Fix channel info extract for 'video game' channels

This commit is contained in:
syeopite 2021-04-02 20:58:39 -07:00
parent c0997f951c
commit fe4eef5855
No known key found for this signature in database
GPG key ID: 6FA616E5A5294A82

View file

@ -821,6 +821,19 @@ def get_about_info(ucid, locale)
raise ChannelRedirect.new(channel_id: browse_endpoint["browseId"].to_s)
end
auto_generated = false
if !initdata.has_key?("metadata")
auto_generated = true
end
if auto_generated
return get_auto_generated_channel_info(initdata, about, ucid)
else
return get_normal_channel_info(initdata, about)
end
end
def get_normal_channel_info(initdata, about)
author = initdata["metadata"]["channelMetadataRenderer"]["title"].as_s
author_url = initdata["metadata"]["channelMetadataRenderer"]["channelUrl"].as_s
author_thumbnail = initdata["metadata"]["channelMetadataRenderer"]["avatar"]["thumbnails"][0]["url"].as_s
@ -929,6 +942,69 @@ def get_about_info(ucid, locale)
})
end
def get_auto_generated_channel_info(initdata, about, ucid)
author = initdata["header"]["interactiveTabbedHeaderRenderer"]["title"]["simpleText"].as_s
author_url = initdata["microformat"]["microformatDataRenderer"]["urlCanonical"].as_s
author_thumbnail = initdata["header"]["interactiveTabbedHeaderRenderer"]["boxArt"]["thumbnails"][0]["url"].as_s
# Raises a KeyError on failure.
banners = initdata["header"]["interactiveTabbedHeaderRenderer"]?.try &.["banner"]?.try &.["thumbnails"]?
banner = banners.try &.[-1]?.try &.["url"].as_s?
description = initdata["header"]["interactiveTabbedHeaderRenderer"]["description"]["simpleText"].as_s
description_html = HTML.escape(description).gsub("\n", "<br>")
paid = false
is_family_friendly = initdata["microformat"]["microformatDataRenderer"]["familySafe"].as_bool
allowed_regions = initdata["microformat"]["microformatDataRenderer"]["availableCountries"].as_a.map { |a| a.as_s }
related_channels = [] of AboutRelatedChannel
total_views = 0_i64
joined = Time.unix(0)
tabs = [] of String
auto_generated = true
tabs_json = initdata["contents"]["twoColumnBrowseResultsRenderer"]["tabs"]?.try &.as_a?
if !tabs_json.nil?
# Retrieve information from the tabs array. The index we are looking for varies between channels.
tabs_json.each do |node|
# Try to find the about section which is located in only one of the tabs.
channel_about_meta = node["tabRenderer"]?.try &.["content"]?.try &.["sectionListRenderer"]?
.try &.["contents"]?.try &.[0]?.try &.["itemSectionRenderer"]?.try &.["contents"]?
.try &.[0]?.try &.["channelAboutFullMetadataRenderer"]?
if !channel_about_meta.nil?
# The joined text is split to several sub strings. The reduce joins those strings before parsing the date.
joined = channel_about_meta["joinedDateText"]?.try &.["runs"]?.try &.as_a.reduce("") { |acc, node| acc + node["text"].as_s }
.try { |text| Time.parse(text, "Joined %b %-d, %Y", Time::Location.local) } || Time.unix(0)
end
end
tabs = tabs_json.reject { |node| node["tabRenderer"]?.nil? }.map { |node| node["tabRenderer"]["title"].as_s.downcase }
end
sub_count = initdata["header"]["c4TabbedHeaderRenderer"]?.try &.["subscriberCountText"]?.try &.["simpleText"]?.try &.as_s?
.try { |text| short_text_to_number(text.split(" ")[0]) } || 0
AboutChannel.new({
ucid: ucid,
author: author,
auto_generated: auto_generated,
author_url: author_url,
author_thumbnail: author_thumbnail,
banner: banner,
description_html: description_html,
paid: paid,
total_views: total_views,
sub_count: sub_count,
joined: joined,
is_family_friendly: is_family_friendly,
allowed_regions: allowed_regions,
related_channels: related_channels,
tabs: tabs,
})
end
def get_channel_videos_response(ucid, page = 1, auto_generated = nil, sort_by = "newest")
continuation = produce_channel_videos_continuation(ucid, page,
auto_generated: auto_generated, sort_by: sort_by, v2: true)