word_count_analyzer 0.0.14 → 1.0.0

data/spec/word_count_analyzer/hyperlink_spec.rb

@@ -1,77 +1,77 @@
 require 'spec_helper'
 
 RSpec.describe WordCountAnalyzer::Hyperlink do
-  context '#hyperlink?' do
+  context '#hyperlink?(string)' do
     it 'returns true if the string is a hyperlink #001' do
       string = "http://www.example.com/this-IS-a_test/hello.html"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(true)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(true)
     end
 
     it 'returns true if the string is a hyperlink #002' do
       string = "http://www.google.co.uk"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(true)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(true)
     end
 
     it 'returns true if the string is a hyperlink #003' do
       string = "https://google.co.uk"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(true)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(true)
     end
 
     it 'returns false if the string is not a hyperlink #004' do
       string = "hello"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(false)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(false)
     end
 
     it 'returns false if the string is not a hyperlink #005' do
       string = "john@gmail.com"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(false)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(false)
     end
 
     it 'returns false if the string is not a hyperlink #006' do
       string = "date:"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(false)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(false)
     end
 
     it 'returns false if the string is not a hyperlink #007' do
       string = 'The file location is c:\Users\johndoe.'
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.hyperlink?).to eq(false)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.hyperlink?(string)).to eq(false)
     end
   end
 
   context '#occurences' do
     it 'returns the occurences of hyperlink tokens in a string #001' do
       string = "Today the date is: Jan 1. Visit https://www.example.com/hello or http://www.google.co.uk"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.occurences).to eq(2)
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.occurences(string)).to eq(2)
     end
   end
 
   context '#replace' do
     it 'replaces the hyperlinks in a string with regular tokens #001' do
       string = "Today the date is: Jan 1. Visit https://www.example.com/hello or http://www.google.co.uk"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.replace).to eq("Today the date is: Jan 1. Visit  wslinkword  or  wslinkword ")
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.replace(string)).to eq("Today the date is: Jan 1. Visit  wslinkword  or  wslinkword ")
     end
 
     it 'replaces the hyperlinks in a string with regular tokens #002' do
       string = 'The file location is c:\Users\johndoe or d:\Users\john\www'
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.replace).to eq('The file location is c:\Users\johndoe or d:\Users\john\www')
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.replace(string)).to eq('The file location is c:\Users\johndoe or d:\Users\john\www')
     end
   end
 
   context '#replace_split_at_period' do
     it 'replaces the hyperlinks in a string with regular tokens, split at periods #001' do
       string = "http://www.google.co.uk"
-      ws = WordCountAnalyzer::Hyperlink.new(string: string)
-      expect(ws.replace_split_at_period).to eq("http://www google co uk")
+      ws = WordCountAnalyzer::Hyperlink.new
+      expect(ws.replace_split_at_period(string)).to eq("http://www google co uk")
     end
   end
 end

data/spec/word_count_analyzer/performance_spec.rb

@@ -0,0 +1,46 @@
+# -*- encoding : utf-8 -*-
+require 'benchmark'
+require 'spec_helper'
+require 'stackprof'
+
+RSpec.describe WordCountAnalyzer::Analyzer do
+  it 'is fast?' do
+    benchmark do
+      text = "This string has a date: Monday, November 3rd, 2011. I was thinking... it also shouldn't have too many contractions, maybe 2. <html> Some HTML and a hyphenated-word</html>. Don't count punctuation ? ? ? Please visit the ____________ ------------ ........ go-to site: https://www.example-site.com today. Let's add a list 1. item a 2. item b 3. item c. Now let's add he/she/it or a c:\\Users\\john. 2/15/2012 is the date! { HYPERLINK 'http://www.hello.com' }"
+      ws = WordCountAnalyzer::Analyzer.new(text: text).analyze
+    end
+  end
+
+  it 'is analyzed' do
+    data = StackProf.run(mode: :cpu, interval: 1000) do
+      text = "This string has a date: Monday, November 3rd, 2011. I was thinking... it also shouldn't have too many contractions, maybe 2. <html> Some HTML and a hyphenated-word</html>. Don't count punctuation ? ? ? Please visit the ____________ ------------ ........ go-to site: https://www.example-site.com today. Let's add a list 1. item a 2. item b 3. item c. Now let's add he/she/it or a c:\\Users\\john. 2/15/2012 is the date! { HYPERLINK 'http://www.hello.com' }"
+      ws = WordCountAnalyzer::Analyzer.new(text: text).analyze
+    end
+    puts StackProf::Report.new(data).print_text
+  end
+
+  it 'is analyzed 2' do
+    data = StackProf.run(mode: :cpu, interval: 1000) do
+      token = "when'd"
+      following_token = nil
+      WordCountAnalyzer::Contraction.new(token: token, following_token: following_token, tgr: EngTagger.new, hyphen: nil).contraction?
+    end
+    puts StackProf::Report.new(data).print_text
+  end
+
+  it 'is analyzed 3' do
+    benchmark do
+      text = "This string has a date: Monday, November 3rd, 2011. I was thinking... it also shouldn't have too many contractions, maybe 2. <html> Some HTML and a hyphenated-word</html>. Don't count punctuation ? ? ? Please visit the ____________ ------------ ........ go-to site: https://www.example-site.com today. Let's add a list 1. item a 2. item b 3. item c. Now let's add he/she/it or a c:\\Users\\john. 2/15/2012 is the date! { HYPERLINK 'http://www.hello.com' }"
+      ws = WordCountAnalyzer::Counter.new(forward_slash: 'count_as_multiple')
+      300.times do
+        ws.count(text)
+      end
+    end
+  end
+end
+
+def benchmark
+  yield
+  time = Benchmark.realtime { yield }
+  puts "RUNTIME: #{time}"
+end

data/word_count_analyzer.gemspec

@@ -22,5 +22,6 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "bundler"
   spec.add_development_dependency "rake", "~> 10.0"
   spec.add_development_dependency "rspec"
+  spec.add_development_dependency "stackprof"
   spec.add_runtime_dependency "engtagger"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: word_count_analyzer
 version: !ruby/object:Gem::Version
-  version: 0.0.14
+  version: 1.0.0
 platform: ruby
 authors:
 - Kevin S. Dias
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-04-14 00:00:00.000000000 Z
+date: 2016-03-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -52,6 +52,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: stackprof
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: engtagger
   requirement: !ruby/object:Gem::Requirement
@@ -108,6 +122,7 @@ files:
 - spec/word_count_analyzer/hyphenated_word_spec.rb
 - spec/word_count_analyzer/number_spec.rb
 - spec/word_count_analyzer/numbered_list_spec.rb
+- spec/word_count_analyzer/performance_spec.rb
 - spec/word_count_analyzer/punctuation_spec.rb
 - spec/word_count_analyzer/slash_spec.rb
 - spec/word_count_analyzer/xhtml_spec.rb
@@ -148,6 +163,7 @@ test_files:
 - spec/word_count_analyzer/hyphenated_word_spec.rb
 - spec/word_count_analyzer/number_spec.rb
 - spec/word_count_analyzer/numbered_list_spec.rb
+- spec/word_count_analyzer/performance_spec.rb
 - spec/word_count_analyzer/punctuation_spec.rb
 - spec/word_count_analyzer/slash_spec.rb
 - spec/word_count_analyzer/xhtml_spec.rb