whatsa 0.2.3 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: aef97bd6c6bbcfbff00e5fe88ab1e6715de035cb
-  data.tar.gz: 4816a33391219a2d92081be925269db130733096
+  metadata.gz: b7cf846d19a9dba5468c21b861abeac0f2eddb90
+  data.tar.gz: ba374012eb9a91e66db755986d55716b4d8bf2b4
 SHA512:
-  metadata.gz: 9661681f7cde6a9b5b17807a2a908e18c0eab23a0b053a3a4c4f7cecfcd77ef2401bb3341af3840cc70b6ebf831f7ce5bbc89709b319cbca4514f91cc2e629f5
-  data.tar.gz: 87d9ba1a98bb9347d8aef6d369009f0303c096d3d4b31bcfcb2387c56c0f77021aad0f61ff3aaa1e89a5c0d20ae1b75da9d5e23f3f43694868e6f489cd9b6a2e
+  metadata.gz: cc3ddde58c65c630c7edf1e0bdfba09e1ad3072d2f81bb212f3c04d0dcecf85d49905a0565d515c03d135de82dfec7270bd816f9a17ae1d6aba782032e2e2bb5
+  data.tar.gz: 3ea318931845cf4e3787c1428cc991a253326080deba40d2b03c869dec56c62131930f12d4426d76b2613ad5e8dffb95119d964bbb73cb19abe2327f56461c48

data/config/environment.rb

@@ -3,6 +3,7 @@ require 'nokogiri'
 require 'pry'
 
 require_relative '../lib/whatsa/version'
+require_relative '../lib/whatsa/format'
 require_relative '../lib/whatsa/cli'
 require_relative '../lib/whatsa/article'
 require_relative '../lib/whatsa/section'

data/lib/whatsa/article.rb

@@ -1,4 +1,10 @@
 class Whatsa::Article
+  # I think this is a bad use of include. I feel like I _should_ make Format a
+  # class and use its methods that way, but I'm going to be using them so often
+  # that I would prefer they not be referenced as Whatsa::Format.blah_blah every
+  # time...
+  include Whatsa::Format
+
   attr_accessor :sections
   attr_reader :contents, :title
 
@@ -53,7 +59,7 @@ class Whatsa::Article
     indices << -1
     secs = [Whatsa::Section.new("#{self.title} - Introduction", intro_pars)]
     indices.each_cons(2) do |i, j|
-      title = self.contents[i].text.gsub('[edit]', '').strip
+      title = heading_to_title(self.contents[i].text)
       par_nodes = self.contents[i...j].select do |e|
         e.name == 'p' || e.name == 'ul' && e.text != ""
       end

data/lib/whatsa/cli.rb

@@ -1,4 +1,9 @@
 class Whatsa::CLI
+  # I think this is a bad use of include. I feel like I _should_ make Format a
+  # class and use its methods that way, but I'm going to be using them so often
+  # that I would prefer they not be referenced as Whatsa::Format.blah_blah every
+  # time...
+  include Whatsa::Format
 
   def clear_screen
     50.times { puts "\n" }
@@ -44,22 +49,6 @@ class Whatsa::CLI
     input
   end
 
-  # setting an indent will indent the lines AFTER the first line of a paragraph
-  def word_wrap(text, indent=0)
-    count = 0
-    words = text.split(/ /)
-    words.each_with_index do |word, index|
-      count += word.length + 1
-      if count > 80
-        words.insert(index, "\n#{' ' * indent}")
-        count = indent
-      elsif word.index("\n")
-        count = word.length
-      end
-    end
-    words.join(" ").gsub(/^ /, "")
-  end
-
   def display_dmb(dmb)
     raise TypeError unless dmb.is_a?(Whatsa::Disambig)
     clear_screen

data/lib/whatsa/format.rb

@@ -0,0 +1,46 @@
+module Whatsa
+  module Format
+
+    def url_friendly(string)
+      string.gsub(/[^A-z0-9\(\)]+/, '+').gsub(/(\A\+|\+\z)/, '')
+    end
+
+    def heading_to_title(string)
+      string.gsub('[edit]', '').strip
+    end
+
+    def remove_citation_markers(string)
+      string.gsub(/\[(\d+|citation needed)\]/, "")
+    end
+
+    def bulletize_lines(string)
+      if string.index("\n")
+        list = string.gsub(/^/, "- ")
+        word_wrap(list, "- ".length)
+      else
+        string
+      end
+    end
+
+    # setting an indent will indent the lines AFTER the first line of a paragraph
+    def word_wrap(text, indent=0)
+      chars = text.split(//)
+      unless text.length < 80
+        count = 1
+        last_space = 80
+        chars.each_with_index do |char, index|
+          count += 1
+          last_space = index if char.match(/ /)
+          if char == "\n"
+            count = indent
+          elsif count == 80
+            chars[last_space] = "\n#{" " * indent}"
+            count = indent + index - last_space
+          end
+        end
+      end
+      chars.join
+    end
+
+  end
+end

data/lib/whatsa/scraper.rb

@@ -1,4 +1,9 @@
 class Whatsa::Scraper
+  # I think this is a bad use of include. I feel like I _should_ make Format a
+  # class and use its methods that way, but I'm going to be using them so often
+  # that I would prefer they not be referenced as Whatsa::Format.blah_blah every
+  # time...
+  include Whatsa::Format
 
   WIKISEARCH = 'https://en.wikipedia.org/w/index.php?search='
 
@@ -7,8 +12,7 @@ class Whatsa::Scraper
   def initialize(term)
     # only keep word chars and parens, turn everything between each 'word'
     # to a single '+' and remove '+'s at the beginning and end if they're there
-    # @query = term.gsub(/\W+/, '+').gsub(/(\A\+|\+\z)/, '')
-    @query = term.gsub(/[^A-z0-9\(\)]+/, '+').gsub(/(\A\+|\+\z)/, '')
+    @query = url_friendly(term)
 
     # store the page in an instance variable so we don't keep polling the site
     @page = Nokogiri::HTML(open(WIKISEARCH + self.query))
@@ -46,5 +50,4 @@ class Whatsa::Scraper
   def make_disambig
     disambig? ? Whatsa::Disambig.new(self.page) : nil
   end
-
 end

data/lib/whatsa/section.rb

@@ -1,4 +1,10 @@
 class Whatsa::Section
+  # I think this is a bad use of include. I feel like I _should_ make Format a
+  # class and use its methods that way, but I'm going to be using them so often
+  # that I would prefer they not be referenced as Whatsa::Format.blah_blah every
+  # time...
+  include Whatsa::Format
+  
   attr_accessor :title, :paragraphs, :article
 
   def initialize(title, paragraphs)
@@ -19,11 +25,11 @@ class Whatsa::Section
   private
 
   def remove_citations
-    self.paragraphs.map! { |par| par.gsub(/\[(\d+|citation needed)\]/, "") }
+    self.paragraphs.map! { |par| remove_citation_markers(par) }
   end
 
   def bullet_list_pars
-    self.paragraphs.map! { |par| par.index("\n") ? par.gsub(/^/, "- ") : par }
+    self.paragraphs.map! { |par| bulletize_lines(par) }
   end
 
 end

data/lib/whatsa/version.rb

@@ -1,3 +1,3 @@
 module Whatsa
-  VERSION = "0.2.3"
+  VERSION = "0.3.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: whatsa
 version: !ruby/object:Gem::Version
-  version: 0.2.3
+  version: 0.3.0
 platform: ruby
 authors:
 - Keegan Leitz
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-11-30 00:00:00.000000000 Z
+date: 2016-12-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -90,6 +90,7 @@ files:
 - lib/whatsa/article.rb
 - lib/whatsa/cli.rb
 - lib/whatsa/disambig.rb
+- lib/whatsa/format.rb
 - lib/whatsa/scraper.rb
 - lib/whatsa/section.rb
 - lib/whatsa/version.rb