commit
db7e7aae46
|
@ -27,6 +27,7 @@ require "yaml"
|
|||
require "compress/zip"
|
||||
require "protodec/utils"
|
||||
require "./invidious/helpers/*"
|
||||
require "./invidious/yt_backend/*"
|
||||
require "./invidious/*"
|
||||
require "./invidious/channels/*"
|
||||
require "./invidious/routes/**"
|
||||
|
|
|
@ -251,43 +251,6 @@ def html_to_content(description_html : String)
|
|||
return description
|
||||
end
|
||||
|
||||
def extract_videos(initial_data : Hash(String, JSON::Any), author_fallback : String? = nil, author_id_fallback : String? = nil)
|
||||
extracted = extract_items(initial_data, author_fallback, author_id_fallback)
|
||||
|
||||
target = [] of SearchItem
|
||||
extracted.each do |i|
|
||||
if i.is_a?(Category)
|
||||
i.contents.each { |cate_i| target << cate_i if !cate_i.is_a? Video }
|
||||
else
|
||||
target << i
|
||||
end
|
||||
end
|
||||
return target.select(&.is_a?(SearchVideo)).map(&.as(SearchVideo))
|
||||
end
|
||||
|
||||
def extract_selected_tab(tabs)
|
||||
# Extract the selected tab from the array of tabs Youtube returns
|
||||
return selected_target = tabs.as_a.select(&.["tabRenderer"]?.try &.["selected"].as_bool)[0]["tabRenderer"]
|
||||
end
|
||||
|
||||
def fetch_continuation_token(items : Array(JSON::Any))
|
||||
# Fetches the continuation token from an array of items
|
||||
return items.last["continuationItemRenderer"]?
|
||||
.try &.["continuationEndpoint"]["continuationCommand"]["token"].as_s
|
||||
end
|
||||
|
||||
def fetch_continuation_token(initial_data : Hash(String, JSON::Any))
|
||||
# Fetches the continuation token from initial data
|
||||
if initial_data["onResponseReceivedActions"]?
|
||||
continuation_items = initial_data["onResponseReceivedActions"][0]["appendContinuationItemsAction"]["continuationItems"]
|
||||
else
|
||||
tab = extract_selected_tab(initial_data["contents"]["twoColumnBrowseResultsRenderer"]["tabs"])
|
||||
continuation_items = tab["content"]["sectionListRenderer"]["contents"][0]["itemSectionRenderer"]["contents"][0]["gridRenderer"]["items"]
|
||||
end
|
||||
|
||||
return fetch_continuation_token(continuation_items.as_a)
|
||||
end
|
||||
|
||||
def check_enum(db, enum_name, struct_type = nil)
|
||||
return # TODO
|
||||
|
||||
|
|
|
@ -1,70 +1,5 @@
|
|||
require "lsquic"
|
||||
require "db"
|
||||
|
||||
def add_yt_headers(request)
|
||||
request.headers["user-agent"] ||= "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.97 Safari/537.36"
|
||||
request.headers["accept-charset"] ||= "ISO-8859-1,utf-8;q=0.7,*;q=0.7"
|
||||
request.headers["accept"] ||= "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"
|
||||
request.headers["accept-language"] ||= "en-us,en;q=0.5"
|
||||
return if request.resource.starts_with? "/sorry/index"
|
||||
request.headers["x-youtube-client-name"] ||= "1"
|
||||
request.headers["x-youtube-client-version"] ||= "2.20200609"
|
||||
# Preserve original cookies and add new YT consent cookie for EU servers
|
||||
request.headers["cookie"] = "#{request.headers["cookie"]?}; CONSENT=YES+"
|
||||
if !CONFIG.cookies.empty?
|
||||
request.headers["cookie"] = "#{(CONFIG.cookies.map { |c| "#{c.name}=#{c.value}" }).join("; ")}; #{request.headers["cookie"]?}"
|
||||
end
|
||||
end
|
||||
|
||||
struct YoutubeConnectionPool
|
||||
property! url : URI
|
||||
property! capacity : Int32
|
||||
property! timeout : Float64
|
||||
property pool : DB::Pool(QUIC::Client | HTTP::Client)
|
||||
|
||||
def initialize(url : URI, @capacity = 5, @timeout = 5.0, use_quic = true)
|
||||
@url = url
|
||||
@pool = build_pool(use_quic)
|
||||
end
|
||||
|
||||
def client(region = nil, &block)
|
||||
if region
|
||||
conn = make_client(url, region)
|
||||
response = yield conn
|
||||
else
|
||||
conn = pool.checkout
|
||||
begin
|
||||
response = yield conn
|
||||
rescue ex
|
||||
conn.close
|
||||
conn = QUIC::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"
|
||||
response = yield conn
|
||||
ensure
|
||||
pool.release(conn)
|
||||
end
|
||||
end
|
||||
|
||||
response
|
||||
end
|
||||
|
||||
private def build_pool(use_quic)
|
||||
DB::Pool(QUIC::Client | HTTP::Client).new(initial_pool_size: 0, max_pool_size: capacity, max_idle_pool_size: capacity, checkout_timeout: timeout) do
|
||||
if use_quic
|
||||
conn = QUIC::Client.new(url)
|
||||
else
|
||||
conn = HTTP::Client.new(url)
|
||||
end
|
||||
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"
|
||||
conn
|
||||
end
|
||||
end
|
||||
end
|
||||
|
||||
# See http://www.evanmiller.org/how-not-to-sort-by-average-rating.html
|
||||
def ci_lower_bound(pos, n)
|
||||
if n == 0
|
||||
|
@ -85,37 +20,6 @@ def elapsed_text(elapsed)
|
|||
"#{(millis * 1000).round(2)}µs"
|
||||
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"
|
||||
client.read_timeout = 10.seconds
|
||||
client.connect_timeout = 10.seconds
|
||||
|
||||
if region
|
||||
PROXY_LIST[region]?.try &.sample(40).each do |proxy|
|
||||
begin
|
||||
proxy = HTTPProxy.new(proxy_host: proxy[:ip], proxy_port: proxy[:port])
|
||||
client.set_proxy(proxy)
|
||||
break
|
||||
rescue ex
|
||||
end
|
||||
end
|
||||
end
|
||||
|
||||
return client
|
||||
end
|
||||
|
||||
def make_client(url : URI, region = nil, &block)
|
||||
client = make_client(url, region)
|
||||
begin
|
||||
yield client
|
||||
ensure
|
||||
client.close
|
||||
end
|
||||
end
|
||||
|
||||
def decode_length_seconds(string)
|
||||
length_seconds = string.gsub(/[^0-9:]/, "").split(":").map &.to_i
|
||||
length_seconds = [0] * (3 - length_seconds.size) + length_seconds
|
||||
|
|
96
src/invidious/yt_backend/connection_pool.cr
Normal file
96
src/invidious/yt_backend/connection_pool.cr
Normal file
|
@ -0,0 +1,96 @@
|
|||
require "lsquic"
|
||||
|
||||
def add_yt_headers(request)
|
||||
request.headers["user-agent"] ||= "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.97 Safari/537.36"
|
||||
request.headers["accept-charset"] ||= "ISO-8859-1,utf-8;q=0.7,*;q=0.7"
|
||||
request.headers["accept"] ||= "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"
|
||||
request.headers["accept-language"] ||= "en-us,en;q=0.5"
|
||||
return if request.resource.starts_with? "/sorry/index"
|
||||
request.headers["x-youtube-client-name"] ||= "1"
|
||||
request.headers["x-youtube-client-version"] ||= "2.20200609"
|
||||
# Preserve original cookies and add new YT consent cookie for EU servers
|
||||
request.headers["cookie"] = "#{request.headers["cookie"]?}; CONSENT=YES+"
|
||||
if !CONFIG.cookies.empty?
|
||||
request.headers["cookie"] = "#{(CONFIG.cookies.map { |c| "#{c.name}=#{c.value}" }).join("; ")}; #{request.headers["cookie"]?}"
|
||||
end
|
||||
end
|
||||
|
||||
struct YoutubeConnectionPool
|
||||
property! url : URI
|
||||
property! capacity : Int32
|
||||
property! timeout : Float64
|
||||
property pool : DB::Pool(QUIC::Client | HTTP::Client)
|
||||
|
||||
def initialize(url : URI, @capacity = 5, @timeout = 5.0, use_quic = true)
|
||||
@url = url
|
||||
@pool = build_pool(use_quic)
|
||||
end
|
||||
|
||||
def client(region = nil, &block)
|
||||
if region
|
||||
conn = make_client(url, region)
|
||||
response = yield conn
|
||||
else
|
||||
conn = pool.checkout
|
||||
begin
|
||||
response = yield conn
|
||||
rescue ex
|
||||
conn.close
|
||||
conn = QUIC::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"
|
||||
response = yield conn
|
||||
ensure
|
||||
pool.release(conn)
|
||||
end
|
||||
end
|
||||
|
||||
response
|
||||
end
|
||||
|
||||
private def build_pool(use_quic)
|
||||
DB::Pool(QUIC::Client | HTTP::Client).new(initial_pool_size: 0, max_pool_size: capacity, max_idle_pool_size: capacity, checkout_timeout: timeout) do
|
||||
if use_quic
|
||||
conn = QUIC::Client.new(url)
|
||||
else
|
||||
conn = HTTP::Client.new(url)
|
||||
end
|
||||
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"
|
||||
conn
|
||||
end
|
||||
end
|
||||
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"
|
||||
client.read_timeout = 10.seconds
|
||||
client.connect_timeout = 10.seconds
|
||||
|
||||
if region
|
||||
PROXY_LIST[region]?.try &.sample(40).each do |proxy|
|
||||
begin
|
||||
proxy = HTTPProxy.new(proxy_host: proxy[:ip], proxy_port: proxy[:port])
|
||||
client.set_proxy(proxy)
|
||||
break
|
||||
rescue ex
|
||||
end
|
||||
end
|
||||
end
|
||||
|
||||
return client
|
||||
end
|
||||
|
||||
def make_client(url : URI, region = nil, &block)
|
||||
client = make_client(url, region)
|
||||
begin
|
||||
yield client
|
||||
ensure
|
||||
client.close
|
||||
end
|
||||
end
|
|
@ -533,37 +533,6 @@ private module HelperExtractors
|
|||
end
|
||||
end
|
||||
|
||||
# Extracts text from InnerTube response
|
||||
#
|
||||
# InnerTube can package text in three different formats
|
||||
# "runs": [
|
||||
# {"text": "something"},
|
||||
# {"text": "cont"},
|
||||
# ...
|
||||
# ]
|
||||
#
|
||||
# "SimpleText": "something"
|
||||
#
|
||||
# Or sometimes just none at all as with the data returned from
|
||||
# category continuations.
|
||||
#
|
||||
# In order to facilitate calling this function with `#[]?`:
|
||||
# A nil will be accepted. Of course, since nil cannot be parsed,
|
||||
# another nil will be returned.
|
||||
def extract_text(item : JSON::Any?) : String?
|
||||
if item.nil?
|
||||
return nil
|
||||
end
|
||||
|
||||
if text_container = item["simpleText"]?
|
||||
return text_container.as_s
|
||||
elsif text_container = item["runs"]?
|
||||
return text_container.as_a.map(&.["text"].as_s).join("")
|
||||
else
|
||||
nil
|
||||
end
|
||||
end
|
||||
|
||||
# Parses an item from Youtube's JSON response into a more usable structure.
|
||||
# The end result can either be a SearchVideo, SearchPlaylist or SearchChannel.
|
||||
def extract_item(item : JSON::Any, author_fallback : String? = "",
|
67
src/invidious/yt_backend/extractors_utils.cr
Normal file
67
src/invidious/yt_backend/extractors_utils.cr
Normal file
|
@ -0,0 +1,67 @@
|
|||
# Extracts text from InnerTube response
|
||||
#
|
||||
# InnerTube can package text in three different formats
|
||||
# "runs": [
|
||||
# {"text": "something"},
|
||||
# {"text": "cont"},
|
||||
# ...
|
||||
# ]
|
||||
#
|
||||
# "SimpleText": "something"
|
||||
#
|
||||
# Or sometimes just none at all as with the data returned from
|
||||
# category continuations.
|
||||
#
|
||||
# In order to facilitate calling this function with `#[]?`:
|
||||
# A nil will be accepted. Of course, since nil cannot be parsed,
|
||||
# another nil will be returned.
|
||||
def extract_text(item : JSON::Any?) : String?
|
||||
if item.nil?
|
||||
return nil
|
||||
end
|
||||
|
||||
if text_container = item["simpleText"]?
|
||||
return text_container.as_s
|
||||
elsif text_container = item["runs"]?
|
||||
return text_container.as_a.map(&.["text"].as_s).join("")
|
||||
else
|
||||
nil
|
||||
end
|
||||
end
|
||||
|
||||
def extract_videos(initial_data : Hash(String, JSON::Any), author_fallback : String? = nil, author_id_fallback : String? = nil)
|
||||
extracted = extract_items(initial_data, author_fallback, author_id_fallback)
|
||||
|
||||
target = [] of SearchItem
|
||||
extracted.each do |i|
|
||||
if i.is_a?(Category)
|
||||
i.contents.each { |cate_i| target << cate_i if !cate_i.is_a? Video }
|
||||
else
|
||||
target << i
|
||||
end
|
||||
end
|
||||
return target.select(&.is_a?(SearchVideo)).map(&.as(SearchVideo))
|
||||
end
|
||||
|
||||
def extract_selected_tab(tabs)
|
||||
# Extract the selected tab from the array of tabs Youtube returns
|
||||
return selected_target = tabs.as_a.select(&.["tabRenderer"]?.try &.["selected"].as_bool)[0]["tabRenderer"]
|
||||
end
|
||||
|
||||
def fetch_continuation_token(items : Array(JSON::Any))
|
||||
# Fetches the continuation token from an array of items
|
||||
return items.last["continuationItemRenderer"]?
|
||||
.try &.["continuationEndpoint"]["continuationCommand"]["token"].as_s
|
||||
end
|
||||
|
||||
def fetch_continuation_token(initial_data : Hash(String, JSON::Any))
|
||||
# Fetches the continuation token from initial data
|
||||
if initial_data["onResponseReceivedActions"]?
|
||||
continuation_items = initial_data["onResponseReceivedActions"][0]["appendContinuationItemsAction"]["continuationItems"]
|
||||
else
|
||||
tab = extract_selected_tab(initial_data["contents"]["twoColumnBrowseResultsRenderer"]["tabs"])
|
||||
continuation_items = tab["content"]["sectionListRenderer"]["contents"][0]["itemSectionRenderer"]["contents"][0]["gridRenderer"]["items"]
|
||||
end
|
||||
|
||||
return fetch_continuation_token(continuation_items.as_a)
|
||||
end
|
Loading…
Reference in a new issue