From 9355c60f6e0121deff51c534b28eff89617aab17 Mon Sep 17 00:00:00 2001 From: Omar Roth Date: Sat, 28 Jul 2018 07:50:22 -0500 Subject: [PATCH] Refactor 'decode_date' --- src/invidious.cr | 22 ++------------- src/invidious/helpers.cr | 61 ++++++++++++++++++++-------------------- 2 files changed, 32 insertions(+), 51 deletions(-) diff --git a/src/invidious.cr b/src/invidious.cr index dd21108cc..dfc01d0f8 100644 --- a/src/invidious.cr +++ b/src/invidious.cr @@ -895,31 +895,13 @@ get "/api/v1/trending" do |env| published, views = node.xpath_nodes(%q(.//ul[@class="yt-lockup-meta-info"]/li)) 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 description = descriptionHtml.gsub("
", "\n") description = description.gsub("
", "\n") description = XML.parse_html(description) - # Time matches format "20 hours ago", "40 minutes ago"... - delta = published.split(" ")[0].to_i - 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 + published = published.content.split(" ")[-3..-1].join(" ") + published = decode_date(published) json.object do json.field "title", title diff --git a/src/invidious/helpers.cr b/src/invidious/helpers.cr index fb7c573b2..793bf0770 100644 --- a/src/invidious/helpers.cr +++ b/src/invidious/helpers.cr @@ -704,46 +704,22 @@ def fetch_channel(ucid, client, db, pull_all_videos = true) end document = XML.parse_html(content_html) - document.xpath_nodes(%q(//li[contains(@class, "channels-content-item")])).each do |item| - root = item.xpath_node(%q(div/div/div[@class="yt-lockup-content"])) - if !root - raise "could not find root" - end - - anchor = root.xpath_node(%q(h3[contains(@class,"yt-lockup-title")]/a)) + document.xpath_nodes(%q(//li[contains(@class, "feed-item-container")])).each do |item| + anchor = item.xpath_node(%q(.//h3[contains(@class,"yt-lockup-title")]/a)) if !anchor raise "could not find anchor" end + title = anchor.content.strip 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 # This happens on Youtube red videos, here we just skip them next end - published = published.content.split(" ") - span = published[0].to_i - 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 + published = published.content + published = decode_date(published) videos << ChannelVideo.new(video_id, title, published, Time.now, ucid, author) end @@ -863,6 +839,29 @@ def decode_time(string) return time 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) ucid = ucid.lchop("UC") ucid = "VLUU" + ucid @@ -899,7 +898,7 @@ end def produce_videos_url(ucid, 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 meta += "\xb8\x01\x00"