okay it actually works this time I promise

schedules_api/db/patriot_web_parser.rb

@@ -36,7 +36,7 @@ module PatriotWeb
     def parse_courses_in_subject(subject)
       response = @networker.fetch_courses_in_subject(subject)
       document = Nokogiri::HTML(response)
-      get_courses(document)
+      get_courses(document, subject)
     end
 
     private
@@ -66,120 +66,74 @@ module PatriotWeb
     # Parse all courses from the subject search page 
     # @param document [Nokogiri::HTML::Document]
     # @return [Array] courses
-    def get_courses(document)
+    def get_courses(document, subject)
       table = document.css('html body div.pagebodydiv table.datadisplaytable')
       rows = table.css('tr')
+      # rows[100..110].each_with_index do |row, i|
+      #   puts i
+      #   puts row
+      # end
+      data_from rows
+    end
+
+    def data_from(rows)
+      i = 0
+      title_index = 0
+      result = []
+      
+      while i < rows.length
+        if is_title(rows[i].text) # check if the row is a title
+          data = {}
 
-      (0..(rows.length/6-1)).map do |i|
-        start = i*5
-        data = {}
-        title = rows[start].text
-        # the title looks this: Survey of Accounting - 71117 - ACCT 203 - 001
-        # so split it by ' - ' and extract
-        title_elements = title.split(' - ')
-        next unless title_elements.length == 4
-        data[:title] = title_elements[0].strip
-        data[:crn] = title_elements[1]
-        
-        full_name = title_elements[2].split(' ')
-        next unless full_name.length == 2
-        data[:subj] = title_elements[2].split(' ')[0]
-        data[:course_number] = title_elements[2].split(' ')[1]
-        
-        data[:section] = title_elements[3].strip
+          title_elements = rows[i].text.split(' - ')
+          data[:title] = title_elements[0].strip
+          data[:crn] = title_elements[1]
+          full_name = title_elements[2].split(' ')
+          next unless full_name.length == 2
+          data[:subj] = full_name[0]
+          data[:course_number] = full_name[1]
+          data[:section] = title_elements[3].strip
 
-        # rows 1 to 3 contain info about registration and drop dates.
-        # for now we're gonna ignore them and skip to row 4, which contains details
-        details = rows[start+2].css('td table tr td')
+          details = rows[i+2].css('td table tr td')
+          unless details.length > 0
+            puts "#{full_name.join(' ')} is fake news"
+            i += 1
+            next
+          end
+          
+          times = details[1].text.split(' - ')
+          if (times.length == 1)
+            data[:start_time] = 'TBA'
+            data[:end_time] = 'TBA'
+          else
+            data[:start_time] = times[0]
+            data[:end_time] = times[1]
+          end
 
-        next unless details.length > 0 # if there are no details, skip this item
-        # details = detail_rows.last.text.split("\n").compact.reject(&:empty?) # skip empty strings
-        
-        times = details[1].text.split(' - ')
-        if (times.length == 1)
-          data[:start_time] = 'TBA'
-          data[:end_time] = 'TBA'
+          data[:days] = details[2].text.strip
+          data[:location] = details[3].text.strip
+          
+          dates = details[4].text.split(' - ')
+          data[:start_date] = dates[0]
+          data[:end_date] = dates[1]
+          
+          data[:type] = details[5].text
+          data[:instructor] = details[6].text
+
+          result << data
+          i += 5 # skip to what we think is the next title
         else
-          data[:start_time] = times[0]
-          data[:end_time] = times[1]
+          i += 1 # try the next row if this one was not a title
         end
-
-        data[:days] = details[2].text.strip
-        data[:location] = details[3].text.strip
-        
-        dates = details[4].text.split(' - ')
-        data[:start_date] = dates[0]
-        data[:end_date] = dates[1]
-        
-        data[:type] = details[5].text
-        data[:instructor] = details[6].text
-        data
       end
-      # puts rows[0].text
-      # puts rows[2].css('td table tr td')
-      # puts rows[5].text
-      # puts rows[7].css('td table tr td')
       
-      # puts rows[10].text
-      # puts rows[12].css('td table tr td')
-      # (0..(rows.length/3-1)).each do |i|
-      #   start = i*3
-        
-      #   puts rows[start].search('th').first.text
-      #   section_data = rows[start+2].css('td table.datadisplaytable').search('td')
-      #   puts section_data[1].text
-      # end
-      # puts rows[3]
-      # puts rows[3].search('th').first.text
-      # section_data = rows[5].css('td table.datadisplaytable').search('td')
-      # puts section_data[1].text
-
-      # end
-      # each section is represented by 6 rows in the table
-      # (0..(rows.length/6 - 1)).map do |i|
-      #   start = i*6
-      #   data = {}
-      #   title = rows[start].text
-      #   # the title looks this: Survey of Accounting - 71117 - ACCT 203 - 001
-      #   # so split it by ' - ' and extract
-      #   title_elements = title.split(' - ')
-      #   next unless title_elements.length == 4
-      #   data[:title] = title_elements[0].strip
-      #   data[:crn] = title_elements[1]
-        
-      #   full_name = title_elements[2].split(' ')
-      #   next unless full_name.length == 2
-      #   data[:subj] = title_elements[2].split(' ')[0]
-      #   data[:course_number] = title_elements[2].split(' ')[1]
-        
-      #   data[:section] = title_elements[3].strip
-
-      #   # rows 1 to 3 contain info about registration and drop dates.
-      #   # for now we're gonna ignore them and skip to row 4, which contains details
-      #   detail_rows = rows[start+4].css('tr')
-      #   next unless detail_rows.length > 0 # if there are no details, skip this item
-      #   details = detail_rows.last.text.split("\n").compact.reject(&:empty?) # skip empty strings
-        
-      #   times = details[1].split(' - ')
-      #   if (times.length == 1)
-      #     data[:start_time] = 'TBA'
-      #     data[:end_time] = 'TBA'
-      #   else
-      #     data[:start_time] = times[0]
-      #     data[:end_time] = times[1]
-      #   end
-
-      #   data[:days] = details[2].strip
-      #   data[:location] = details[3].strip
-        
-      #   dates = details[4].split(' - ')
-      #   data[:start_date] = dates[0]
-      #   data[:end_date] = dates[1]
-        
-      #   data[:type] = details[5]
-      #   data[:instructor] = details[6]
-      #   data
-      # end
+      result
+    end
+    
+    # a title looks this: Survey of Accounting - 71117 - ACCT 203 - 001
+    def is_title(text)
+      elements = text.split(' - ')
+      elements.length == 4 && elements[2].split(' ').length == 2
     end
   end
 end

schedules_api/db/seeds.rb

@@ -19,17 +19,17 @@ puts "DDOSing Patriot Web, buckle up kids"
 # parse all subjects and their courses in the semester
 parser.parse_subjects(semester).each do |subject|
   puts "Getting courses for #{subject}"
-  threads << Thread.new {
+  # threads << Thread.new {
     total[subject] = parser.parse_courses_in_subject(subject)    
-  }
+  # }
 end
 
 # For testing, only get first subject
-# subject = parser.parse_subjects(semester).first
+# subject = parser.parse_subjects(semester)[20]
 # total[subject] = parser.parse_courses_in_subject(subject)
 
 # wait for all the threads to finish
-ThreadsWait.all_waits(*threads)
+# ThreadsWait.all_waits(*threads)
 
 # delete everything in the current database
 Closure.delete_all
@@ -44,7 +44,10 @@ semester.save!
 total.each do |subject, sections|
   puts "Adding courses for #{subject}..."
   sections.each do |section|
-    next if section.nil? || !section.key?(:subj) || !section.key?(:course_number)
+    if section.nil? || !section.key?(:subj) || !section.key?(:course_number)
+      puts "#{subject} failed section: #{section.class}"
+      next
+    end
     
     # Find or create a course and set its semester
     # TODO: this breaks when you try to do more than one semester,