hnruby 0.01.2 → 0.02

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 4b6d68737114cfb703d4d7007b6719583bc99331
-  data.tar.gz: 4f93b627eb99063c28fc9a0e870e87d399663e7d
+  metadata.gz: f1bb2f5a4c8b8fa3b1b78dc9438863f0506be8a3
+  data.tar.gz: dc9ab3e324b16719efd2f5e77734606d40ee76ad
 SHA512:
-  metadata.gz: f2404fb955d07e5aa329766ba34d4561e5e04d20aa49930d759a787152e05dfb62401157bd7b15ea3438076fc04c78bf2999f6cf38453f4826b1c0be61ce40d5
-  data.tar.gz: 3ffe6c85c9e4df51dabf563a270a8c863f26b8238d0ec79179ba5b9e9bda930a032c4843dc912fd265c6da67d4226a4d1d5bf7c09e88f3ebbafab6d1910c81ad
+  metadata.gz: 14030d713a3aeb6a065d6515913ae299f70f751ad1ea8d48ce841c4a0bc59fb01203c880c381b41cbd84b667b2624a8ec5d40659f0d05266bd8f2e0c4fa8220e
+  data.tar.gz: 7904c487d791b218a9306f8405e1587779d5409c885fd53a6eb4a9025c4460bdb0cb38520e9ba496b1ba6de610df98c950844b43eaa02ddfa0c1d3b287537387

data/lib/hnruby.rb

@@ -43,7 +43,7 @@ require 'hnruby/storylist.rb'
 # From a CommentPage, specific comments can be accessed via index:
 #  comment = cpage[1]	#=> <Comment> by user 1 day ago
 module HackerNews
-#Because clearly this is in danger of changing at any moment.
-HN_URL = "https://news.ycombinator.com/"
+	#Because clearly this is in danger of changing at any moment.
+	HN_URL = "https://news.ycombinator.com/"
 
 end

data/lib/hnruby/comment.rb

@@ -82,12 +82,16 @@ class Comment
 end
 
 # Represents the list of comments pertaining to a particluar HN story.
+#---
+#TODO: Support self-post content, polls, etc.
 class CommentPage
 	include Enumerable
 	# Returns this comment page's title.
 	attr_reader :title
 	# Returns the URL for this comment page on Hacker News.
 	attr_reader :url
+	# Returns the URL for the story corresponding to this comment page.
+	attr_reader :story_url
 
 	# Returns a new CommentPage corresponding to <tt>url</tt>.
 	#  cpage = HackerNews::CommentPage.new\
@@ -97,13 +101,14 @@ class CommentPage
 	# the story whose ID is +url+.
 	#  cpage = HackerNews::CommentPage.new 6621679	#=> "C--" <12 Comments>
 	def initialize(url)
-		url = "#{HN_URL}/item?id=#{url.to_i}" if url.is_a? Numeric
+		@url = url.is_a?(Numeric) ? "#{HN_URL}item?id=#{url.to_i}" : url
 
-		html = Nokogiri::HTML(open(url), nil, "UTF-8")
+		html = Nokogiri::HTML(open(@url), nil, "UTF-8")
 		@title = html.title.chomp " | Hacker News"
-		@url = url
+		@story_url = ((html / "tr")[4] / "td")[1].at("a").attribute("href").value
 
-		html = (html / "tr")[8..-3]
+		html = (html / "tr")[(@story_url.start_with?("item?id=") ? 10 : 8)..-3].
+				drop_while { |i| i.children.length != 1 } # Fixes polls and such
 		@comments = []
 		lineage = []
 		(0...html.length).select{ |i| i.even? }.each do |i|
@@ -120,11 +125,6 @@ class CommentPage
 		self
 	end
 
-	# Returns array containing every Comment object within the page.
-	def to_a
-		@comments
-	end
-
 	# Returns the Comment object at <tt>index</tt>.
 	def [](index)
 		@comments[index]
@@ -136,6 +136,11 @@ class CommentPage
 		@comments[index]
 	end
 
+	# Returns the number of comments in +self+.
+	def length
+		@comments.length
+	end
+
 	# :nodoc:
 	def inspect
 		"\"#{@title}\" <#{@comments.length} comments>"

data/lib/hnruby/newsitem.rb

@@ -38,38 +38,32 @@ class NewsItem
 
 		@number = article.first.content.to_i
 		@url = article[2].at("a").attributes["href"].value
-		@url = HN_URL + @url if url.start_with? "item?id="
+		@url.prepend(HN_URL) if url.start_with? "item?id="
 		@title = article[2].children.first.content
 
 		if meta.children.length == 5
 			#This is a normal article
-			@category = :article
-
-			@points = meta[1].children.first.content.to_i
-			@submitter = meta[1].children[2].content
-			@time = meta[1].children[3].content.delete("|").strip
-			@comment_url = meta[1].children[4].attributes["href"].value
-			@comment_page = nil #Don't needlessly download until asked
-			@comments = meta[1].children[4].content.to_i
+			@points       = meta[1].children.first.content.to_i
+			@submitter    = meta[1].children[2].content
+			@time         = meta[1].children[3].content.delete("|").strip
+			@comment_url  = meta[1].children[4].attributes["href"].value
+			@comments     = meta[1].children[4].content.to_i
 		else
 			#This, by contrast, is a job offer, which lacks all meta-info aside
 			#from submission time
-			@category = :joboffer
-
 			@time = meta[1].content
-			@points, @submitter, @comment_url, @comment_page, @comments = nil
 		end
 	end
 	# :startdoc:
 
 	# Returns true if <tt>self</tt> is an article (with points, comments, etc.).
 	def article?
-		@category == :article
+		!@points.nil?
 	end
 	# Returns true if <tt>self</tt> is a job offer (without points, comments,
 	# etc.).
 	def joboffer?
-		@category == :joboffer
+		@points.nil?
 	end
 
 	# Returns the URL to the HN profile of the submitter.
@@ -85,12 +79,13 @@ class NewsItem
 	def comment_url
 		"#{HN_URL}#{@comment_url}" if @comment_url
 	end
-	# Returns the CommentPage corresponding to <tt>self</tt>
+	# Returns the CommentPage corresponding to <tt>self</tt>.
+	#
+	# If +self+ is a job offer, returns +nil+.
 	#  news_item.comment_page\
 	#  #=> "Amazon and the "profitless business model" fallacy" <21 comments>
 	def comment_page
-		return @comment_page if @comment_page
-		@comment_page = CommentPage.new comment_url if @comment_url
+		@comment_page ||= CommentPage.new comment_url if @comment_url
 	end
 
 	# :nodoc:

data/lib/hnruby/storylist.rb

@@ -16,13 +16,20 @@ module HackerNews
 #  story == HackerNews::StoryList[1]	#=> false
 module StoryList
 	@@items = {}
+	@@page_urls = {} #sketchiness ensues
 
 private
+	def self.get_page_url(index)
+		update_page(index - 1) unless @@page_urls[index]
+		HN_URL + @@page_urls[index]
+	end
+
 	def self.update_page(index)
-		url = HN_URL + "#{"news#{index}" if index > 1}"
-		html = (Nokogiri::HTML(open(url), nil, "UTF-8") / "tr").to_a[4..-6]
+		url = (index > 1) ? get_page_url(index) : HN_URL
+		html = Nokogiri::HTML(open(url), nil, "UTF-8") / "tr"
+		@@page_urls[index+1] = html[-3].at("a").attribute("href").content.sub "/",""
 
-		(0...html.length).step(3) do |i|
+		(4...(html.length - 6)).step(3) do |i|
 			item = NewsItem.new(html[i], html[i + 1])
 			@@items[item.number] = item
 		end
@@ -33,6 +40,7 @@ public
 	#updated, though.
 	def self.update
 		@@items.clear
+		@@page_urls.clear
 	end
 
 	# If <tt>index</tt> is a number, returns a NewsItem representing the story
@@ -48,10 +56,11 @@ public
 	#  #=> ["C--" <article: 6621679>, "Profitless Prosperity" <article: 6622394>]
 	# :arg: index
 	def self.[](i)
-		if i.is_a? Numeric
-			self.at i.to_i
-		elsif i.is_a? Enumerable
-			self.stories i
+		case i
+		when Numeric
+			at(i.to_i)
+		when Enumerable
+			stories(i)
 		else
 			raise TypeError, "#{i.class} objects cannot be used to access stories"
 		end
@@ -81,6 +90,23 @@ public
 	def self.stories(range)
 		range.map{ |i| at i }
 	end
+
+	# Returns (up to) the first +count+ NewsItems that return +true+ in the
+	# accompanying block--- if no argument is given, returns the first such
+	# NewsItem.
+	#
+	# May return fewer than +count+ items, if a smaller number than are requested
+	# can be found.
+	def self.get_by(count = 1)
+		i = 0
+		items = []
+		until items.count == count
+			i += 1
+			items << at(i) if yield at(i)
+		end
+	rescue OpenURI::HTTPError #NOTE: May be unnecessary since page-loading's fixed
+		items
+	end
 end
 
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hnruby
 version: !ruby/object:Gem::Version
-  version: 0.01.2
+  version: '0.02'
 platform: ruby
 authors:
 - Jem Orgun
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-11-02 00:00:00.000000000 Z
+date: 2013-11-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ruby-terminfo