // `enabled` is the data attribute that the toggle uses internally
+          fastIndexingEnabled = enabled
+          $watch('enabled', value => fastIndexingEnabled = !!value)
+        "
       />
     </div>
+  </section>
 
   <h3 class="mt-8 text-2xl text-black dark:text-white">
     Downloading Options
@@ -80,6 +88,14 @@
     help="Only download media uploaded after this date. Leave blank to download all media. Must be in YYYY-MM-DD format"
   />
 
+  <.input
+    field={f[:retention_period_days]}
+    type="number"
+    label="Retention Period (days)"
+    min="0"
+    help="Days between when media is *downloaded* and when it's deleted. Leave blank to keep media indefinitely"
+  />
+
   <section x-show="advancedMode">
     <h3 class="mt-8 text-2xl text-black dark:text-white">
       Advanced Options

lib/pinchflat_web/endpoint.ex

@@ -20,7 +20,7 @@ defmodule PinchflatWeb.Endpoint do
   plug Plug.Static,
     at: "/",
     from: :pinchflat,
-    gzip: false,
+    gzip: Mix.env() == :prod,
     only: PinchflatWeb.static_paths()
 
   # Code reloading can be explicitly enabled under the
@@ -47,5 +47,54 @@ defmodule PinchflatWeb.Endpoint do
   plug Plug.MethodOverride
   plug Plug.Head
   plug Plug.Session, @session_options
+
+  plug :override_base_url
+  plug :strip_trailing_extension
+
   plug PinchflatWeb.Router
+
+  # URLs need to be generated using the host of the current page being accessed
+  # for things like Podcast RSS feeds to contain links to the right location.
+  #
+  # Normally you'd set the `url` option in the Endpoint configuration, but
+  # since this is self-hosted and often accessed at multiple different URLs,
+  # that would probably be more difficult for end-users to set up than just
+  # having the application figure it out.
+  defp override_base_url(conn, _opts) do
+    new_scheme =
+      case get_req_header(conn, "x-forwarded-proto") do
+        [scheme] -> scheme
+        _ -> to_string(conn.scheme)
+      end
+
+    new_port = if conn.port in [80, 443], do: "", else: ":#{conn.port}"
+    new_base_url = "#{new_scheme}://#{conn.host}#{new_port}"
+
+    Phoenix.Controller.put_router_url(conn, new_base_url)
+  end
+
+  defp strip_trailing_extension(%{path_info: []} = conn, _opts), do: conn
+
+  defp strip_trailing_extension(conn, _opts) do
+    path =
+      conn.path_info
+      |> List.last()
+      |> String.split(".")
+      |> Enum.reverse()
+
+    case path do
+      [_] ->
+        conn
+
+      [_format | fragments] ->
+        new_path =
+          fragments
+          |> Enum.reverse()
+          |> Enum.join(".")
+
+        path_fragments = List.replace_at(conn.path_info, -1, new_path)
+
+        %{conn | path_info: path_fragments}
+    end
+  end
 end

lib/pinchflat_web/router.ex

@@ -1,6 +1,9 @@
 defmodule PinchflatWeb.Router do
   use PinchflatWeb, :router
+  import Phoenix.LiveDashboard.Router
 
+  # IMPORTANT: `strip_trailing_extension` in endpoint.ex removes
+  # the extension from the path
   pipeline :browser do
     plug :basic_auth
     plug :accepts, ["html"]
@@ -9,12 +12,17 @@ defmodule PinchflatWeb.Router do
     plug :put_root_layout, html: {PinchflatWeb.Layouts, :root}
     plug :protect_from_forgery
     plug :put_secure_browser_headers
+    plug :allow_iframe_embed
   end
 
   pipeline :api do
     plug :accepts, ["json"]
   end
 
+  pipeline :feeds do
+    plug :maybe_basic_auth
+  end
+
   scope "/", PinchflatWeb do
     pipe_through :browser
 
@@ -24,29 +32,46 @@ defmodule PinchflatWeb.Router do
     resources "/search", Searches.SearchController, only: [:show], singleton: true
 
     resources "/sources", Sources.SourceController do
-      resources "/media", MediaItems.MediaItemController, only: [:show, :delete]
+      post "/force_download", Sources.SourceController, :force_download
+      post "/force_index", Sources.SourceController, :force_index
+
+      resources "/media", MediaItems.MediaItemController, only: [:show, :edit, :update, :delete] do
+        post "/force_download", MediaItems.MediaItemController, :force_download
+      end
     end
   end
 
-  # Other scopes may use custom stacks.
-  # scope "/api", PinchflatWeb do
-  #   pipe_through :api
-  # end
+  # Routes in here _may not be_ protected by basic auth. This is necessary for
+  # media streaming to work for RSS podcast feeds.
+  scope "/", PinchflatWeb do
+    pipe_through :feeds
 
-  # Enable LiveDashboard and Swoosh mailbox preview in development
-  if Application.compile_env(:pinchflat, :dev_routes) do
-    # If you want to use the LiveDashboard in production, you should put
-    # it behind authentication and allow only admins to access it.
-    # If your application does not have an admins-only section yet,
-    # you can use Plug.BasicAuth to set up some basic authentication
-    # as long as you are also using SSL (which you should anyway).
-    import Phoenix.LiveDashboard.Router
+    get "/sources/:uuid/feed", Podcasts.PodcastController, :rss_feed
+    get "/sources/:uuid/feed_image", Podcasts.PodcastController, :feed_image
+
+    get "/media/:uuid/stream", MediaItems.MediaItemController, :stream
+  end
+
+  # No auth or CSRF protection for the health check endpoint
+  scope "/", PinchflatWeb do
+    pipe_through :api
+
+    get "/healthcheck", HealthController, :check
+  end
 
   scope "/dev" do
     pipe_through :browser
 
-      live_dashboard "/dashboard", metrics: PinchflatWeb.Telemetry
-      forward "/mailbox", Plug.Swoosh.MailboxPreview
+    live_dashboard "/dashboard",
+      metrics: PinchflatWeb.Telemetry,
+      ecto_repos: [Pinchflat.Repo]
+  end
+
+  defp maybe_basic_auth(conn, opts) do
+    if Application.get_env(:pinchflat, :expose_feed_endpoints) do
+      conn
+    else
+      basic_auth(conn, opts)
     end
   end
 
@@ -64,4 +89,8 @@ defmodule PinchflatWeb.Router do
   defp credential_set?(credential) do
     credential && credential != ""
   end
+
+  defp allow_iframe_embed(conn, _opts) do
+    delete_resp_header(conn, "x-frame-options")
+  end
 end

mix.exs

@@ -4,7 +4,7 @@ defmodule Pinchflat.MixProject do
   def project do
     [
       app: :pinchflat,
-      version: "0.1.3",
+      version: "0.1.9",
       elixir: "~> 1.16",
       elixirc_paths: elixirc_paths(Mix.env()),
       start_permanent: Mix.env() == :prod,
@@ -33,7 +33,7 @@ defmodule Pinchflat.MixProject do
   def application do
     [
       mod: {Pinchflat.Application, []},
-      extra_applications: [:logger, :runtime_tools, :inets]
+      extra_applications: [:logger, :runtime_tools, :inets, :os_mon]
     ]
   end
 
@@ -50,6 +50,7 @@ defmodule Pinchflat.MixProject do
       {:phoenix_ecto, "~> 4.4"},
       {:ecto_sql, "~> 3.10"},
       {:ecto_sqlite3, ">= 0.0.0"},
+      {:ecto_sqlite3_extras, "~> 1.2.0"},
       {:phoenix_html, "~> 3.3"},
       {:phoenix_live_reload, "~> 1.2", only: :dev},
       {:phoenix_live_view, "~> 0.20.1"},

mix.lock

@@ -13,6 +13,7 @@
   "ecto": {:hex, :ecto, "3.11.1", "4b4972b717e7ca83d30121b12998f5fcdc62ba0ed4f20fd390f16f3270d85c3e", [:mix], [{:decimal, "~> 2.0", [hex: :decimal, repo: "hexpm", optional: false]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: true]}, {:telemetry, "~> 0.4 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "ebd3d3772cd0dfcd8d772659e41ed527c28b2a8bde4b00fe03e0463da0f1983b"},
   "ecto_sql": {:hex, :ecto_sql, "3.11.1", "e9abf28ae27ef3916b43545f9578b4750956ccea444853606472089e7d169470", [:mix], [{:db_connection, "~> 2.4.1 or ~> 2.5", [hex: :db_connection, repo: "hexpm", optional: false]}, {:ecto, "~> 3.11.0", [hex: :ecto, repo: "hexpm", optional: false]}, {:myxql, "~> 0.6.0", [hex: :myxql, repo: "hexpm", optional: true]}, {:postgrex, "~> 0.16.0 or ~> 0.17.0 or ~> 1.0", [hex: :postgrex, repo: "hexpm", optional: true]}, {:tds, "~> 2.1.1 or ~> 2.2", [hex: :tds, repo: "hexpm", optional: true]}, {:telemetry, "~> 0.4.0 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "ce14063ab3514424276e7e360108ad6c2308f6d88164a076aac8a387e1fea634"},
   "ecto_sqlite3": {:hex, :ecto_sqlite3, "0.15.1", "40f2fbd9e246455f8c42e7e0a77009ef806caa1b3ce6f717b2a0a80e8432fcfd", [:mix], [{:decimal, "~> 1.6 or ~> 2.0", [hex: :decimal, repo: "hexpm", optional: false]}, {:ecto, "~> 3.11", [hex: :ecto, repo: "hexpm", optional: false]}, {:ecto_sql, "~> 3.11", [hex: :ecto_sql, repo: "hexpm", optional: false]}, {:exqlite, "~> 0.19", [hex: :exqlite, repo: "hexpm", optional: false]}], "hexpm", "28b16e177123c688948357176662bf9ff9084daddf950ef5b6baf3ee93707064"},
+  "ecto_sqlite3_extras": {:hex, :ecto_sqlite3_extras, "1.2.2", "36e60b561a11441d15f26c791817999269fb578b985162207ebb08b04ca71e40", [:mix], [{:exqlite, ">= 0.13.2", [hex: :exqlite, repo: "hexpm", optional: false]}, {:table_rex, "~> 4.0", [hex: :table_rex, repo: "hexpm", optional: false]}], "hexpm", "2b66ba7246bb4f7e39e2578acd4a0e4e4be54f60ff52d450a01be95eeb78ff1e"},
   "elixir_make": {:hex, :elixir_make, "0.7.8", "505026f266552ee5aabca0b9f9c229cbb496c689537c9f922f3eb5431157efc7", [:mix], [{:castore, "~> 0.1 or ~> 1.0", [hex: :castore, repo: "hexpm", optional: true]}, {:certifi, "~> 2.0", [hex: :certifi, repo: "hexpm", optional: true]}], "hexpm", "7a71945b913d37ea89b06966e1342c85cfe549b15e6d6d081e8081c493062c07"},
   "esbuild": {:hex, :esbuild, "0.8.1", "0cbf919f0eccb136d2eeef0df49c4acf55336de864e63594adcea3814f3edf41", [:mix], [{:castore, ">= 0.0.0", [hex: :castore, repo: "hexpm", optional: false]}, {:jason, "~> 1.4", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "25fc876a67c13cb0a776e7b5d7974851556baeda2085296c14ab48555ea7560f"},
   "ex_check": {:hex, :ex_check, "0.14.0", "d6fbe0bcc51cf38fea276f5bc2af0c9ae0a2bb059f602f8de88709421dae4f0e", [:mix], [], "hexpm", "8a602e98c66e6a4be3a639321f1f545292042f290f91fa942a285888c6868af0"},
@@ -46,6 +47,7 @@
   "ranch": {:hex, :ranch, "1.8.0", "8c7a100a139fd57f17327b6413e4167ac559fbc04ca7448e9be9057311597a1d", [:make, :rebar3], [], "hexpm", "49fbcfd3682fab1f5d109351b61257676da1a2fdbe295904176d5e521a2ddfe5"},
   "sobelow": {:hex, :sobelow, "0.13.0", "218afe9075904793f5c64b8837cc356e493d88fddde126a463839351870b8d1e", [:mix], [{:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "cd6e9026b85fc35d7529da14f95e85a078d9dd1907a9097b3ba6ac7ebbe34a0d"},
   "swoosh": {:hex, :swoosh, "1.14.4", "94e9dba91f7695a10f49b0172c4a4cb658ef24abef7e8140394521b7f3bbb2d4", [:mix], [{:cowboy, "~> 1.1 or ~> 2.4", [hex: :cowboy, repo: "hexpm", optional: true]}, {:ex_aws, "~> 2.1", [hex: :ex_aws, repo: "hexpm", optional: true]}, {:finch, "~> 0.6", [hex: :finch, repo: "hexpm", optional: true]}, {:gen_smtp, "~> 0.13 or ~> 1.0", [hex: :gen_smtp, repo: "hexpm", optional: true]}, {:hackney, "~> 1.9", [hex: :hackney, repo: "hexpm", optional: true]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: false]}, {:mail, "~> 0.2", [hex: :mail, repo: "hexpm", optional: true]}, {:mime, "~> 1.1 or ~> 2.0", [hex: :mime, repo: "hexpm", optional: false]}, {:plug, "~> 1.9", [hex: :plug, repo: "hexpm", optional: true]}, {:plug_cowboy, ">= 1.0.0", [hex: :plug_cowboy, repo: "hexpm", optional: true]}, {:req, "~> 0.4 or ~> 1.0", [hex: :req, repo: "hexpm", optional: true]}, {:telemetry, "~> 0.4.2 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "081c5a590e4ba85cc89baddf7b2beecf6c13f7f84a958f1cd969290815f0f026"},
+  "table_rex": {:hex, :table_rex, "4.0.0", "3c613a68ebdc6d4d1e731bc973c233500974ec3993c99fcdabb210407b90959b", [:mix], [], "hexpm", "c35c4d5612ca49ebb0344ea10387da4d2afe278387d4019e4d8111e815df8f55"},
   "tailwind": {:hex, :tailwind, "0.2.2", "9e27288b568ede1d88517e8c61259bc214a12d7eed271e102db4c93fcca9b2cd", [:mix], [{:castore, ">= 0.0.0", [hex: :castore, repo: "hexpm", optional: false]}], "hexpm", "ccfb5025179ea307f7f899d1bb3905cd0ac9f687ed77feebc8f67bdca78565c4"},
   "telemetry": {:hex, :telemetry, "1.2.1", "68fdfe8d8f05a8428483a97d7aab2f268aaff24b49e0f599faa091f1d4e7f61c", [:rebar3], [], "hexpm", "dad9ce9d8effc621708f99eac538ef1cbe05d6a874dd741de2e689c47feafed5"},
   "telemetry_metrics": {:hex, :telemetry_metrics, "0.6.2", "2caabe9344ec17eafe5403304771c3539f3b6e2f7fb6a6f602558c825d0d0bfb", [:mix], [{:telemetry, "~> 0.4 or ~> 1.0", [hex: :telemetry, repo: "hexpm", optional: false]}], "hexpm", "9b43db0dc33863930b9ef9d27137e78974756f5f198cae18409970ed6fa5b561"},

priv/repo/migrations/20240325222242_add_uuid_to_source_and_media.exs

@@ -0,0 +1,13 @@
+defmodule Pinchflat.Repo.Migrations.AddUuidToSourceAndMedia do
+  use Ecto.Migration
+
+  def change do
+    alter table(:sources) do
+      add :uuid, :uuid
+    end
+
+    alter table(:media_items) do
+      add :uuid, :uuid
+    end
+  end
+end

priv/repo/migrations/20240325222243_backfill_content_uuids.exs

@@ -0,0 +1,13 @@
+defmodule Pinchflat.Repo.Migrations.BackfillContentUuids do
+  use Ecto.Migration
+
+  def up do
+    execute("UPDATE sources SET uuid = gen_random_uuid() WHERE uuid IS NULL")
+    execute("UPDATE media_items SET uuid = gen_random_uuid() WHERE uuid IS NULL")
+  end
+
+  def down do
+    execute("UPDATE sources SET uuid = NULL")
+    execute("UPDATE media_items SET uuid = NULL")
+  end
+end

priv/repo/migrations/20240327165841_add_index_to_uuid_fields.exs

@@ -0,0 +1,8 @@
+defmodule Pinchflat.Repo.Migrations.AddIndexToUuidFields do
+  use Ecto.Migration
+
+  def change do
+    create unique_index(:sources, [:uuid])
+    create unique_index(:media_items, [:uuid])
+  end
+end

priv/repo/migrations/20240328160057_add_sponsorblock_to_media_profiles.exs

@@ -0,0 +1,10 @@
+defmodule Pinchflat.Repo.Migrations.AddSponsorblockToMediaProfiles do
+  use Ecto.Migration
+
+  def change do
+    alter table(:media_profiles) do
+      add :sponsorblock_behaviour, :string, default: "disabled"
+      add :sponsorblock_categories, {:array, :string}, default: []
+    end
+  end
+end

priv/repo/migrations/20240328221845_add_duration_to_media_items.exs

@@ -0,0 +1,9 @@
+defmodule Pinchflat.Repo.Migrations.AddDurationToMediaItems do
+  use Ecto.Migration
+
+  def change do
+    alter table(:media_items) do
+      add :duration_seconds, :integer
+    end
+  end
+end

priv/repo/migrations/20240401170440_re_re_add_source_uniqueness_index.exs

@@ -0,0 +1,19 @@
+defmodule Pinchflat.Repo.Migrations.ReReAddSourceUniquenessIndex do
+  use Ecto.Migration
+
+  def up do
+    execute """
+      CREATE UNIQUE INDEX sources_collection_id_media_profile_id_title_filter_regex_index ON sources (
+        collection_id,
+        media_profile_id,
+        IFNULL(title_filter_regex, '')
+      );
+    """
+  end
+
+  def down do
+    execute """
+      DROP INDEX sources_collection_id_media_profile_id_title_filter_regex_index;
+    """
+  end
+end

priv/repo/migrations/20240402192417_add_prevent_download_to_media_items.exs

@@ -0,0 +1,9 @@
+defmodule Pinchflat.Repo.Migrations.AddPreventDownloadToMediaItems do
+  use Ecto.Migration
+
+  def change do
+    alter table(:media_items) do
+      add :prevent_download, :boolean, default: false, null: false
+    end
+  end
+end

priv/repo/migrations/20240402221638_add_retention_period_to_sources.exs

@@ -0,0 +1,9 @@
+defmodule Pinchflat.Repo.Migrations.AddRetentionPeriodToSources do
+  use Ecto.Migration
+
+  def change do
+    alter table(:sources) do
+      add :retention_period_days, :integer
+    end
+  end
+end

priv/repo/migrations/20240403164943_add_culled_at_to_media_items.exs

@@ -0,0 +1,10 @@
+defmodule Pinchflat.Repo.Migrations.AddCulledAtToMediaItems do
+  use Ecto.Migration
+
+  def change do
+    alter table(:media_items) do
+      add :culled_at, :utc_datetime
+      add :prevent_culling, :boolean, default: false
+    end
+  end
+end

priv/repo/migrations/20240404172256_rename_settings_table.exs

@@ -0,0 +1,7 @@
+defmodule Pinchflat.Repo.Migrations.RenameSettingsTable do
+  use Ecto.Migration
+
+  def change do
+    rename table(:settings), to: table(:settings_backup)
+  end
+end

priv/repo/migrations/20240404174144_create_new_settings.exs

@@ -0,0 +1,29 @@
+defmodule Pinchflat.Repo.Migrations.CreateNewSettings do
+  use Ecto.Migration
+
+  def up do
+    create table(:settings) do
+      add :onboarding, :boolean, default: true, null: false
+      add :pro_enabled, :boolean, default: false, null: false
+      add :yt_dlp_version, :string
+    end
+
+    # Make an initial record because this will be the only one ever inserted
+    execute "INSERT INTO settings (onboarding, pro_enabled, yt_dlp_version) VALUES (true, false, NULL)"
+
+    # Set the value of onboarding to the previous version set in `settings_backup`
+    execute """
+      UPDATE settings
+      SET onboarding = COALESCE((SELECT value = 'true' FROM settings_backup WHERE name = 'onboarding'), true)
+    """
+
+    execute """
+      UPDATE settings
+      SET pro_enabled = COALESCE((SELECT value = 'true' FROM settings_backup WHERE name = 'pro_enabled'), false)
+    """
+  end
+
+  def down do
+    drop table(:settings)
+  end
+end

selfhosted.Dockerfile

@@ -95,11 +95,9 @@ ENV LANGUAGE en_US:en
 ENV LC_ALL en_US.UTF-8
 
 WORKDIR "/app"
-RUN chown nobody /app
 
 # Set up data volumes
 RUN mkdir /config /downloads
-RUN chown nobody /config /downloads
 
 # set runner ENV
 ENV MIX_ENV="prod"
@@ -108,7 +106,7 @@ ENV RUN_CONTEXT="selfhosted"
 EXPOSE ${PORT}
 
 # Only copy the final release from the build stage
-COPY --from=builder --chown=nobody:root /app/_build/${MIX_ENV}/rel/pinchflat ./
+COPY --from=builder /app/_build/${MIX_ENV}/rel/pinchflat ./
 
 # NEVER do this if you're running in an environment where you don't trust the user
 # (ie: most environments). This is only acceptable in a self-hosted environment.
@@ -121,12 +119,13 @@ COPY --from=builder --chown=nobody:root /app/_build/${MIX_ENV}/rel/pinchflat ./
 # root, use --user 0:0 or something.
 RUN passwd -d root
 
-USER nobody
-
 # If using an environment that doesn't automatically reap zombie processes, it is
 # advised to add an init process such as tini via `apt-get install`
 # above and adding an entrypoint. See https://github.com/krallin/tini for details
 # ENTRYPOINT ["/tini", "--"]
 
+HEALTHCHECK --interval=120s --start-period=10s \
+  CMD curl --fail http://localhost:${PORT}/healthcheck || exit 1
+
 # Start the app
 CMD ["/app/bin/docker_start"]

test/pinchflat/boot/data_backfill_worker_test.exs

@@ -1,46 +0,0 @@
-defmodule Pinchflat.Boot.DataBackfillWorkerTest do
-  use Pinchflat.DataCase
-
-  alias Pinchflat.Boot.DataBackfillWorker
-  alias Pinchflat.JobFixtures.TestJobWorker
-
-  describe "cancel_pending_backfill_jobs/0" do
-    test "cancels all pending backfill jobs" do
-      %{}
-      |> DataBackfillWorker.new()
-      |> Repo.insert_unique_job()
-
-      assert_enqueued(worker: DataBackfillWorker)
-
-      DataBackfillWorker.cancel_pending_backfill_jobs()
-
-      refute_enqueued(worker: DataBackfillWorker)
-    end
-
-    test "does not cancel jobs for other workers" do
-      %{id: 0}
-      |> TestJobWorker.new()
-      |> Repo.insert_unique_job()
-
-      assert_enqueued(worker: TestJobWorker)
-
-      DataBackfillWorker.cancel_pending_backfill_jobs()
-
-      assert_enqueued(worker: TestJobWorker)
-    end
-  end
-
-  describe "perform/1" do
-    setup do
-      DataBackfillWorker.cancel_pending_backfill_jobs()
-
-      :ok
-    end
-
-    test "reschedules itself once complete" do
-      perform_job(DataBackfillWorker, %{})
-
-      assert_enqueued(worker: DataBackfillWorker, scheduled_at: now_plus(60, :minutes))
-    end
-  end
-end

test/pinchflat/boot/pre_job_startup_tasks_test.exs

@@ -1,25 +1,47 @@
 defmodule Pinchflat.Boot.PreJobStartupTasksTest do
   use Pinchflat.DataCase
 
+  import Pinchflat.JobFixtures
+
   alias Pinchflat.Settings
-  alias Pinchflat.Settings.Setting
   alias Pinchflat.Boot.PreJobStartupTasks
 
-  describe "apply_default_settings" do
-    setup do
-      Repo.delete_all(Setting)
+  describe "reset_executing_jobs" do
+    test "resets executing jobs" do
+      job = job_fixture()
+      Repo.update_all(Oban.Job, set: [state: "executing"])
 
-      :ok
-    end
-
-    test "sets default settings" do
-      assert_raise Ecto.NoResultsError, fn -> Settings.get!(:onboarding) end
-      assert_raise Ecto.NoResultsError, fn -> Settings.get!(:pro_enabled) end
+      assert Repo.reload!(job).state == "executing"
 
       PreJobStartupTasks.start_link()
 
-      assert Settings.get!(:onboarding)
-      refute Settings.get!(:pro_enabled)
+      assert Repo.reload!(job).state == "retryable"
+    end
+  end
+
+  describe "create_blank_cookie_file" do
+    test "creates a blank cookie file" do
+      base_dir = Application.get_env(:pinchflat, :extras_directory)
+      filepath = Path.join(base_dir, "cookies.txt")
+      File.rm(filepath)
+
+      refute File.exists?(filepath)
+
+      PreJobStartupTasks.start_link()
+
+      assert File.exists?(filepath)
+    end
+  end
+
+  describe "apply_default_settings" do
+    test "sets default settings" do
+      Settings.set(yt_dlp_version: nil)
+
+      refute Settings.get!(:yt_dlp_version)
+
+      PreJobStartupTasks.start_link()
+
+      assert Settings.get!(:yt_dlp_version)
     end
   end
 end

test/pinchflat/downloading/download_option_builder_test.exs

@@ -222,6 +222,45 @@ defmodule Pinchflat.Downloading.DownloadOptionBuilderTest do
     end
   end
 
+  describe "build/1 when testing sponsorblock options" do
+    test "includes :sponsorblock_remove option when specified", %{media_item: media_item} do
+      media_item =
+        update_media_profile_attribute(media_item, %{
+          sponsorblock_behaviour: :remove,
+          sponsorblock_categories: ["sponsor", "intro"]
+        })
+
+      assert {:ok, res} = DownloadOptionBuilder.build(media_item)
+
+      assert {:sponsorblock_remove, "sponsor,intro"} in res
+    end
+
+    test "does not include :sponsorblock_remove option without categories", %{media_item: media_item} do
+      media_item =
+        update_media_profile_attribute(media_item, %{
+          sponsorblock_behaviour: :remove,
+          sponsorblock_categories: []
+        })
+
+      assert {:ok, res} = DownloadOptionBuilder.build(media_item)
+
+      refute {:sponsorblock_remove, ""} in res
+      refute {:sponsorblock_remove, []} in res
+      refute :sponsorblock_remove in res
+    end
+
+    test "does not include any sponsorblock options when disabled", %{media_item: media_item} do
+      media_item =
+        update_media_profile_attribute(media_item, %{sponsorblock_behaviour: :disabled})
+
+      assert {:ok, res} = DownloadOptionBuilder.build(media_item)
+
+      refute {:sponsorblock_remove, ""} in res
+      refute {:sponsorblock_remove, []} in res
+      refute :sponsorblock_remove in res
+    end
+  end
+
   describe "build_output_path_for/1" do
     test "builds an output path for a source", %{media_item: media_item} do
       path = DownloadOptionBuilder.build_output_path_for(media_item.source)

test/pinchflat/downloading/media_download_worker_test.exs

@@ -4,6 +4,7 @@ defmodule Pinchflat.Downloading.MediaDownloadWorkerTest do
   import Mox
   import Pinchflat.MediaFixtures
 
+  alias Pinchflat.Media
   alias Pinchflat.Sources
   alias Pinchflat.Filesystem.FilesystemHelpers
   alias Pinchflat.Downloading.MediaDownloadWorker
@@ -34,11 +35,28 @@ defmodule Pinchflat.Downloading.MediaDownloadWorkerTest do
       assert {:ok, task} = MediaDownloadWorker.kickoff_with_task(media_item)
       assert task.media_item_id == media_item.id
     end
+
+    test "can be called with additional job arguments", %{media_item: media_item} do
+      job_args = %{"force" => true}
+
+      assert {:ok, _} = MediaDownloadWorker.kickoff_with_task(media_item, job_args)
+
+      assert_enqueued(worker: MediaDownloadWorker, args: %{"id" => media_item.id, "force" => true})
+    end
+
+    test "can be called with additional job options", %{media_item: media_item} do
+      job_opts = [max_attempts: 5]
+
+      assert {:ok, _} = MediaDownloadWorker.kickoff_with_task(media_item, %{}, job_opts)
+
+      [job] = all_enqueued(worker: MediaDownloadWorker, args: %{"id" => media_item.id})
+      assert job.max_attempts == 5
+    end
   end
 
   describe "perform/1" do
     test "it saves attributes to the media_item", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         {:ok, render_metadata(:media_metadata)}
       end)
 
@@ -48,7 +66,7 @@ defmodule Pinchflat.Downloading.MediaDownloadWorkerTest do
     end
 
     test "it saves the metadata to the media_item", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         {:ok, render_metadata(:media_metadata)}
       end)
 
@@ -65,7 +83,7 @@ defmodule Pinchflat.Downloading.MediaDownloadWorkerTest do
     end
 
     test "it sets the job to retryable if the download fails", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot -> {:error, "error"} end)
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl -> {:error, "error"} end)
 
       Oban.Testing.with_testing_mode(:inline, fn ->
         {:ok, job} = Oban.insert(MediaDownloadWorker.new(%{id: media_item.id}))
@@ -74,16 +92,51 @@ defmodule Pinchflat.Downloading.MediaDownloadWorkerTest do
       end)
     end
 
+    test "sets the job to retryable if the download failed and was retried", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
+        {:error, "Unable to communicate with SponsorBlock", 1}
+      end)
+
+      Oban.Testing.with_testing_mode(:inline, fn ->
+        {:ok, job} = Oban.insert(MediaDownloadWorker.new(%{id: media_item.id}))
+
+        assert job.state == "retryable"
+      end)
+    end
+
+    test "it ensures error are returned in a 2-item tuple", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl -> {:error, "error", 1} end)
+
+      assert {:error, :download_failed} = perform_job(MediaDownloadWorker, %{id: media_item.id})
+    end
+
     test "it does not download if the source is set to not download", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, 0, fn _url, _opts, _ot -> :ok end)
+      expect(YtDlpRunnerMock, :run, 0, fn _url, _opts, _ot, _addl -> :ok end)
 
       Sources.update_source(media_item.source, %{download_media: false})
 
       perform_job(MediaDownloadWorker, %{id: media_item.id})
     end
 
+    test "does not download if the media item is set to not download", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, 0, fn _url, _opts, _ot, _addl -> :ok end)
+
+      Media.update_media_item(media_item, %{prevent_download: true})
+
+      perform_job(MediaDownloadWorker, %{id: media_item.id})
+    end
+
+    test "downloads anyway if forced", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl -> :ok end)
+
+      Sources.update_source(media_item.source, %{download_media: false})
+      Media.update_media_item(media_item, %{prevent_download: true})
+
+      perform_job(MediaDownloadWorker, %{id: media_item.id, force: true})
+    end
+
     test "it saves the file's size to the database", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         metadata = render_parsed_metadata(:media_metadata)
         FilesystemHelpers.write_p!(metadata["filepath"], "test")
 

test/pinchflat/downloading/media_downloader_test.exs

@@ -25,9 +25,11 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
 
   describe "download_for_media_item/3" do
     test "it calls the backend runner", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn url, _opts, ot ->
+      expect(YtDlpRunnerMock, :run, fn url, _opts, ot, addl ->
         assert url == media_item.original_url
         assert ot == "after_move:%()j"
+        assert [{:output_filepath, filepath}] = addl
+        assert is_binary(filepath)
 
         {:ok, render_metadata(:media_metadata)}
       end)
@@ -35,8 +37,8 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
       assert {:ok, _} = MediaDownloader.download_for_media_item(media_item)
     end
 
-    test "it saves the metadata filepatha to the database", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+    test "it saves the metadata filepath to the database", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         {:ok, render_metadata(:media_metadata)}
       end)
 
@@ -47,18 +49,56 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
       assert updated_media_item.metadata.thumbnail_filepath =~ "media_items/#{media_item.id}/maxresdefault.jpg"
     end
 
-    test "errors are passed through", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
-        {:error, :some_error}
+    test "non-recoverable errors are passed through", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
+        {:error, :some_error, 1}
       end)
 
       assert {:error, :some_error} = MediaDownloader.download_for_media_item(media_item)
     end
+
+    test "unknown errors are passed through", %{media_item: media_item} do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
+        {:error, :some_error}
+      end)
+
+      assert {:error, message} = MediaDownloader.download_for_media_item(media_item)
+      assert message == "Unknown error: {:error, :some_error}"
+    end
+  end
+
+  describe "download_for_media_item/3 when testing retries" do
+    test "returns a recovered tuple on recoverable errors", %{media_item: media_item} do
+      message = "Unable to communicate with SponsorBlock"
+
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
+        {:error, message, 1}
+      end)
+
+      assert {:recovered, ^message} = MediaDownloader.download_for_media_item(media_item)
+    end
+
+    test "attempts to update the media item on recoverable errors", %{media_item: media_item} do
+      message = "Unable to communicate with SponsorBlock"
+
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, addl ->
+        [{:output_filepath, filepath}] = addl
+        File.write(filepath, render_metadata(:media_metadata))
+
+        {:error, message, 1}
+      end)
+
+      assert {:recovered, ^message} = MediaDownloader.download_for_media_item(media_item)
+      media_item = Repo.reload(media_item)
+
+      assert DateTime.diff(DateTime.utc_now(), media_item.media_downloaded_at) < 2
+      assert String.ends_with?(media_item.media_filepath, ".mkv")
+    end
   end
 
   describe "download_for_media_item/3 when testing media_item attributes" do
     setup do
-      stub(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      stub(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         {:ok, render_metadata(:media_metadata)}
       end)
 
@@ -93,8 +133,14 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
       assert [["de", _], ["en", _] | _rest] = updated_media_item.subtitle_filepaths
     end
 
+    test "it extracts the duration_seconds", %{media_item: media_item} do
+      assert media_item.duration_seconds == nil
+      assert {:ok, updated_media_item} = MediaDownloader.download_for_media_item(media_item)
+      assert is_integer(updated_media_item.duration_seconds)
+    end
+
     test "it extracts the thumbnail_filepath", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         metadata = render_parsed_metadata(:media_metadata)
 
         thumbnail_filepath =
@@ -118,7 +164,7 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
     end
 
     test "it extracts the metadata_filepath", %{media_item: media_item} do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         metadata = render_parsed_metadata(:media_metadata)
 
         infojson_filepath = metadata["infojson_filename"]
@@ -137,7 +183,7 @@ defmodule Pinchflat.Downloading.MediaDownloaderTest do
 
   describe "download_for_media_item/3 when testing NFO generation" do
     setup do
-      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot ->
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl ->
         {:ok, render_metadata(:media_metadata)}
       end)
 

test/pinchflat/downloading/media_retention_worker_test.exs

@@ -0,0 +1,70 @@
+defmodule Pinchflat.Downloading.MediaRetentionWorkerTest do
+  use Pinchflat.DataCase
+
+  import Pinchflat.MediaFixtures
+  import Pinchflat.SourcesFixtures
+
+  alias Pinchflat.Media
+  alias Pinchflat.Downloading.MediaRetentionWorker
+
+  describe "perform/1" do
+    test "deletes media files that are past their retention date" do
+      {_source, old_media_item, new_media_item} = prepare_records()
+
+      perform_job(MediaRetentionWorker, %{})
+
+      assert File.exists?(new_media_item.media_filepath)
+      refute File.exists?(old_media_item.media_filepath)
+      assert Repo.reload!(new_media_item).media_filepath
+      refute Repo.reload!(old_media_item).media_filepath
+    end
+
+    test "sets deleted media to not re-download" do
+      {_source, old_media_item, new_media_item} = prepare_records()
+
+      perform_job(MediaRetentionWorker, %{})
+
+      refute Repo.reload!(new_media_item).prevent_download
+      assert Repo.reload!(old_media_item).prevent_download
+    end
+
+    test "sets culled_at timestamp on deleted media" do
+      {_source, old_media_item, new_media_item} = prepare_records()
+
+      perform_job(MediaRetentionWorker, %{})
+
+      refute Repo.reload!(new_media_item).culled_at
+      assert Repo.reload!(old_media_item).culled_at
+      assert DateTime.diff(now(), Repo.reload!(old_media_item).culled_at) < 1
+    end
+
+    test "doesn't cull media items that have prevent_culling set" do
+      {_source, old_media_item, _new_media_item} = prepare_records()
+
+      Media.update_media_item(old_media_item, %{prevent_culling: true})
+
+      perform_job(MediaRetentionWorker, %{})
+
+      assert File.exists?(old_media_item.media_filepath)
+      assert Repo.reload!(old_media_item).media_filepath
+    end
+  end
+
+  defp prepare_records do
+    source = source_fixture(%{retention_period_days: 2})
+
+    old_media_item =
+      media_item_with_attachments(%{
+        source_id: source.id,
+        media_downloaded_at: now_minus(3, :days)
+      })
+
+    new_media_item =
+      media_item_with_attachments(%{
+        source_id: source.id,
+        media_downloaded_at: now_minus(1, :day)
+      })
+
+    {source, old_media_item, new_media_item}
+  end
+end

test/pinchflat/media_test.exs

@@ -38,29 +38,95 @@ defmodule Pinchflat.MediaTest do
     end
   end
 
-  describe "list_media_items_for/1" do
-    test "it returns media_items for a given source" do
-      source = source_fixture()
-      media_item = media_item_fixture(%{source_id: source.id})
+  describe "list_cullable_media_items/0" do
+    test "returns media items where the source has a retention period" do
+      source_one = source_fixture(%{retention_period_days: 2})
+      source_two = source_fixture(%{retention_period_days: 0})
+      source_three = source_fixture(%{retention_period_days: nil})
 
-      assert Media.list_media_items_for(source) == [media_item]
-    end
+      _media_item =
+        media_item_fixture(%{
+          source_id: source_two.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      _media_item =
+        media_item_fixture(%{
+          source_id: source_three.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      expected_media_item =
+        media_item_fixture(%{
+          source_id: source_one.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      assert Media.list_cullable_media_items() == [expected_media_item]
     end
 
-  describe "list_media_items_by_media_id_for/2" do
-    test "returns media_items for a given source and media_ids" do
-      source = source_fixture()
-      media_item = media_item_fixture(%{source_id: source.id, media_id: "123"})
+    test "returns media_items with a media_filepath" do
+      source = source_fixture(%{retention_period_days: 2})
 
-      assert Media.list_media_items_by_media_id_for(source, ["123"]) == [media_item]
+      _media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: nil,
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      expected_media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      assert Media.list_cullable_media_items() == [expected_media_item]
     end
 
-    test "does not return matching media_ids for a different source" do
-      source = source_fixture()
-      other_source = source_fixture()
-      _media_item = media_item_fixture(%{source_id: other_source.id, media_id: "123"})
+    test "returns items that have passed their retention period" do
+      source = source_fixture(%{retention_period_days: 2})
 
-      assert Media.list_media_items_by_media_id_for(source, ["123"]) == []
+      _media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(2, :days)
+        })
+
+      expected_media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      assert Media.list_cullable_media_items() == [expected_media_item]
+    end
+
+    test "doesn't return items that are set to prevent culling" do
+      source = source_fixture(%{retention_period_days: 2})
+
+      _media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days),
+          prevent_culling: true
+        })
+
+      expected_media_item =
+        media_item_fixture(%{
+          source_id: source.id,
+          media_filepath: "/video/#{Faker.File.file_name(:video)}",
+          media_downloaded_at: now_minus(3, :days)
+        })
+
+      assert Media.list_cullable_media_items() == [expected_media_item]
     end
   end
 
@@ -83,14 +149,6 @@ defmodule Pinchflat.MediaTest do
 
       assert Media.list_pending_media_items_for(source) == []
     end
-
-    test "optionally accepts a limit" do
-      source = source_fixture()
-      media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil})
-
-      assert Media.list_pending_media_items_for(source, limit: 1) == [media_item]
-      assert Media.list_pending_media_items_for(source, limit: 0) == []
-    end
   end
 
   describe "list_pending_media_items_for/1 when testing shorts" do
@@ -259,22 +317,13 @@ defmodule Pinchflat.MediaTest do
     end
   end
 
-  describe "list_downloaded_media_items_for/1" do
-    test "returns only media items with a media_filepath" do
+  describe "list_pending_media_items_for/1 when testing download prevention" do
+    test "returns only media items that are not prevented from downloading" do
       source = source_fixture()
-      _media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil})
-      media_item = media_item_fixture(%{source_id: source.id, media_filepath: "/video/#{Faker.File.file_name(:video)}"})
+      _prevented_media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil, prevent_download: true})
+      media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil, prevent_download: false})
 
-      assert Media.list_downloaded_media_items_for(source) == [media_item]
-    end
-
-    test "optionally accepts a limit" do
-      source = source_fixture()
-      _media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil})
-      media_item = media_item_fixture(%{source_id: source.id, media_filepath: "/video/#{Faker.File.file_name(:video)}"})
-
-      assert Media.list_downloaded_media_items_for(source, limit: 1) == [media_item]
-      assert Media.list_downloaded_media_items_for(source, limit: 0) == []
+      assert Media.list_pending_media_items_for(source) == [media_item]
     end
   end
 
@@ -325,6 +374,39 @@ defmodule Pinchflat.MediaTest do
 
       assert Media.pending_download?(media_item)
     end
+
+    test "returns true if the content matches the title regex" do
+      source = source_fixture(%{title_filter_regex: "(?i)^FOO$"})
+      media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil, title: "foo"})
+
+      assert Media.pending_download?(media_item)
+    end
+
+    test "returns false if the content doesn't match the title regex" do
+      source = source_fixture(%{title_filter_regex: "(?i)^FOO$"})
+      media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil, title: "bar"})
+
+      refute Media.pending_download?(media_item)
+    end
+
+    test "return true if there is no title regex" do
+      source = source_fixture(%{title_filter_regex: nil})
+      media_item = media_item_fixture(%{source_id: source.id, media_filepath: nil, title: "foo"})
+
+      assert Media.pending_download?(media_item)
+    end
+
+    test "returns true if the media item is not prevented from downloading" do
+      media_item = media_item_fixture(%{media_filepath: nil, prevent_download: false})
+
+      assert Media.pending_download?(media_item)
+    end
+
+    test "returns false if the media item is prevented from downloading" do
+      media_item = media_item_fixture(%{media_filepath: nil, prevent_download: true})
+
+      refute Media.pending_download?(media_item)
+    end
   end
 
   describe "search/1" do
@@ -398,6 +480,37 @@ defmodule Pinchflat.MediaTest do
       assert media_item.media_filepath == valid_attrs.media_filepath
     end
 
+    test "automatically sets the UUID" do
+      valid_attrs = %{
+        media_id: Faker.String.base64(12),
+        title: Faker.Commerce.product_name(),
+        media_filepath: "/video/#{Faker.File.file_name(:video)}",
+        source_id: source_fixture().id,
+        original_url: "https://www.youtube.com/channel/#{Faker.String.base64(12)}",
+        upload_date: Date.utc_today()
+      }
+
+      assert {:ok, %MediaItem{} = media_item} = Media.create_media_item(valid_attrs)
+
+      assert String.length(media_item.uuid) == 36
+    end
+
+    test "UUID is not writable by the user" do
+      valid_attrs = %{
+        media_id: Faker.String.base64(12),
+        title: Faker.Commerce.product_name(),
+        media_filepath: "/video/#{Faker.File.file_name(:video)}",
+        source_id: source_fixture().id,
+        original_url: "https://www.youtube.com/channel/#{Faker.String.base64(12)}",
+        upload_date: Date.utc_today(),
+        uuid: "some-uuid"
+      }
+
+      assert {:ok, %MediaItem{} = media_item} = Media.create_media_item(valid_attrs)
+
+      assert String.length(media_item.uuid) == 36
+    end
+
     test "creating with invalid data returns error changeset" do
       assert {:error, %Ecto.Changeset{}} = Media.create_media_item(@invalid_attrs)
     end
@@ -561,6 +674,63 @@ defmodule Pinchflat.MediaTest do
     end
   end
 
+  describe "delete_media_files/2" do
+    test "does not delete the media_item" do
+      media_item = media_item_fixture()
+
+      assert {:ok, %MediaItem{}} = Media.delete_media_files(media_item)
+      assert Repo.reload!(media_item)
+    end
+
+    test "deletes attached tasks" do
+      media_item = media_item_fixture()
+      task = task_fixture(%{media_item_id: media_item.id})
+
+      assert {:ok, %MediaItem{}} = Media.delete_media_files(media_item)
+      assert_raise Ecto.NoResultsError, fn -> Repo.reload!(task) end
+    end
+
+    test "deletes the media_item's files" do
+      media_item = media_item_with_attachments()
+
+      assert File.exists?(media_item.media_filepath)
+      assert {:ok, _} = Media.delete_media_files(media_item)
+      refute File.exists?(media_item.media_filepath)
+    end
+
+    test "does not delete the media item's metadata files" do
+      stub(HTTPClientMock, :get, fn _url, _headers, _opts -> {:ok, ""} end)
+      media_item = Repo.preload(media_item_with_attachments(), :metadata)
+
+      update_attrs = %{
+        metadata: %{
+          metadata_filepath: MetadataFileHelpers.compress_and_store_metadata_for(media_item, %{}),
+          thumbnail_filepath:
+            MetadataFileHelpers.download_and_store_thumbnail_for(media_item, %{
+              "thumbnail" => "https://example.com/thumbnail.jpg"
+            })
+        }
+      }
+
+      {:ok, updated_media_item} = Media.update_media_item(media_item, update_attrs)
+      metadata = Repo.preload(updated_media_item, :metadata).metadata
+
+      assert {:ok, _} = Media.delete_media_files(updated_media_item)
+      assert Repo.reload(metadata)
+      assert File.exists?(updated_media_item.metadata.metadata_filepath)
+
+      # cleanup
+      Media.delete_media_item(updated_media_item, delete_files: true)
+    end
+
+    test "can take additional attributes update media item" do
+      media_item = media_item_with_attachments()
+
+      assert {:ok, updated_media_item} = Media.delete_media_files(media_item, %{prevent_download: true})
+      assert updated_media_item.prevent_download
+    end
+  end
+
   describe "change_media_item/1" do
     test "change_media_item/1 returns a media_item changeset" do
       media_item = media_item_fixture()

test/pinchflat/metadata/metadata_parser_test.exs

@@ -45,6 +45,12 @@ defmodule Pinchflat.Metadata.MetadataParserTest do
 
       assert result.livestream == metadata["was_live"]
     end
+
+    test "it extracts the duration in seconds", %{metadata: metadata} do
+      result = Parser.parse_for_media_item(metadata)
+
+      assert result.duration_seconds == round(metadata["duration"])
+    end
   end
 
   describe "parse_for_media_item/1 when testing subtitle metadata" do

test/pinchflat/metadata/nfo_builder_test.exs

@@ -30,6 +30,21 @@ defmodule Pinchflat.Metadata.NfoBuilderTest do
       assert String.contains?(nfo, ~S(<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>))
       assert String.contains?(nfo, "<title>#{metadata["title"]}</title>")
     end
+
+    test "escapes invalid characters", %{filepath: filepath} do
+      metadata = %{
+        "title" => "hello' & <world>",
+        "uploader" => "uploader",
+        "id" => "id",
+        "description" => "description",
+        "upload_date" => "20210101"
+      }
+
+      result = NfoBuilder.build_and_store_for_media_item(filepath, metadata)
+      nfo = File.read!(result)
+
+      assert String.contains?(nfo, "hello&#39; &amp; &lt;world&gt;")
+    end
   end
 
   describe "build_and_store_for_source/2" do
@@ -46,5 +61,18 @@ defmodule Pinchflat.Metadata.NfoBuilderTest do
       assert String.contains?(nfo, ~S(<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>))
       assert String.contains?(nfo, "<title>#{metadata["title"]}</title>")
     end
+
+    test "escapes invalid characters", %{filepath: filepath} do
+      metadata = %{
+        "title" => "hello' & <world>",
+        "description" => "description",
+        "id" => "id"
+      }
+
+      result = NfoBuilder.build_and_store_for_source(filepath, metadata)
+      nfo = File.read!(result)
+
+      assert String.contains?(nfo, "hello&#39; &amp; &lt;world&gt;")
+    end
   end
 end

test/pinchflat/metadata/source_metadata_storage_worker_test.exs

@@ -57,7 +57,7 @@ defmodule Pinchflat.Metadata.SourceMetadataStorageWorkerTest do
         _url, _opts, ot when ot == @metadata_ot -> {:ok, render_metadata(:channel_source_metadata)}
       end)
 
-      source = source_fixture()
+      source = source_fixture(%{description: nil})
 
       refute source.description
       perform_job(SourceMetadataStorageWorker, %{id: source.id})

test/pinchflat/podcasts/podcast_helpers_test.exs

@@ -0,0 +1,64 @@
+defmodule Pinchflat.Podcasts.PodcastHelpersTest do
+  use Pinchflat.DataCase
+
+  import Pinchflat.SourcesFixtures
+  import Pinchflat.MediaFixtures
+
+  alias Pinchflat.Podcasts.PodcastHelpers
+
+  describe "persisted_media_items_for/2" do
+    test "returns media items with files that exist on-disk" do
+      source = source_fixture()
+      good_media = media_item_with_attachments(%{source_id: source.id})
+      _bad_media = media_item_fixture(%{source_id: source.id, media_filepath: "/tmp/existing_file.mp3"})
+
+      assert [persisted_media] = PodcastHelpers.persisted_media_items_for(source)
+      assert persisted_media.id == good_media.id
+    end
+
+    test "lets you specify a limit" do
+      source = source_fixture()
+      _good_media = media_item_with_attachments(%{source_id: source.id})
+
+      assert [] = PodcastHelpers.persisted_media_items_for(source, limit: 0)
+    end
+  end
+
+  describe "select_cover_image/2" do
+    test "returns a source's poster, if present" do
+      source = source_with_metadata_attachments()
+
+      {:ok, res} = PodcastHelpers.select_cover_image(source, [])
+
+      assert res == source.metadata.poster_filepath
+    end
+
+    test "falls back to a source's fanart, if present" do
+      source = source_with_metadata_attachments()
+
+      File.rm(source.metadata.poster_filepath)
+
+      {:ok, res} = PodcastHelpers.select_cover_image(source, [])
+
+      assert res == source.metadata.fanart_filepath
+    end
+
+    test "falls back to a media item's thumbnail, if present" do
+      source = source_with_metadata_attachments()
+      media_item = media_item_with_metadata_attachments(%{source_id: source.id})
+
+      File.rm(source.metadata.poster_filepath)
+      File.rm(source.metadata.fanart_filepath)
+
+      {:ok, res} = PodcastHelpers.select_cover_image(source, [media_item])
+
+      assert res == media_item.metadata.thumbnail_filepath
+    end
+
+    test "returns error if no artwork can be found" do
+      source = source_fixture()
+
+      assert PodcastHelpers.select_cover_image(source, []) == {:error, :no_suitable_image}
+    end
+  end
+end

test/pinchflat/podcasts/rss_feed_builder_test.exs

@@ -0,0 +1,145 @@
+defmodule Pinchflat.Podcasts.RssFeedBuilderTest do
+  use Pinchflat.DataCase
+
+  import Pinchflat.MediaFixtures
+  import Pinchflat.SourcesFixtures
+
+  alias Pinchflat.Podcasts.RssFeedBuilder
+
+  @datetime_format "%a, %d %b %Y %H:%M:%S %z"
+
+  setup do
+    source = source_fixture()
+
+    {:ok, source: source}
+  end
+
+  describe "build/2" do
+    test "returns an XML document", %{source: source} do
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(res, ~s(<?xml version="1.0" encoding="UTF-8"?>))
+    end
+
+    test "escapes illegal characters" do
+      source = source_fixture(%{custom_name: "A & B"})
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(res, ~s(<title>A &amp; B</title>))
+    end
+
+    test "can optionally apply a limit to media items", %{source: source} do
+      good_media = media_item_with_attachments(%{source_id: source.id})
+
+      res = RssFeedBuilder.build(source, limit: 0)
+
+      refute String.contains?(res, ~s(<title>#{good_media.title}</title>))
+    end
+
+    test "can optionally specify a URL base", %{source: source} do
+      res = RssFeedBuilder.build(source, url_base: "http://example.com")
+
+      assert String.contains?(res, ~s(http://example.com/sources/#{source.uuid}/feed.xml))
+    end
+  end
+
+  describe "build/2 when testing source XML" do
+    test "returns XML for static source attributes", %{source: source} do
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(res, ~s(<title>#{source.custom_name}</title>))
+      assert String.contains?(res, ~s(<link>#{source.original_url}</link>))
+      assert String.contains?(res, ~s(<description>#{source.description}</description>))
+      assert String.contains?(res, ~s(<itunes:author>#{source.custom_name}</itunes:author>))
+      assert String.contains?(res, ~s(<itunes:subtitle>#{source.custom_name}</itunes:subtitle>))
+      assert String.contains?(res, ~s(<description>#{source.description}</description>))
+      assert String.contains?(res, ~s(<podcast:guid>#{source.uuid}</podcast:guid>))
+    end
+
+    test "returns the lastBuildDate and pubDate based off the source's timestamps", %{source: source} do
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(res, ~s(<lastBuildDate>#{format_date(source.updated_at)}</lastBuildDate>))
+      assert String.contains?(res, ~s(<pubDate>#{format_date(source.inserted_at)}</pubDate>))
+    end
+
+    test "returns a self-link", %{source: source} do
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(
+               res,
+               ~s(<atom:link href="http://localhost:8945/sources/#{source.uuid}/feed.xml" rel="self" type="application/rss+xml" />)
+             )
+    end
+
+    test "returns a link to the feed image" do
+      source = source_with_metadata_attachments()
+
+      res = RssFeedBuilder.build(source)
+      [_before, image_block, _after] = String.split(res, ~r(</?image>))
+
+      assert String.contains?(image_block, ~s(<url>http://localhost:8945/sources/#{source.uuid}/feed_image.jpg</url>))
+      assert String.contains?(image_block, ~s(<title>#{source.custom_name}</title>))
+      assert String.contains?(image_block, ~s(<link>#{source.original_url}</link>))
+
+      assert String.contains?(
+               res,
+               ~s(<itunes:image href="http://localhost:8945/sources/#{source.uuid}/feed_image.jpg"></itunes:image>)
+             )
+    end
+  end
+
+  describe "build/2 when testing media XML" do
+    test "only includes media persisted to disk", %{source: source} do
+      good_media = media_item_with_attachments(%{source_id: source.id})
+      bad_media = media_item_fixture(%{source_id: source.id, media_filepath: "/tmp/existing_file.mp3"})
+      pending_media = media_item_fixture(%{source_id: source.id, media_filepath: nil})
+
+      res = RssFeedBuilder.build(source)
+
+      assert String.contains?(res, ~s(<title>#{good_media.title}</title>))
+      refute String.contains?(res, ~s(<title>#{bad_media.title}</title>))
+      refute String.contains?(res, ~s(<title>#{pending_media.title}</title>))
+    end
+
+    test "returns XML for static media attributes", %{source: source} do
+      media_item = media_item_with_attachments(%{source_id: source.id})
+
+      res = RssFeedBuilder.build(source)
+      [_before, item_xml, _after] = String.split(res, ~r(</?item>))
+
+      assert String.contains?(item_xml, ~s(<guid isPermaLink="false">#{media_item.uuid}</guid>))
+      assert String.contains?(item_xml, ~s(<title>#{media_item.title}</title>))
+      assert String.contains?(item_xml, ~s(<link>#{media_item.original_url}</link>))
+      assert String.contains?(item_xml, ~s(<description>#{media_item.description}</description>))
+      assert String.contains?(item_xml, ~s(<itunes:author>#{source.custom_name}</itunes:author>))
+      assert String.contains?(item_xml, ~s(<itunes:subtitle>#{media_item.title}</itunes:subtitle>))
+      assert String.contains?(item_xml, ~s(<itunes:summary><![CDATA[#{media_item.description}]]></itunes:summary>))
+    end
+
+    test "returns pubDate based off the media's upload_date", %{source: source} do
+      media_item_with_attachments(%{source_id: source.id, upload_date: ~D[2020-01-01]})
+
+      res = RssFeedBuilder.build(source)
+      [_before, item_xml, _after] = String.split(res, ~r(</?item>))
+
+      assert String.contains?(item_xml, ~s(<pubDate>Wed, 01 Jan 2020 00:00:00 +0000</pubDate>))
+    end
+
+    test "returns an enclosure tag with the media's stream URL", %{source: source} do
+      media_item = media_item_with_attachments(%{source_id: source.id, media_size_bytes: 1234})
+
+      res = RssFeedBuilder.build(source)
+      [_before, item_xml, _after] = String.split(res, ~r(</?item>))
+
+      assert String.contains?(item_xml, ~s(<enclosure))
+      assert String.contains?(item_xml, ~s(url="http://localhost:8945/media/#{media_item.uuid}/stream.mp4"))
+      assert String.contains?(item_xml, ~s(length="1234"))
+      assert String.contains?(item_xml, ~s(type="video/mp4"))
+    end
+  end
+
+  defp format_date(date) do
+    Calendar.strftime(date, @datetime_format)
+  end
+end

test/pinchflat/settings_test.exs

@@ -9,100 +9,55 @@ defmodule Pinchflat.SettingsTest do
   # are always created on app boot (including in the test env),
   # so we can't treat these like a clean slate.
 
-  describe "list_settings/0" do
-    test "returns all settings" do
-      Settings.set!("foo", "bar")
-      results = Settings.list_settings()
+  setup do
+    # Ensure we have a clean slate
+    Settings.set(onboarding: false)
+    Settings.set(pro_enabled: false)
+    Settings.set(yt_dlp_version: nil)
 
-      assert Enum.all?(results, fn setting -> match?(%Setting{}, setting) end)
+    :ok
+  end
+
+  describe "record/0" do
+    test "returns the only setting" do
+      assert %Setting{} = Settings.record()
     end
   end
 
-  describe "set/2" do
-    test "creates a new setting if one does not exist" do
-      original = Repo.aggregate(Setting, :count, :id)
-      Settings.set!("foo", "bar")
-      assert Repo.aggregate(Setting, :count, :id) == original + 1
+  describe "set/1" do
+    test "updates the setting" do
+      assert {:ok, true} = Settings.set(onboarding: true)
+      assert {:ok, true} = Settings.get(:onboarding)
     end
 
-    test "updates an existing setting if one exists" do
-      Settings.set!("foo", "bar")
-      original = Repo.aggregate(Setting, :count, :id)
-      Settings.set!("foo", "baz")
-      assert Repo.aggregate(Setting, :count, :id) == original
-      assert Settings.get!("foo") == "baz"
+    test "returns an error if the setting key doesn't exist" do
+      assert {:error, :invalid_key} = Settings.set(foo: "bar")
     end
 
-    test "returns the parsed value" do
-      assert Settings.set!("foo", true) == true
-      assert Settings.set!("foo", false) == false
-      assert Settings.set!("foo", 123) == 123
-      assert Settings.set!("foo", 12.34) == 12.34
-      assert Settings.set!("foo", "bar") == "bar"
-    end
-
-    test "allows for atom keys" do
-      assert Settings.set!(:foo, "bar") == "bar"
-    end
-
-    test "blows up when an unsupported datatype is used" do
-      assert_raise FunctionClauseError, fn ->
-        Settings.set!("foo", nil)
-      end
-    end
-  end
-
-  describe "set/3" do
-    test "allows manual specification of datatype" do
-      assert Settings.set!("foo", "true", :boolean) == true
-      assert Settings.set!("foo", "false", :boolean) == false
-      assert Settings.set!("foo", "123", :integer) == 123
-      assert Settings.set!("foo", "12.34", :float) == 12.34
+    test "returns an error if the setting value is invalid" do
+      assert {:error, %Ecto.Changeset{}} = Settings.set(onboarding: "bar")
     end
   end
 
   describe "get/1" do
-    test "returns the value of the setting" do
-      Settings.set!("str", "bar")
-      Settings.set!("bool", true)
-      Settings.set!("int", 123)
-      Settings.set!("float", 12.34)
-
-      assert Settings.get!("str") == "bar"
-      assert Settings.get!("bool") == true
-      assert Settings.get!("int") == 123
-      assert Settings.get!("float") == 12.34
+    test "returns the setting value" do
+      assert {:ok, false} = Settings.get(:onboarding)
     end
 
-    test "allows for atom keys" do
-      Settings.set!("str", "bar")
-      assert Settings.get!(:str) == "bar"
-    end
-
-    test "blows up when the setting does not exist" do
-      assert_raise Ecto.NoResultsError, fn ->
-        Settings.get!("foo")
-      end
+    test "returns an error if the setting key doesn't exist" do
+      assert {:error, :invalid_key} = Settings.get(:foo)
     end
   end
 
-  describe "fetch/2" do
-    test "creates a setting if one doesn't exist" do
-      original = Repo.aggregate(Setting, :count, :id)
-      assert Settings.fetch!("foo", "bar") == "bar"
-      assert Repo.aggregate(Setting, :count, :id) == original + 1
+  describe "get!/1" do
+    test "returns the setting value" do
+      assert Settings.get!(:onboarding) == false
     end
 
-    test "returns an existing setting if one does exist" do
-      Settings.set!("foo", "bar")
-
-      assert Settings.fetch!("foo", "baz") == "bar"
+    test "raises an error if the setting key doesn't exist" do
+      assert_raise RuntimeError, "Setting `foo` not found", fn ->
+        Settings.get!(:foo)
       end
     end
-
-  describe "fetch/3" do
-    test "allows manual specification of datatype" do
-      assert Settings.fetch!("foo", "true", :boolean) == true
-    end
   end
 end

test/pinchflat/slow_indexing/media_collection_indexing_worker_test.exs

@@ -14,6 +14,42 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorkerTest do
 
   setup :verify_on_exit!
 
+  describe "kickoff_with_task/3" do
+    setup do
+      source = source_fixture(index_frequency_minutes: 10)
+
+      {:ok, %{source: source}}
+    end
+
+    test "starts the worker", %{source: source} do
+      assert [] = all_enqueued(worker: MediaCollectionIndexingWorker)
+      assert {:ok, _} = MediaCollectionIndexingWorker.kickoff_with_task(source)
+      assert [_] = all_enqueued(worker: MediaCollectionIndexingWorker)
+    end
+
+    test "attaches a task", %{source: source} do
+      assert {:ok, task} = MediaCollectionIndexingWorker.kickoff_with_task(source)
+      assert task.source_id == source.id
+    end
+
+    test "can be called with additional job arguments", %{source: source} do
+      job_args = %{"force" => true}
+
+      assert {:ok, _} = MediaCollectionIndexingWorker.kickoff_with_task(source, job_args)
+
+      assert_enqueued(worker: MediaCollectionIndexingWorker, args: %{"id" => source.id, "force" => true})
+    end
+
+    test "can be called with additional job options", %{source: source} do
+      job_opts = [max_attempts: 5]
+
+      assert {:ok, _} = MediaCollectionIndexingWorker.kickoff_with_task(source, %{}, job_opts)
+
+      [job] = all_enqueued(worker: MediaCollectionIndexingWorker, args: %{"id" => source.id})
+      assert job.max_attempts == 5
+    end
+  end
+
   describe "perform/1" do
     test "it indexes the source if it should be indexed" do
       expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl_opts -> {:ok, ""} end)
@@ -31,6 +67,14 @@ defmodule Pinchflat.SlowIndexing.MediaCollectionIndexingWorkerTest do
       perform_job(MediaCollectionIndexingWorker, %{id: source.id})
     end
 
+    test "it indexes the source no matter what if the 'force' arg is passed" do
+      expect(YtDlpRunnerMock, :run, fn _url, _opts, _ot, _addl_opts -> {:ok, ""} end)
+
+      source = source_fixture(index_frequency_minutes: 0, last_indexed_at: DateTime.utc_now())
+
+      perform_job(MediaCollectionIndexingWorker, %{id: source.id, force: true})
+    end
+
     test "it does not do any indexing if the source has been indexed and shouldn't be rescheduled" do
       expect(YtDlpRunnerMock, :run, 0, fn _url, _opts, _ot, _addl_opts -> {:ok, ""} end)
 

test/pinchflat/slow_indexing/slow_indexing_helpers_test.exs

@@ -18,7 +18,7 @@ defmodule Pinchflat.SlowIndexing.SlowIndexingHelpersTest do
 
   setup :verify_on_exit!
 
-  describe "kickoff_indexing_task/1" do
+  describe "kickoff_indexing_task/3" do
     test "it schedules a job" do
       source = source_fixture(index_frequency_minutes: 1)
 
@@ -64,6 +64,25 @@ defmodule Pinchflat.SlowIndexing.SlowIndexingHelpersTest do
 
       assert_raise Ecto.NoResultsError, fn -> Repo.reload!(task) end
     end
+
+    test "can be called with additional job arguments" do
+      source = source_fixture(index_frequency_minutes: 1)
+      job_args = %{"force" => true}
+
+      assert {:ok, _} = SlowIndexingHelpers.kickoff_indexing_task(source, job_args)
+
+      assert_enqueued(worker: MediaCollectionIndexingWorker, args: %{"id" => source.id, "force" => true})
+    end
+
+    test "can be called with additional job options" do
+      source = source_fixture(index_frequency_minutes: 1)
+      job_opts = [max_attempts: 5]
+
+      assert {:ok, _} = SlowIndexingHelpers.kickoff_indexing_task(source, %{}, job_opts)
+
+      [job] = all_enqueued(worker: MediaCollectionIndexingWorker, args: %{"id" => source.id})
+      assert job.max_attempts == 5
+    end
   end
 
   describe "index_and_enqueue_download_for_media_items/1" do

test/pinchflat/sources_test.exs

@@ -59,6 +59,31 @@ defmodule Pinchflat.SourcesTest do
   end
 
   describe "create_source/2" do
+    test "automatically sets the UUID" do
+      expect(YtDlpRunnerMock, :run, &channel_mock/3)
+
+      valid_attrs = %{
+        media_profile_id: media_profile_fixture().id,
+        original_url: "https://www.youtube.com/channel/abc123"
+      }
+
+      assert {:ok, %Source{} = source} = Sources.create_source(valid_attrs)
+      assert String.length(source.uuid) == 36
+    end
+
+    test "UUID is not writable by the user" do
+      expect(YtDlpRunnerMock, :run, &channel_mock/3)
+
+      valid_attrs = %{
+        media_profile_id: media_profile_fixture().id,
+        original_url: "https://www.youtube.com/channel/abc123",
+        uuid: "some_uuid"
+      }
+
+      assert {:ok, %Source{} = source} = Sources.create_source(valid_attrs)
+      assert String.length(source.uuid) == 36
+    end
+
     test "creates a source and adds name + ID from runner response for channels" do
       expect(YtDlpRunnerMock, :run, &channel_mock/3)
 
@@ -112,6 +137,75 @@ defmodule Pinchflat.SourcesTest do
       assert source.custom_name == "some channel name"
     end
 
+    test "creation enforces uniqueness of collection_id scoped to the media_profile and title regex" do
+      expect(YtDlpRunnerMock, :run, 2, fn _url, _opts, _ot ->
+        {:ok,
+         Phoenix.json_library().encode!(%{
+           channel: "some channel name",
+           channel_id: "some_channel_id_12345678",
+           playlist_id: "some_channel_id_12345678",
+           playlist_title: "some channel name - videos"
+         })}
+      end)
+
+      valid_once_attrs = %{
+        media_profile_id: media_profile_fixture().id,
+        original_url: "https://www.youtube.com/channel/abc123",
+        title_filter_regex: nil
+      }
+
+      assert {:ok, %Source{}} = Sources.create_source(valid_once_attrs)
+      assert {:error, %Ecto.Changeset{}} = Sources.create_source(valid_once_attrs)
+    end
+
+    test "creation lets you duplicate collection_ids and profiles as long as the regex is different" do
+      expect(YtDlpRunnerMock, :run, 2, fn _url, _opts, _ot ->
+        {:ok,
+         Phoenix.json_library().encode!(%{
+           channel: "some channel name",
+           channel_id: "some_channel_id_12345678",
+           playlist_id: "some_channel_id_12345678",
+           playlist_title: "some channel name - videos"
+         })}
+      end)
+
+      valid_attrs = %{
+        media_profile_id: media_profile_fixture().id,
+        name: "some name",
+        original_url: "https://www.youtube.com/channel/abc123"
+      }
+
+      source_1_attrs = Map.merge(valid_attrs, %{title_filter_regex: "foo"})
+      source_2_attrs = Map.merge(valid_attrs, %{title_filter_regex: "bar"})
+
+      assert {:ok, %Source{}} = Sources.create_source(source_1_attrs)
+      assert {:ok, %Source{}} = Sources.create_source(source_2_attrs)
+    end
+
+    test "creation lets you duplicate collection_ids as long as the media profile is different" do
+      expect(YtDlpRunnerMock, :run, 2, fn _url, _opts, _ot ->
+        {:ok,
+         Phoenix.json_library().encode!(%{
+           channel: "some channel name",
+           channel_id: "some_channel_id_12345678",
+           playlist_id: "some_channel_id_12345678",
+           playlist_title: "some channel name - videos"
+         })}
+      end)
+
+      valid_attrs = %{
+        name: "some name",
+        original_url: "https://www.youtube.com/channel/abc123",
+        title_filter_regex: "TEST"
+      }
+
+      source_1_attrs = Map.merge(valid_attrs, %{media_profile_id: media_profile_fixture().id})
+      source_2_attrs = Map.merge(valid_attrs, %{media_profile_id: media_profile_fixture().id})
+
+      assert {:ok, %Source{}} = Sources.create_source(source_1_attrs)
+      assert {:ok, %Source{}} = Sources.create_source(source_2_attrs)
+    end
+
     test "collection_type is inferred from source details" do
       expect(YtDlpRunnerMock, :run, &channel_mock/3)
       expect(YtDlpRunnerMock, :run, &playlist_mock/3)

test/pinchflat/utils/datetime_utils_test.exs

@@ -0,0 +1,14 @@
+defmodule Pinchflat.Utils.DatetimeUtilsTest do
+  use ExUnit.Case, async: true
+
+  alias Pinchflat.Utils.DatetimeUtils
+
+  describe "date_to_datetime/1" do
+    test "converts a Date to a DateTime" do
+      date = ~D[2022-01-01]
+      datetime = DatetimeUtils.date_to_datetime(date)
+
+      assert datetime == ~U[2022-01-01 00:00:00Z]
+    end
+  end
+end

test/pinchflat/utils/xml_utils_test.exs

@@ -0,0 +1,16 @@
+defmodule Pinchflat.Utils.XmlUtilsTest do
+  use ExUnit.Case, async: true
+
+  alias Pinchflat.Utils.XmlUtils
+
+  describe "safe/1" do
+    test "escapes invalid characters" do
+      assert XmlUtils.safe("hello' & <world>") == "hello&#39; &amp; &lt;world&gt;"
+    end
+
+    test "converts input to string" do
+      assert XmlUtils.safe(42) == "42"
+      assert XmlUtils.safe(nil) == ""
+    end
+  end
+end