diff options
Diffstat (limited to 'src')
| -rw-r--r-- | src/invidious.cr | 2 | ||||
| -rw-r--r-- | src/invidious/channels/community.cr | 72 | ||||
| -rw-r--r-- | src/invidious/config.cr | 2 | ||||
| -rw-r--r-- | src/invidious/routes/images.cr | 142 | ||||
| -rw-r--r-- | src/invidious/videos/parser.cr | 10 | ||||
| -rw-r--r-- | src/invidious/yt_backend/connection_pool.cr | 37 | ||||
| -rw-r--r-- | src/invidious/yt_backend/extractors.cr | 26 | ||||
| -rw-r--r-- | src/invidious/yt_backend/youtube_api.cr | 14 |
8 files changed, 89 insertions, 216 deletions
diff --git a/src/invidious.cr b/src/invidious.cr index 84e1895d..e0bd0101 100644 --- a/src/invidious.cr +++ b/src/invidious.cr @@ -90,7 +90,7 @@ SOFTWARE = { "branch" => "#{CURRENT_BRANCH}", } -YT_POOL = YoutubeConnectionPool.new(YT_URL, capacity: CONFIG.pool_size, use_quic: CONFIG.use_quic) +YT_POOL = YoutubeConnectionPool.new(YT_URL, capacity: CONFIG.pool_size) # CLI Kemal.config.extra_options do |parser| diff --git a/src/invidious/channels/community.cr b/src/invidious/channels/community.cr index aac4bc8a..791f1641 100644 --- a/src/invidious/channels/community.cr +++ b/src/invidious/channels/community.cr @@ -1,49 +1,31 @@ private IMAGE_QUALITIES = {320, 560, 640, 1280, 2000} # TODO: Add "sort_by" -def fetch_channel_community(ucid, continuation, locale, format, thin_mode) - response = YT_POOL.client &.get("/channel/#{ucid}/community?gl=US&hl=en") - if response.status_code != 200 - response = YT_POOL.client &.get("/user/#{ucid}/community?gl=US&hl=en") - end - - if response.status_code != 200 - raise NotFoundException.new("This channel does not exist.") - end - - ucid = response.body.match(/https:\/\/www.youtube.com\/channel\/(?<ucid>UC[a-zA-Z0-9_-]{22})/).not_nil!["ucid"] - - if !continuation || continuation.empty? - initial_data = extract_initial_data(response.body) - body = extract_selected_tab(initial_data["contents"]["twoColumnBrowseResultsRenderer"]["tabs"])["content"]["sectionListRenderer"]["contents"][0]["itemSectionRenderer"] - - if !body - raise InfoException.new("Could not extract community tab.") +def fetch_channel_community(ucid, cursor, locale, format, thin_mode) + if cursor.nil? + # Egljb21tdW5pdHk%3D is the protobuf object to load "community" + initial_data = YoutubeAPI.browse(ucid, params: "Egljb21tdW5pdHk%3D") + + items = [] of JSON::Any + extract_items(initial_data) do |item| + items << item end else - continuation = produce_channel_community_continuation(ucid, continuation) + continuation = produce_channel_community_continuation(ucid, cursor) + initial_data = YoutubeAPI.browse(continuation: continuation) - headers = HTTP::Headers.new - headers["cookie"] = response.cookies.add_request_headers(headers)["cookie"] + container = initial_data.dig?("continuationContents", "itemSectionContinuation", "contents") - session_token = response.body.match(/"XSRF_TOKEN":"(?<session_token>[^"]+)"/).try &.["session_token"]? || "" - post_req = { - session_token: session_token, - } + raise InfoException.new("Can't extract community data") if container.nil? - body = YoutubeAPI.browse(continuation) - - body = body.dig?("continuationContents", "itemSectionContinuation") || - body.dig?("continuationContents", "backstageCommentsContinuation") - - if !body - raise InfoException.new("Could not extract continuation.") - end + items = container.as_a end - posts = body["contents"].as_a + return extract_channel_community(items, ucid: ucid, locale: locale, format: format, thin_mode: thin_mode) +end - if message = posts[0]["messageRenderer"]? +def extract_channel_community(items, *, ucid, locale, format, thin_mode) + if message = items[0]["messageRenderer"]? error_message = (message["text"]["simpleText"]? || message["text"]["runs"]?.try &.[0]?.try &.["text"]?) .try &.as_s || "" @@ -59,7 +41,7 @@ def fetch_channel_community(ucid, continuation, locale, format, thin_mode) json.field "authorId", ucid json.field "comments" do json.array do - posts.each do |post| + items.each do |post| comments = post["backstagePostThreadRenderer"]?.try &.["comments"]? || post["backstageCommentsContinuation"]? @@ -216,6 +198,22 @@ def fetch_channel_community(ucid, continuation, locale, format, thin_mode) parse_item(attachment) .as(SearchPlaylist) .to_json(locale, json) + when .has_key?("quizRenderer") + json.object do + attachment = attachment["quizRenderer"] + json.field "type", "quiz" + json.field "totalVotes", short_text_to_number(attachment["totalVotes"]["simpleText"].as_s.split(" ")[0]) + json.field "choices" do + json.array do + attachment["choices"].as_a.each do |choice| + json.object do + json.field "text", choice.dig("text", "runs", 0, "text").as_s + json.field "isCorrect", choice["isCorrect"].as_bool + end + end + end + end + end else json.object do json.field "type", "unknown" @@ -242,7 +240,7 @@ def fetch_channel_community(ucid, continuation, locale, format, thin_mode) end end end - if cont = posts.dig?(-1, "continuationItemRenderer", "continuationEndpoint", "continuationCommand", "token") + if cont = items.dig?(-1, "continuationItemRenderer", "continuationEndpoint", "continuationCommand", "token") json.field "continuation", extract_channel_community_cursor(cont.as_s) end end diff --git a/src/invidious/config.cr b/src/invidious/config.cr index e5f1e822..cee33ce1 100644 --- a/src/invidious/config.cr +++ b/src/invidious/config.cr @@ -126,8 +126,6 @@ class Config property host_binding : String = "0.0.0.0" # Pool size for HTTP requests to youtube.com and ytimg.com (each domain has a separate pool of `pool_size`) property pool_size : Int32 = 100 - # Use quic transport for youtube api - property use_quic : Bool = false # Saved cookies in "name1=value1; name2=value2..." format @[YAML::Field(converter: Preferences::StringToCookies)] diff --git a/src/invidious/routes/images.cr b/src/invidious/routes/images.cr index 594a7869..b6a2e110 100644 --- a/src/invidious/routes/images.cr +++ b/src/invidious/routes/images.cr @@ -3,17 +3,7 @@ module Invidious::Routes::Images def self.ggpht(env) url = env.request.path.lchop("/ggpht") - headers = ( - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - HTTP::Headers{":authority" => "yt3.ggpht.com"} - else - HTTP::Headers.new - end - {% else %} - HTTP::Headers.new - {% end %} - ) + headers = HTTP::Headers.new REQUEST_HEADERS_WHITELIST.each do |header| if env.request.headers[header]? @@ -42,22 +32,9 @@ module Invidious::Routes::Images } begin - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - YT_POOL.client &.get(url, headers) do |resp| - return request_proc.call(resp) - end - else - HTTP::Client.get("https://yt3.ggpht.com#{url}") do |resp| - return request_proc.call(resp) - end - end - {% else %} - # This can likely be optimized into a (small) pool sometime in the future. - HTTP::Client.get("https://yt3.ggpht.com#{url}") do |resp| - return request_proc.call(resp) - end - {% end %} + HTTP::Client.get("https://yt3.ggpht.com#{url}") do |resp| + return request_proc.call(resp) + end rescue ex end end @@ -78,10 +55,6 @@ module Invidious::Routes::Images headers = HTTP::Headers.new - {% unless flag?(:disable_quic) %} - headers[":authority"] = "#{authority}.ytimg.com" - {% end %} - REQUEST_HEADERS_WHITELIST.each do |header| if env.request.headers[header]? headers[header] = env.request.headers[header] @@ -107,22 +80,9 @@ module Invidious::Routes::Images } begin - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - YT_POOL.client &.get(url, headers) do |resp| - return request_proc.call(resp) - end - else - HTTP::Client.get("https://#{authority}.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - end - {% else %} - # This can likely be optimized into a (small) pool sometime in the future. - HTTP::Client.get("https://#{authority}.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - {% end %} + HTTP::Client.get("https://#{authority}.ytimg.com#{url}") do |resp| + return request_proc.call(resp) + end rescue ex end end @@ -133,17 +93,7 @@ module Invidious::Routes::Images name = env.params.url["name"] url = env.request.resource - headers = ( - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - HTTP::Headers{":authority" => "i9.ytimg.com"} - else - HTTP::Headers.new - end - {% else %} - HTTP::Headers.new - {% end %} - ) + headers = HTTP::Headers.new REQUEST_HEADERS_WHITELIST.each do |header| if env.request.headers[header]? @@ -169,22 +119,9 @@ module Invidious::Routes::Images } begin - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - YT_POOL.client &.get(url, headers) do |resp| - return request_proc.call(resp) - end - else - HTTP::Client.get("https://i9.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - end - {% else %} - # This can likely be optimized into a (small) pool sometime in the future. - HTTP::Client.get("https://i9.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - {% end %} + HTTP::Client.get("https://i9.ytimg.com#{url}") do |resp| + return request_proc.call(resp) + end rescue ex end end @@ -223,41 +160,16 @@ module Invidious::Routes::Images id = env.params.url["id"] name = env.params.url["name"] - headers = ( - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - HTTP::Headers{":authority" => "i.ytimg.com"} - else - HTTP::Headers.new - end - {% else %} - HTTP::Headers.new - {% end %} - ) + headers = HTTP::Headers.new if name == "maxres.jpg" build_thumbnails(id).each do |thumb| thumbnail_resource_path = "/vi/#{id}/#{thumb[:url]}.jpg" - # Logic here is short enough that manually typing them out should be fine. - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - if YT_POOL.client &.head(thumbnail_resource_path, headers).status_code == 200 - name = thumb[:url] + ".jpg" - break - end - else - if HTTP::Client.head("https://i.ytimg.com#{thumbnail_resource_path}").status_code == 200 - name = thumb[:url] + ".jpg" - break - end - end - {% else %} - # This can likely be optimized into a (small) pool sometime in the future. - if HTTP::Client.head("https://i.ytimg.com#{thumbnail_resource_path}").status_code == 200 - name = thumb[:url] + ".jpg" - break - end - {% end %} + # This can likely be optimized into a (small) pool sometime in the future. + if HTTP::Client.head("https://i.ytimg.com#{thumbnail_resource_path}").status_code == 200 + name = thumb[:url] + ".jpg" + break + end end end @@ -287,22 +199,10 @@ module Invidious::Routes::Images } begin - {% unless flag?(:disable_quic) %} - if CONFIG.use_quic - YT_POOL.client &.get(url, headers) do |resp| - return request_proc.call(resp) - end - else - HTTP::Client.get("https://i.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - end - {% else %} - # This can likely be optimized into a (small) pool sometime in the future. - HTTP::Client.get("https://i.ytimg.com#{url}") do |resp| - return request_proc.call(resp) - end - {% end %} + # This can likely be optimized into a (small) pool sometime in the future. + HTTP::Client.get("https://i.ytimg.com#{url}") do |resp| + return request_proc.call(resp) + end rescue ex end end diff --git a/src/invidious/videos/parser.cr b/src/invidious/videos/parser.cr index 9cc0ffdc..2a09d187 100644 --- a/src/invidious/videos/parser.cr +++ b/src/invidious/videos/parser.cr @@ -55,8 +55,9 @@ def extract_video_info(video_id : String, proxy_region : String? = nil) client_config = YoutubeAPI::ClientConfig.new(proxy_region: proxy_region) # Fetch data from the player endpoint - # 8AEB param is used to fetch YouTube stories - player_response = YoutubeAPI.player(video_id: video_id, params: "8AEB", client_config: client_config) + # CgIQBg is a workaround for streaming URLs that returns a 403. + # See https://github.com/iv-org/invidious/issues/4027#issuecomment-1666944520 + player_response = YoutubeAPI.player(video_id: video_id, params: "CgIQBg", client_config: client_config) playability_status = player_response.dig?("playabilityStatus", "status").try &.as_s @@ -135,8 +136,9 @@ end def try_fetch_streaming_data(id : String, client_config : YoutubeAPI::ClientConfig) : Hash(String, JSON::Any)? LOGGER.debug("try_fetch_streaming_data: [#{id}] Using #{client_config.client_type} client.") - # 8AEB param is used to fetch YouTube stories - response = YoutubeAPI.player(video_id: id, params: "8AEB", client_config: client_config) + # CgIQBg is a workaround for streaming URLs that returns a 403. + # See https://github.com/iv-org/invidious/issues/4027#issuecomment-1666944520 + response = YoutubeAPI.player(video_id: id, params: "CgIQBg", client_config: client_config) playability_status = response["playabilityStatus"]["status"] LOGGER.debug("try_fetch_streaming_data: [#{id}] Got playabilityStatus == #{playability_status}.") diff --git a/src/invidious/yt_backend/connection_pool.cr b/src/invidious/yt_backend/connection_pool.cr index 658731cf..e9eb726c 100644 --- a/src/invidious/yt_backend/connection_pool.cr +++ b/src/invidious/yt_backend/connection_pool.cr @@ -1,11 +1,3 @@ -{% unless flag?(:disable_quic) %} - require "lsquic" - - alias HTTPClientType = QUIC::Client | HTTP::Client -{% else %} - alias HTTPClientType = HTTP::Client -{% end %} - def add_yt_headers(request) if request.headers["User-Agent"] == "Crystal" request.headers["User-Agent"] ||= "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36" @@ -26,11 +18,11 @@ struct YoutubeConnectionPool property! url : URI property! capacity : Int32 property! timeout : Float64 - property pool : DB::Pool(HTTPClientType) + property pool : DB::Pool(HTTP::Client) - def initialize(url : URI, @capacity = 5, @timeout = 5.0, use_quic = true) + def initialize(url : URI, @capacity = 5, @timeout = 5.0) @url = url - @pool = build_pool(use_quic) + @pool = build_pool() end def client(region = nil, &block) @@ -43,11 +35,7 @@ struct YoutubeConnectionPool response = yield conn rescue ex conn.close - {% unless flag?(:disable_quic) %} - conn = CONFIG.use_quic ? QUIC::Client.new(url) : HTTP::Client.new(url) - {% else %} - conn = HTTP::Client.new(url) - {% end %} + conn = HTTP::Client.new(url) conn.family = (url.host == "www.youtube.com") ? CONFIG.force_resolve : Socket::Family::INET conn.family = Socket::Family::INET if conn.family == Socket::Family::UNSPEC @@ -61,19 +49,9 @@ struct YoutubeConnectionPool response end - private def build_pool(use_quic) - DB::Pool(HTTPClientType).new(initial_pool_size: 0, max_pool_size: capacity, max_idle_pool_size: capacity, checkout_timeout: timeout) do - conn = nil # Declare - {% unless flag?(:disable_quic) %} - if use_quic - conn = QUIC::Client.new(url) - else - conn = HTTP::Client.new(url) - end - {% else %} - conn = HTTP::Client.new(url) - {% end %} - + private def build_pool + DB::Pool(HTTP::Client).new(initial_pool_size: 0, max_pool_size: capacity, max_idle_pool_size: capacity, checkout_timeout: timeout) do + conn = HTTP::Client.new(url) conn.family = (url.host == "www.youtube.com") ? CONFIG.force_resolve : Socket::Family::INET conn.family = Socket::Family::INET if conn.family == Socket::Family::UNSPEC conn.before_request { |r| add_yt_headers(r) } if url.host == "www.youtube.com" @@ -83,7 +61,6 @@ struct YoutubeConnectionPool end def make_client(url : URI, region = nil) - # TODO: Migrate any applicable endpoints to QUIC client = HTTPClient.new(url, OpenSSL::SSL::Context::Client.insecure) client.family = (url.host == "www.youtube.com") ? CONFIG.force_resolve : Socket::Family::UNSPEC client.before_request { |r| add_yt_headers(r) } if url.host == "www.youtube.com" diff --git a/src/invidious/yt_backend/extractors.cr b/src/invidious/yt_backend/extractors.cr index 8456313b..aaf7772e 100644 --- a/src/invidious/yt_backend/extractors.cr +++ b/src/invidious/yt_backend/extractors.cr @@ -659,19 +659,25 @@ private module Extractors private def self.unpack_section_list(contents) raw_items = [] of JSON::Any - contents.as_a.each do |renderer_container| - renderer_container_contents = renderer_container["itemSectionRenderer"]["contents"][0] - - # Category extraction - if items_container = renderer_container_contents["shelfRenderer"]? - raw_items << renderer_container_contents - next - elsif items_container = renderer_container_contents["gridRenderer"]? + contents.as_a.each do |item| + if item_section_content = item.dig?("itemSectionRenderer", "contents") + raw_items += self.unpack_item_section(item_section_content) else - items_container = renderer_container_contents + raw_items << item end + end - items_container["items"]?.try &.as_a.each do |item| + return raw_items + end + + private def self.unpack_item_section(contents) + raw_items = [] of JSON::Any + + contents.as_a.each do |item| + # Category extraction + if container = item.dig?("gridRenderer", "items") || item.dig?("items") + raw_items += container.as_a + else raw_items << item end end diff --git a/src/invidious/yt_backend/youtube_api.cr b/src/invidious/yt_backend/youtube_api.cr index 3dd9e9d8..aef9ddd9 100644 --- a/src/invidious/yt_backend/youtube_api.cr +++ b/src/invidious/yt_backend/youtube_api.cr @@ -595,17 +595,9 @@ module YoutubeAPI LOGGER.trace("YoutubeAPI: POST data: #{data}") # Send the POST request - if {{ !flag?(:disable_quic) }} && CONFIG.use_quic - # Using QUIC client - body = YT_POOL.client(client_config.proxy_region, - &.post(url, headers: headers, body: data.to_json) - ).body - else - # Using HTTP client - body = YT_POOL.client(client_config.proxy_region) do |client| - client.post(url, headers: headers, body: data.to_json) do |response| - self._decompress(response.body_io, response.headers["Content-Encoding"]?) - end + body = YT_POOL.client(client_config.proxy_region) do |client| + client.post(url, headers: headers, body: data.to_json) do |response| + self._decompress(response.body_io, response.headers["Content-Encoding"]?) end end |