Memory optimization

.rspec

@@ -0,0 +1 @@
+--color

lib/twitter_ebooks.rb

@@ -1,5 +1,3 @@
-gem 'minitest'
-
 $debug = false
 
 def log(*args)

lib/twitter_ebooks/model.rb

@@ -18,14 +18,28 @@ module Ebooks
       Marshal.load(File.open(path, 'rb') { |f| f.read })
     end
 
+    def mass_tokenize(text)
+      sentences = NLP.sentences(text)
+      tokens = []
+
+      sentences.each do |s|
+        tokens << NLP.tokenize(s).reject do |t|
+          # Don't include usernames/urls as tokens
+          t.include?('@') || t.include?('http')
+        end
+      end
+
+      tokens
+    end
+
     def consume(path)
       content = File.read(path, :encoding => 'utf-8')
       @hash = Digest::MD5.hexdigest(content)
 
       if path.split('.')[-1] == "json"
         log "Reading json corpus from #{path}"
-        lines = JSON.parse(content, symbolize_names: true).map do |tweet|
+        lines = JSON.parse(content).map do |tweet|
-          tweet[:text]
+          tweet['text']
         end
       elsif path.split('.')[-1] == "csv"
         log "Reading CSV corpus from #{path}"
@@ -42,44 +56,31 @@ module Ebooks
 
       log "Removing commented lines and sorting mentions"
 
-      keeping = []
+      statements = []
       mentions = []
       lines.each do |l|
         next if l.start_with?('#') # Remove commented lines
         next if l.include?('RT') || l.include?('MT') # Remove soft retweets
 
         if l.include?('@')
-          mentions << l
+          statements << NLP.normalize(l)
         else
-          keeping << l
+          mentions << NLP.normalize(l)
-        end
-      end
-      text = NLP.normalize(keeping.join("\n")) # Normalize weird characters
-      mention_text = NLP.normalize(mentions.join("\n"))
-
-      log "Segmenting text into sentences"
-
-      statements = NLP.sentences(text)
-      mentions = NLP.sentences(mention_text)
-
-      log "Tokenizing #{statements.length} statements and #{mentions.length} mentions"
-      @sentences = []
-      @mentions = []
-
-      statements.each do |s|
-        @sentences << NLP.tokenize(s).reject do |t|
-          t.include?('@') || t.include?('http')
         end
       end
 
-      mentions.each do |s|
+      text = statements.join("\n")
-        @mentions << NLP.tokenize(s).reject do |t|
+      mention_text = mentions.join("\n")
-          t.include?('@') || t.include?('http')
-        end
-      end
 
-      log "Ranking keywords"
+      lines = nil; statements = nil; mentions = nil # Allow garbage collection
-      @keywords = NLP.keywords(@sentences)
+
+      log "Tokenizing #{text.count('\n')} statements and #{mention_text.count('\n')} mentions"
+
+      @sentences = mass_tokenize(text)
+      @mentions = mass_tokenize(mention_text)
+
+      #log "Ranking keywords"
+      #@keywords = NLP.keywords(@sentences)
 
       self
     end

lib/twitter_ebooks/nlp.rb

@@ -43,7 +43,7 @@ module Ebooks
     end
 
     ### Utility functions
-    
+
     # We don't really want to deal with all this weird unicode punctuation
     def self.normalize(text)
       htmlentities.decode text.gsub('“', '"').gsub('”', '"').gsub('’', "'").gsub('…', '...')

spec/model_spec.rb

@@ -1,8 +1,14 @@
 require 'spec_helper'
-require 'objspace'
+require 'memory_profiler'
 
 describe Ebooks::Model do
-  it "does stuff" do
+  it "does not use a ridiculous amount of memory" do
-    model = Ebooks::Model.load(path("data/0xabad1dea.model"))
+    RubyProf.start
+    Ebooks::Model.consume(path("data/0xabad1dea.json"))
+    result = RubyProf.stop
+
+    require 'pry'; binding.pry
+
+    expect(report.total_retained).to be < 100000
   end
 end

twitter_ebooks.gemspec

@@ -16,11 +16,12 @@ Gem::Specification.new do |gem|
   gem.version       = Ebooks::VERSION
 
   gem.add_development_dependency 'rspec'
-  gem.add_development_dependency 'memory_profiler'
+  gem.add_development_dependency 'ruby-prof'
+  gem.add_development_dependency 'pry-byebug'
 
 
-  gem.add_runtime_dependency 'twitter', '~> 4.5'
+  gem.add_runtime_dependency 'twitter', '~> 5.1'
-  gem.add_runtime_dependency 'tweetstream', '= 2.5'
+  gem.add_runtime_dependency 'tweetstream'
   gem.add_runtime_dependency 'rufus-scheduler'
   gem.add_runtime_dependency 'gingerice'
   gem.add_runtime_dependency 'htmlentities'