adjusted analyzer to use new facebook data format

lib/facebook_data_analyzer/analyzeables/contacts.rb

@@ -5,8 +5,8 @@ class Contacts < Analyzeable
     attr_reader :contacts
     def initialize(catalog:)
       @catalog = catalog
-      @directory = "#{catalog}/html/"
-      @file_pattern = 'contact_info.htm'
+      @directory = "#{catalog}/about_you/"
+      @file_pattern = 'your_address_books.html'
       @contacts = []
 
       super()
@@ -17,14 +17,10 @@ def analyze
         content = File.open(@file_pattern).read
         doc = Nokogiri::HTML(content)
 
-        contacts_rows = doc.css('div.contents tr')
+        contacts_rows = doc.at_css('._4t5n').children
 
         unique_contacts = contacts_rows.each_with_object({}) do |contact, seen_contacts|
-          text = contact.text
-
-          next if text == 'NameContacts'
-
-          seen_contacts[text] = Contact.parse(contact_text: text)
+          seen_contacts[contact.children[0].text] = Contact.parse(contact_text: contact)
         end
 
         unique_contacts.values.each do |contact|

lib/facebook_data_analyzer/analyzeables/friends.rb

@@ -12,8 +12,8 @@ class Friends < Analyzeable
 
     def initialize(catalog:)
       @catalog = catalog
-      @directory = "#{catalog}/html/"
-      @file_pattern = 'friends.htm'
+      @directory = "#{catalog}/friends/"
+      @file_pattern = 'friends.html'
       @friends = []
 
       super()
@@ -23,7 +23,7 @@ def analyze
       Dir.chdir(@directory) do
         content = File.open(@file_pattern).read
         doc = Nokogiri::HTML(content)
-        friends_list = doc.css('div.contents > ul')[0].css('li')
+        friends_list = doc.at_css('._4t5n').children
 
         friends_list.each do |friend_element|
           friend = Friend.parse(friend_element: friend_element)

lib/facebook_data_analyzer/analyzeables/messages.rb

@@ -26,21 +26,23 @@ class Messages < Analyzeable
     def initialize(catalog:, options: {})
       @verbose = options.fetch(:verbose) { false }
       @catalog = catalog
-      @directory = "#{catalog}/messages"
-      @file_pattern = '*.html'
+      @directory = "#{catalog}/messages/inbox"
+      @file_pattern = '**/*.html'
       @messages = []
 
       # super(parallel: options.fetch(:parallel) { false })
       super(parallel: options.fetch(:parallel, false))
     end
 
     def me
-      @me ||= Nokogiri::HTML(File.open("#{@catalog}/index.htm")).title.split(' - Profile')[0].to_sym
+      @me ||= Nokogiri::HTML(File.open("#{@catalog}/profile_information/profile_information.html")).css('table._4nkx td')[0].text.to_sym
     end
 
     def analyze
       Dir.chdir(@directory) do
-        messages_files = Dir.glob(@file_pattern)
+        messages_files = Dir.glob(@file_pattern).map{|f| File.realpath(f)}
+
+        puts me
 
         # This block will be skipped if all message files have already been parsed
         ::Parallel.each(messages_files, in_processes: @processes_supported, progress: 'Parsing Messages') do |file|
@@ -51,13 +53,13 @@ def analyze
               content: message.content }
           end
 
-          File.open("_#{file}.json", 'w') do |json|
+          File.open("#{file}.json", 'w') do |json|
             json.write(conversation_messages.to_json)
           end
-        end unless @verbose || (Dir.glob('_*.json').count == messages_files.count)
+        end unless @verbose || (Dir.glob('**/*.json').count == messages_files.count)
 
         semaphore = Mutex.new
-        parsed_message_files = Dir.glob('_*.json')
+        parsed_message_files = Dir.glob('**/*.json')
         ::Parallel.each(parsed_message_files, in_threads: @threads_supported, progress: 'Analyzing Messages') do |json_file|
           json_message_array = JSON.parse(File.read(json_file))
           messages = json_message_array.map do |message|
@@ -340,21 +342,24 @@ def most_popular_english_words
     def extract_messages(file:)
       content = File.open(file)
       doc = Nokogiri::HTML(content)
-      conversation_name = doc.title.split('Conversation with ')[1]
+      conversation_name = doc.title
+
+      #puts conversation_name
 
       return [] if conversation_name.nil?
 
-      conversation = doc.at_css('.thread').children
+      conversation = doc.at_css('._4t5n').children
       conversation_senders = []
       conversation_contents = []
       messages = []
 
       conversation.each do |node|
-        if node.name == 'div' && node['class'] == 'message'
-          conversation_senders << node
-        elsif node.name == 'p'
-          # There are empty <p> as padding around images
-          conversation_contents << node unless node.children.count == 0
+        if node.name == 'div' && node['class'] == 'pam _3-95 _2pi0 _2lej uiBoxWhite noborder' && node.children.count >= 3
+          conversation_senders << [node.children[0], node.children[2]]
+          conversation_contents << node.children[1]
+        #elsif node.name == 'p'
+        #  # There are empty <p> as padding around images
+        #  conversation_contents << node unless node.children.count == 0
         end
       end
 

lib/facebook_data_analyzer/contact.rb

@@ -3,9 +3,10 @@
 module FacebookDataAnalyzer
   class Contact
     def self.parse(contact_text:)
-      contact_info = contact_text.split('contact: ')
+      name_text = contact_text.children[0].text
+      details = contact_text.children.drop(1)
 
-      Contact.new(name: String(contact_info[0]), details: contact_info[1..3].join(' '))
+      Contact.new(name: String(name_text), details: details.join(' '))
     end
 
     attr_reader :name, :details

lib/facebook_data_analyzer/friend.rb

@@ -5,13 +5,8 @@ class Friend
     attr_reader :name, :date_added
 
     def self.parse(friend_element:)
-      friend_with_email = friend_element.text.match(/(.*)\s\((.*)\)\s\((.*)\)/)
-
-      if friend_with_email
-        name, date_added = friend_with_email.captures
-      else
-        name, date_added = friend_element.text.match(/(.*)\s\((.*)\)/).captures
-      end
+      name = friend_element.children[0].text
+      date_added = friend_element.children.last.text
 
       date = if date_added == 'Today'
                Date.today

lib/facebook_data_analyzer/message.rb

@@ -3,13 +3,8 @@
 module FacebookDataAnalyzer
   class Message
     def self.parse(sender_info:, content:, conversation:)
-      # To avoid searching, making a hash of child.name.child.class
-      hashed_children = {}
-      message_header = sender_info.children[0]
-      message_header.children.each { |c| hashed_children["#{c.name}.#{c['class']}"] = c }
-
-      sender = hashed_children['span.user'].text
-      date_sent = DateTime.parse(hashed_children['span.meta'].text)
+      sender = sender_info[0].text
+      date_sent = DateTime.parse(sender_info[1].text)
       # There are some legit messages that are empty <p>'s for some reason
       raw_content = (content&.text || 'messageremoved').downcase
       # Removes everything that's not alphanumeric (except for spaces and $)