Refactor 'decode_date'

pull/46/head
Omar Roth 2018-07-28 07:50:22 -05:00
rodzic e767651836
commit 9355c60f6e
2 zmienionych plików z 32 dodań i 51 usunięć

Wyświetl plik

@ -895,31 +895,13 @@ get "/api/v1/trending" do |env|
published, views = node.xpath_nodes(%q(.//ul[@class="yt-lockup-meta-info"]/li)) published, views = node.xpath_nodes(%q(.//ul[@class="yt-lockup-meta-info"]/li))
views = views.content.rchop(" views").delete(",").to_i views = views.content.rchop(" views").delete(",").to_i
published = published.content
published = published.split(" ")
published = published[-3..-1].join(" ")
descriptionHtml = node.xpath_node(%q(.//div[contains(@class, "yt-lockup-description")])).not_nil!.to_s descriptionHtml = node.xpath_node(%q(.//div[contains(@class, "yt-lockup-description")])).not_nil!.to_s
description = descriptionHtml.gsub("<br>", "\n") description = descriptionHtml.gsub("<br>", "\n")
description = description.gsub("<br/>", "\n") description = description.gsub("<br/>", "\n")
description = XML.parse_html(description) description = XML.parse_html(description)
# Time matches format "20 hours ago", "40 minutes ago"... published = published.content.split(" ")[-3..-1].join(" ")
delta = published.split(" ")[0].to_i published = decode_date(published)
case published
when .includes? "minute"
published = Time.now - delta.minutes
when .includes? "hour"
published = Time.now - delta.hours
when .includes? "day"
published = Time.now - delta.days
when .includes? "week"
published = Time.now - delta.weeks
when .includes? "month"
published = Time.now - delta.weeks
else
raise "Could not parse #{published}"
end
json.object do json.object do
json.field "title", title json.field "title", title

Wyświetl plik

@ -704,46 +704,22 @@ def fetch_channel(ucid, client, db, pull_all_videos = true)
end end
document = XML.parse_html(content_html) document = XML.parse_html(content_html)
document.xpath_nodes(%q(//li[contains(@class, "channels-content-item")])).each do |item| document.xpath_nodes(%q(//li[contains(@class, "feed-item-container")])).each do |item|
root = item.xpath_node(%q(div/div/div[@class="yt-lockup-content"])) anchor = item.xpath_node(%q(.//h3[contains(@class,"yt-lockup-title")]/a))
if !root
raise "could not find root"
end
anchor = root.xpath_node(%q(h3[contains(@class,"yt-lockup-title")]/a))
if !anchor if !anchor
raise "could not find anchor" raise "could not find anchor"
end end
title = anchor.content.strip title = anchor.content.strip
video_id = anchor["href"].lchop("/watch?v=") video_id = anchor["href"].lchop("/watch?v=")
published = root.xpath_node(%q(div[@class="yt-lockup-meta"]/ul/li[2])) published = item.xpath_node(%q(.//div[@class="yt-lockup-meta"]/ul/li[1]))
if !published if !published
# This happens on Youtube red videos, here we just skip them # This happens on Youtube red videos, here we just skip them
next next
end end
published = published.content.split(" ") published = published.content
span = published[0].to_i published = decode_date(published)
case published[1]
when .includes? "second"
span = span.seconds
when .includes? "minute"
span = span.minutes
when .includes? "hour"
span = span.hours
when .includes? "day"
span = span.days
when .includes? "week"
span = span.weeks
when .includes? "month"
span = span.months
when .includes? "year"
span = span.years
else
raise "Unrecognized time: #{published[1]}"
end
published = Time.now - span
videos << ChannelVideo.new(video_id, title, published, Time.now, ucid, author) videos << ChannelVideo.new(video_id, title, published, Time.now, ucid, author)
end end
@ -863,6 +839,29 @@ def decode_time(string)
return time return time
end end
def decode_date(date : String)
# Time matches format "20 hours ago", "40 minutes ago"...
delta = date.split(" ")[0].to_i
case date
when .includes? "minute"
delta = delta.minutes
when .includes? "hour"
delta = delta.hours
when .includes? "day"
delta = delta.days
when .includes? "week"
delta = delta.weeks
when .includes? "month"
delta = delta.months
when .includes? "year"
delta = delta.years
else
raise "Could not parse #{date}"
end
return Time.now - delta
end
def produce_playlist_url(ucid, index) def produce_playlist_url(ucid, index)
ucid = ucid.lchop("UC") ucid = ucid.lchop("UC")
ucid = "VLUU" + ucid ucid = "VLUU" + ucid
@ -899,7 +898,7 @@ end
def produce_videos_url(ucid, page) def produce_videos_url(ucid, page)
page = "#{page}" page = "#{page}"
meta = "\x12\x06videos \x00\x30\x01\x38\x01\x60\x01\x6a\x00\x7a" meta = "\x12\x06videos \x00\x30\x02\x38\x01\x60\x01\x6a\x00\x7a"
meta += page.size.to_u8.unsafe_chr meta += page.size.to_u8.unsafe_chr
meta += page meta += page
meta += "\xb8\x01\x00" meta += "\xb8\x01\x00"