InExpress360.com crawler

inexpress360-detail.rb

# encoding: UTF-8

require "nokogiri"
require "open-uri"
require "terminal-table"
require "csv"

table_headings = ["STT", "Mã BC", "Tên Bưu cục", "BC cấp", "Địa chỉ", "Điện thoại"]
province = ARGV[0]
document = Nokogiri::HTML(open("https://inxpress360.com/ma-buu-dien-#{province}/"))
table_rows = []

document.css("article table.tve_table tbody tr").each do |tr_elm|
  table_rows << tr_elm.css("td").map{|td_elm| td_elm.content}
end

File.open("#{province}.csv", "w:UTF-8") do |f|
  csv_content = CSV.generate do |csv|
    csv << table_headings
    table_rows.each do |row|
      csv << row
    end
  end

  f.write(csv_content)
end

terminal_table = Terminal::Table.new headings: table_headings, rows: table_rows
puts "#{terminal_table}\n"

inexpress360-list.rb

require "nokogiri"
require "open-uri"
require "terminal-table"

table_headings = %w(No. Province Code)
document = Nokogiri::HTML(open("https://inxpress360.com/ma-buu-dien/"))
table_rows = []

document.css("article .thrv_wrapper table.tve_table tbody tr").each do |tr_elm|
  table_rows << tr_elm.css("td").map{|td_elm| td_elm.content}
end

terminal_table = Terminal::Table.new headings: table_headings, rows: table_rows
puts "#{terminal_table.to_s}\n"