scrape-pool-schedule.rb

#!/usr/bin/env ruby

require 'date'
require 'time'
require 'mechanize'
# do this or it complains
require 'active_support/isolated_execution_state'
# wah wah wah crybaby
require 'icalendar'
require 'icalendar/tzinfo'
require 'icalendar-recurrence'

class Scrapinator

  private

  def scrape_table table
    table.xpath('tbody/tr').map do |tr|
      tr.elements.map do |td|
        text = td.text.gsub(/\u{a0}+/, ' ').strip
        text.empty? ? nil : text
      end
    end
  end

  def convert_arrays tables
    # { "Weekday" => [ [start, end, description, location], ... ] }
    out = {}

    tables.each do |t|
      # skip unless the table has four cells
      next unless t.first and t.first.length == 4

      today = nil

      t.each do |row|
        # a row of all nulls signifies a new day, and subsquent days are
        # in the first column *next* to the first entry in columns 2-4
        next if row.all? &:nil?

        # the first row has the day in the first column and labels for
        # subsequent columns which we can discard

        has_data = true

        unless row.first.nil?
          has_data = false unless today
          today = row.first

          out[today] ||= []
        end

        if has_data
          time, desc, location = row.drop 1
          start, finish = time.split(?-).map(&:strip)
          unless /[ap]m/i.match?(start)
            if m = /([ap]m)/i.match(finish)
              start << m.captures.first
            end
          end

          out[today] << [start, finish, desc, location]
        end
      end
    end

    out
  end

  def make_ical basis, struct
    ical = Icalendar::Calendar.new

    tz = TZInfo::Timezone.get @tzid
    timezone = tz.ical_timezone DateTime.now

    ical.add_timezone timezone

    (0..6).each do |i|
      day = (basis + i).strftime '%A'

      if struct[day]
        struct[day].each do |record|
          start, finish, desc, location = record

          start, finish = [start, finish].map { |t| Time.parse(t, basis + i) }

          ical.event do |e|
            e.dtstart  = start
            e.dtend    = finish
            e.summary  = desc
            e.location = location
            e.rrule    = 'FREQ=WEEKLY;COUNT=12'
          end
        end
      end
    end

    ical
  end

  public

  def initialize tz
    @tzid = tz
    @mech = Mechanize.new
  end

  def run basis, url
    # fetch the webpage
    page = @mech.get url

    # find the stupid tables
    tables = page.search('//table').map { |t| scrape_table t }

    warn tables.inspect

    # turn them into a struct
    struct = convert_arrays tables


    # punt out ical file
    make_ical basis, struct
  end
end

if $0 == __FILE__
  basis = Date.new 2022, 1, 31
  url  = 'https://kpe.utoronto.ca/sport-recreationrecreational-workouts-activitiesdrop-sports-activities/lane-swimming'
  ical = Scrapinator.new('America/Toronto').run basis, url
  ical.publish

  print ical.to_ical
end
	#!/usr/bin/env ruby

	require 'date'
	require 'time'
	require 'mechanize'
	# do this or it complains
	require 'active_support/isolated_execution_state'
	# wah wah wah crybaby
	require 'icalendar'
	require 'icalendar/tzinfo'
	require 'icalendar-recurrence'

	class Scrapinator

	private

	def scrape_table table
	table.xpath('tbody/tr').map do \|tr\|
	tr.elements.map do \|td\|
	text = td.text.gsub(/\u{a0}+/, ' ').strip
	text.empty? ? nil : text
	end
	end
	end

	def convert_arrays tables
	# { "Weekday" => [ [start, end, description, location], ... ] }
	out = {}

	tables.each do \|t\|
	# skip unless the table has four cells
	next unless t.first and t.first.length == 4

	today = nil

	t.each do \|row\|
	# a row of all nulls signifies a new day, and subsquent days are
	# in the first column next to the first entry in columns 2-4
	next if row.all? &:nil?

	# the first row has the day in the first column and labels for
	# subsequent columns which we can discard

	has_data = true

	unless row.first.nil?
	has_data = false unless today
	today = row.first

	out[today] \|\|= []
	end

	if has_data
	time, desc, location = row.drop 1
	start, finish = time.split(?-).map(&:strip)
	unless /[ap]m/i.match?(start)
	if m = /([ap]m)/i.match(finish)
	start << m.captures.first
	end
	end

	out[today] << [start, finish, desc, location]
	end
	end
	end

	out
	end

	def make_ical basis, struct
	ical = Icalendar::Calendar.new

	tz = TZInfo::Timezone.get @tzid
	timezone = tz.ical_timezone DateTime.now

	ical.add_timezone timezone

	(0..6).each do \|i\|
	day = (basis + i).strftime '%A'

	if struct[day]
	struct[day].each do \|record\|
	start, finish, desc, location = record

	start, finish = [start, finish].map { \|t\| Time.parse(t, basis + i) }

	ical.event do \|e\|
	e.dtstart = start
	e.dtend = finish
	e.summary = desc
	e.location = location
	e.rrule = 'FREQ=WEEKLY;COUNT=12'
	end
	end
	end
	end

	ical
	end

	public

	def initialize tz
	@tzid = tz
	@mech = Mechanize.new
	end

	def run basis, url
	# fetch the webpage
	page = @mech.get url

	# find the stupid tables
	tables = page.search('//table').map { \|t\| scrape_table t }

	warn tables.inspect

	# turn them into a struct
	struct = convert_arrays tables


	# punt out ical file
	make_ical basis, struct
	end
	end

	if $0 == __FILE__
	basis = Date.new 2022, 1, 31
	url = 'https://kpe.utoronto.ca/sport-recreationrecreational-workouts-activitiesdrop-sports-activities/lane-swimming'
	ical = Scrapinator.new('America/Toronto').run basis, url
	ical.publish

	print ical.to_ical
	end