Merge branch 'feature/crawl_all_job_using_parallel_mutex' into 'master'

Crawl all job using parallel and mutex

See merge request !20

Gemfile

@@ -29,11 +29,13 @@ gem 'rsolr'
 
 gem 'carrierwave'
 
-gem 'activerecord-import'
+gem 'activerecord-import', require: false
 
 gem 'will_paginate'
 
 gem 'settingslogic'
+
+gem 'parallel', require: false
 # Use Active Storage variant
 # gem 'image_processing', '~> 1.2'
 

Gemfile.lock

@@ -160,6 +160,7 @@ GEM
     nokogiri (1.10.9-x86-mingw32)
       mini_portile2 (~> 2.4.0)
     orm_adapter (0.5.0)
+    parallel (1.19.1)
     pry (0.13.0)
       coderay (~> 1.1)
       method_source (~> 1.0)
@@ -305,6 +306,7 @@ DEPENDENCIES
   listen (>= 3.0.5, < 3.2)
   meta-tags
   mysql2
+  parallel
   pry
   puma (~> 4.1)
   rails (~> 6.0.2, >= 6.0.2.2)

lib/tasks/crawler.rake

 require "nokogiri"
 require "open-uri"
+require "parallel"
+require "activerecord-import"
 
 namespace :crawler do
   desc "Crawler Careerbuilder"
@@ -7,13 +9,15 @@ namespace :crawler do
   task job: :environment do
     # Define crawler logger
     logger = Logger.new("log/crawler_logger.log")
-
+    html_careerbuilder_list_jobs = Nokogiri::HTML.parse(URI.open("https://careerbuilder.vn/viec-lam/tat-ca-viec-lam-vi.html"))
+    total_page = (html_careerbuilder_list_jobs.at_css(".search-result-list .container .job-found .job-found-amout p").text.tr(",việc làm","").to_i / 50.0).ceil
     # Loop page
-    (1..2).each do |page|
+    (1..total_page).each do |page|
       # Fetch and parse HTML document
       html_jobs = Nokogiri::HTML.parse(URI.open("https://careerbuilder.vn/viec-lam/tat-ca-viec-lam-trang-#{page}-vi.html"))
       # Loop item
-      html_jobs.css(".jobs-side-list .job-item").each do |item|
+      Parallel.each(html_jobs.css(".jobs-side-list .job-item"), in_threads: 5) { |item|
+        begin
           url = item.css(".figure .figcaption .title .job_link @href").text
           html_job_detail = Nokogiri::HTML.parse(URI.open(URI.encode(url)))
           if html_job_detail.at_css(".search-result-list-detail .tabs div#tab-1").nil?
@@ -62,7 +66,7 @@ namespace :crawler do
           end
           next if item.at_css(".figure .image a @href").text == "javascript:void(0);"
           # Company attributes
-        html_company_detail = Nokogiri::HTML.parse(open(URI.encode(item.css(".figure .image a @href").text)))
+          html_company_detail = Nokogiri::HTML.parse(URI.open(URI.encode(item.css(".figure .image a @href").text)))
           next if html_company_detail.at_css(".jobsby-company").nil?
           company_css = ".jobsby-company .company-introduction .company-info .info "
           company_attributes = {
@@ -80,19 +84,24 @@ namespace :crawler do
             industry.text.tr(",","").squish
           end              
 
+          sleep rand
+          Mutex.new.synchronize {
             result = CrawlerService.imports(job_attributes, company_attributes, cities, industries)
+          }
+          sleep rand
 
           logger.info "Crawl success url : #{url}"
-        rescue Exception => e
+        rescue => e
           logger.error e
           next
         end
+      }
     end
   end
 
   task city: :environment do
     # Fetch and parse HTML document
-    html_cities = Nokogiri::HTML.parse(open("https://careerbuilder.vn/tim-viec-lam.html"))
+    html_cities = Nokogiri::HTML.parse(URI.open("https://careerbuilder.vn/tim-viec-lam.html"))
     # Get city in country
     cities_in_country = html_cities.css(".find-jobsby-categories .main-jobs-by-location .jobs-in-country .list-jobs-by-country li a").map do |title|
       {
@@ -115,7 +124,7 @@ namespace :crawler do
 
   task industry: :environment do
     # Fetch and parse HTML document
-    html_industries = Nokogiri::HTML.parse(open("https://careerbuilder.vn/tim-viec-lam.html"))
+    html_industries = Nokogiri::HTML.parse(URI.open("https://careerbuilder.vn/tim-viec-lam.html"))
     # Get industry
     industries = html_industries.css(".find-jobsby-categories .list-of-working-positions .list-jobs li a").map do |title|
       {