[Enhancement] Use cookies on a per-source basis (#354)

* Added 'use_cookies' column to source

* Added cookies to the source form

* Updated every command to optionally include cookies

* LOTS of tests

* Made YT cookies off by default

lib/pinchflat/downloading/media_downloader.ex

@@ -106,8 +106,9 @@ defmodule Pinchflat.Downloading.MediaDownloader do
 
   defp download_with_options(url, item_with_preloads, output_filepath, override_opts) do
     {:ok, options} = DownloadOptionBuilder.build(item_with_preloads, override_opts)
+    runner_opts = [output_filepath: output_filepath, use_cookies: item_with_preloads.source.use_cookies]
 
-    YtDlpMedia.download(url, options, output_filepath: output_filepath)
+    YtDlpMedia.download(url, options, runner_opts)
   end
 
   defp recoverable_errors do

lib/pinchflat/fast_indexing/fast_indexing_helpers.ex

@@ -68,7 +68,7 @@ defmodule Pinchflat.FastIndexing.FastIndexingHelpers do
   defp create_media_item_from_media_id(source, media_id) do
     url = "https://www.youtube.com/watch?v=#{media_id}"
 
-    case YtDlpMedia.get_media_attributes(url) do
+    case YtDlpMedia.get_media_attributes(url, use_cookies: source.use_cookies) do
       {:ok, media_attrs} ->
         Media.create_media_item_from_backend_attrs(source, media_attrs)
 

lib/pinchflat/metadata/metadata_file_helpers.ex

@@ -62,11 +62,13 @@ defmodule Pinchflat.Metadata.MetadataFileHelpers do
 
   Returns binary() | nil
   """
-  def download_and_store_thumbnail_for(database_record) do
-    yt_dlp_filepath = generate_filepath_for(database_record, "thumbnail.%(ext)s")
-    real_filepath = generate_filepath_for(database_record, "thumbnail.jpg")
+  def download_and_store_thumbnail_for(media_item_with_preloads) do
+    yt_dlp_filepath = generate_filepath_for(media_item_with_preloads, "thumbnail.%(ext)s")
+    real_filepath = generate_filepath_for(media_item_with_preloads, "thumbnail.jpg")
+    command_opts = [output: yt_dlp_filepath]
+    addl_opts = [use_cookies: media_item_with_preloads.source.use_cookies]
 
-    case YtDlpMedia.download_thumbnail(database_record.original_url, output: yt_dlp_filepath) do
+    case YtDlpMedia.download_thumbnail(media_item_with_preloads.original_url, command_opts, addl_opts) do
       {:ok, _} -> real_filepath
       _ -> nil
     end

lib/pinchflat/metadata/source_metadata_storage_worker.ex

@@ -92,7 +92,9 @@ defmodule Pinchflat.Metadata.SourceMetadataStorageWorker do
 
   defp determine_series_directory(source) do
     output_path = DownloadOptionBuilder.build_output_path_for(source)
-    {:ok, %{filepath: filepath}} = MediaCollection.get_source_details(source.original_url, output: output_path)
+    runner_opts = [output: output_path]
+    addl_opts = [use_cookies: source.use_cookies]
+    {:ok, %{filepath: filepath}} = MediaCollection.get_source_details(source.original_url, runner_opts, addl_opts)
 
     case MetadataFileHelpers.series_directory_from_media_filepath(filepath) do
       {:ok, series_directory} -> series_directory
@@ -119,7 +121,7 @@ defmodule Pinchflat.Metadata.SourceMetadataStorageWorker do
         base_opts ++ [:write_thumbnail, playlist_items: 1]
       end
 
-    MediaCollection.get_source_metadata(source.original_url, opts)
+    MediaCollection.get_source_metadata(source.original_url, opts, use_cookies: source.use_cookies)
   end
 
   defp tmp_directory do

lib/pinchflat/slow_indexing/slow_indexing_helpers.ex

@@ -94,7 +94,8 @@ defmodule Pinchflat.SlowIndexing.SlowIndexingHelpers do
     {:ok, pid} = FileFollowerServer.start_link()
 
     handler = fn filepath -> setup_file_follower_watcher(pid, filepath, source) end
-    result = MediaCollection.get_media_attributes_for_collection(source.original_url, file_listener_handler: handler)
+    runner_opts = [file_listener_handler: handler, use_cookies: source.use_cookies]
+    result = MediaCollection.get_media_attributes_for_collection(source.original_url, runner_opts)
 
     FileFollowerServer.stop(pid)
 

lib/pinchflat/sources/source.ex

@@ -27,6 +27,7 @@ defmodule Pinchflat.Sources.Source do
     series_directory
     index_frequency_minutes
     fast_index
+    use_cookies
     download_media
     last_indexed_at
     original_url
@@ -73,6 +74,7 @@ defmodule Pinchflat.Sources.Source do
     field :collection_type, Ecto.Enum, values: [:channel, :playlist]
     field :index_frequency_minutes, :integer, default: 60 * 24
     field :fast_index, :boolean, default: false
+    field :use_cookies, :boolean, default: false
     field :download_media, :boolean, default: true
     field :last_indexed_at, :utc_datetime
     # Only download media items that were published after this date

lib/pinchflat/sources/sources.ex

@@ -180,7 +180,9 @@ defmodule Pinchflat.Sources do
   end
 
   defp add_source_details_to_changeset(source, changeset) do
-    case MediaCollection.get_source_details(changeset.changes.original_url) do
+    use_cookies = Ecto.Changeset.get_field(changeset, :use_cookies)
+
+    case MediaCollection.get_source_details(changeset.changes.original_url, [], use_cookies: use_cookies) do
       {:ok, source_details} ->
         add_source_details_by_collection_type(source, changeset, source_details)
 

lib/pinchflat/yt_dlp/command_runner.ex

@@ -18,6 +18,8 @@ defmodule Pinchflat.YtDlp.CommandRunner do
     - :output_filepath - the path to save the output to. If not provided, a temporary
       file will be created and used. Useful for if you need a reference to the file
       for a file watcher.
+    - :use_cookies - if true, will add a cookie file to the command options. Will not
+      attach a cookie file if the user hasn't set one up.
 
   Returns {:ok, binary()} | {:error, output, status}.
   """
@@ -28,7 +30,7 @@ defmodule Pinchflat.YtDlp.CommandRunner do
 
     output_filepath = generate_output_filepath(addl_opts)
     print_to_file_opts = [{:print_to_file, output_template}, output_filepath]
-    user_configured_opts = cookie_file_options()
+    user_configured_opts = cookie_file_options(addl_opts)
     # These must stay in exactly this order, hence why I'm giving it its own variable.
     all_opts = command_opts ++ print_to_file_opts ++ user_configured_opts ++ global_options()
     formatted_command_opts = [url] ++ CliUtils.parse_options(all_opts)
@@ -78,7 +80,14 @@ defmodule Pinchflat.YtDlp.CommandRunner do
     ]
   end
 
-  defp cookie_file_options do
+  defp cookie_file_options(addl_opts) do
+    case Keyword.get(addl_opts, :use_cookies) do
+      true -> add_cookie_file()
+      _ -> []
+    end
+  end
+
+  defp add_cookie_file do
     base_dir = Application.get_env(:pinchflat, :extras_directory)
     filename_options_map = %{cookies: "cookies.txt"}
 

lib/pinchflat/yt_dlp/media.ex

@@ -37,9 +37,9 @@ defmodule Pinchflat.YtDlp.Media do
   Returns {:ok, map()} | {:error, any, ...}.
   """
   def download(url, command_opts \\ [], addl_opts \\ []) do
-    opts = [:no_simulate] ++ command_opts
+    all_command_opts = [:no_simulate] ++ command_opts
 
-    with {:ok, output} <- backend_runner().run(url, opts, "after_move:%()j", addl_opts),
+    with {:ok, output} <- backend_runner().run(url, all_command_opts, "after_move:%()j", addl_opts),
          {:ok, parsed_json} <- Phoenix.json_library().decode(output) do
       {:ok, parsed_json}
     else
@@ -53,12 +53,12 @@ defmodule Pinchflat.YtDlp.Media do
 
   Returns {:ok, ""} | {:error, any, ...}.
   """
-  def download_thumbnail(url, command_opts \\ []) do
-    opts = [:no_simulate, :skip_download, :write_thumbnail, convert_thumbnail: "jpg"] ++ command_opts
+  def download_thumbnail(url, command_opts \\ [], addl_opts \\ []) do
+    all_command_opts = [:no_simulate, :skip_download, :write_thumbnail, convert_thumbnail: "jpg"] ++ command_opts
 
     # NOTE: it doesn't seem like this command actually returns anything in `after_move` since
     # we aren't downloading the main media file
-    backend_runner().run(url, opts, "after_move:%()j")
+    backend_runner().run(url, all_command_opts, "after_move:%()j", addl_opts)
   end
 
   @doc """
@@ -66,12 +66,12 @@ defmodule Pinchflat.YtDlp.Media do
 
   Returns {:ok, %Media{}} | {:error, any, ...}.
   """
-  def get_media_attributes(url) do
+  def get_media_attributes(url, addl_opts \\ []) do
     runner = Application.get_env(:pinchflat, :yt_dlp_runner)
     command_opts = [:simulate, :skip_download]
     output_template = indexing_output_template()
 
-    case runner.run(url, command_opts, output_template) do
+    case runner.run(url, command_opts, output_template, addl_opts) do
       {:ok, output} ->
         output
         |> Phoenix.json_library().decode!()

lib/pinchflat/yt_dlp/media_collection.ex

@@ -25,15 +25,17 @@ defmodule Pinchflat.YtDlp.MediaCollection do
     # the first video has not released yet (ie: is a premier). We don't care about
     # available formats since we're just getting the media details
     command_opts = [:simulate, :skip_download, :ignore_no_formats_error, :no_warnings]
+    use_cookies = Keyword.get(addl_opts, :use_cookies, false)
     output_template = YtDlpMedia.indexing_output_template()
     output_filepath = FilesystemUtils.generate_metadata_tmpfile(:json)
     file_listener_handler = Keyword.get(addl_opts, :file_listener_handler, false)
+    runner_opts = [output_filepath: output_filepath, use_cookies: use_cookies]
 
     if file_listener_handler do
       file_listener_handler.(output_filepath)
     end
 
-    case runner.run(url, command_opts, output_template, output_filepath: output_filepath) do
+    case runner.run(url, command_opts, output_template, runner_opts) do
       {:ok, output} ->
         parsed_lines =
           output
@@ -64,7 +66,7 @@ defmodule Pinchflat.YtDlp.MediaCollection do
 
   Returns {:ok, map()} | {:error, any, ...}.
   """
-  def get_source_details(source_url, addl_opts \\ []) do
+  def get_source_details(source_url, command_opts \\ [], addl_opts \\ []) do
     # `ignore_no_formats_error` is necessary because yt-dlp will error out if
     # the first video has not released yet (ie: is a premier). We don't care about
     # available formats since we're just getting the source details
@@ -75,10 +77,10 @@ defmodule Pinchflat.YtDlp.MediaCollection do
       playlist_end: 1
     ]
 
-    command_opts = default_opts ++ addl_opts
+    all_command_opts = default_opts ++ command_opts
     output_template = "%(.{channel,channel_id,playlist_id,playlist_title,filename})j"
 
-    with {:ok, output} <- backend_runner().run(source_url, command_opts, output_template),
+    with {:ok, output} <- backend_runner().run(source_url, all_command_opts, output_template, addl_opts),
          {:ok, parsed_json} <- Phoenix.json_library().decode(output) do
       {:ok, format_source_details(parsed_json)}
     else
@@ -109,14 +111,14 @@ defmodule Pinchflat.YtDlp.MediaCollection do
 
   Returns {:ok, map()} | {:error, any, ...}.
   """
-  def get_source_metadata(source_url, addl_opts \\ [playlist_items: 0]) do
+  def get_source_metadata(source_url, command_opts, addl_opts \\ []) do
     # This only validates that the `playlist_items` key is present. It's otherwise unused
-    _playlist_items = Keyword.fetch!(addl_opts, :playlist_items)
+    _playlist_items = Keyword.fetch!(command_opts, :playlist_items)
 
-    opts = [:skip_download] ++ addl_opts
+    all_command_opts = [:skip_download] ++ command_opts
     output_template = "playlist:%()j"
 
-    with {:ok, output} <- backend_runner().run(source_url, opts, output_template),
+    with {:ok, output} <- backend_runner().run(source_url, all_command_opts, output_template, addl_opts),
          {:ok, parsed_json} <- Phoenix.json_library().decode(output) do
       {:ok, parsed_json}
     else