remote_job_scraper 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5bbc3d06e54e5b828ef7817799a7305d914eac90
-  data.tar.gz: 72a6243b5a1f0568a7e96d64c1f55bdd4036b846
+  metadata.gz: b880f59b8ef9959438e8025736371ee13b88ec7c
+  data.tar.gz: a7e1aff1067ea8bdc33d79962cfe749c76c111b6
 SHA512:
-  metadata.gz: a8b357f4170c3b1a4d70571c15598d2437b2f9eb24dc9c0185b37d074fd2e18ccb36f2a99230fcd28cc6f14367484b06d9ceaf2be735cb61d882175c9d25c1c6
-  data.tar.gz: 6789a0f0f69246f93d5390ecd6ae09e2059dea4be51770eeafef2479ed147d60c8bf3dc34b55c47e766123f1a9785383793a61d2bc53beaace05127a7016af4e
+  metadata.gz: 92d3ba352d3f42be0f4840cb95fb46521eda24fcc669697d4f1cb1882bc15435fbbc55e096adab99e11e61a20e9596231789fd735e719f4af576e3912ff1b6bf
+  data.tar.gz: 980d533d890bd8ebc3ba2da1e69d9a951c98ce8290aea2a2b136919281d1e5bf1c11c4a8e0110e5cc595e75360ef86c4a9cd0a1cbafddc261791b020a6346568

data/README.md

@@ -11,7 +11,7 @@ Going through many job listings and finding the right one may be a time-consumin
 
 ## Usage
 
-    $ bundle exec exe/remote_job_scraper
+    $ remote_job_scraper
 
 * Tested with Ruby versions:
 

data/lib/remote_job_scraper/version.rb

@@ -1,3 +1,3 @@
 module RemoteJobScraper
-  VERSION = "0.2.0"
+  VERSION = "0.3.0"
 end

data/lib/sites/jobs_rails42.rb

@@ -12,19 +12,18 @@ module Sites
     JOB_ITEM_SELECTOR = 'li.job-offers__item a'.freeze
     STORE_DIR = 'data/jobs_rails42'.freeze
 
-    NUMBER_OF_PAGES = 10
-
-    def initialize(job_type: :programming)
+    def initialize(job_type: :programming, total_pages: 4)
       @job_type = job_type
       @url = build_url
       @doc = nil
       @current_time = Time.new
       @timestamp = @current_time.strftime("%Y%m%d%H%M%S")
+      @total_pages = total_pages
       @count = get_count
     end
 
     def collect_jobs
-      (1..NUMBER_OF_PAGES).to_a.each do |page|
+      (1..@total_pages).to_a.each do |page|
         current_page = "#{@url}?page=#{page}"
         doc = Nokogiri::HTML(open_page(current_page))
         process_page(doc, current_page, page)
@@ -45,20 +44,19 @@ module Sites
           offer_text = job_page.css('.job-offer__description').to_s
 
           location = Support::OfferParser.get_location(offer_text)
-          region   = nil
           keywords = Support::OfferParser.get_keywords(offer_text)
 
-          csv << [job_url, location, region, keywords]
+          csv << [job_url, location, keywords]
         end
       end
 
-      puts "[Done] Collected #{@count} job offers from #{url}. Data stores in: #{filepath}." if page == NUMBER_OF_PAGES
+      puts "[Done] Collected #{@count} job offers from #{url}. Data stores in: #{filepath}." if page == @total_pages
     end
 
     private
 
     def get_count
-      25 * NUMBER_OF_PAGES
+      25 * @total_pages
     end
   end
 end

data/lib/sites/remote_ok.rb

@@ -24,10 +24,9 @@ module Sites
           offer_text = job_page.css('td.heading').to_s
 
           location = Support::OfferParser.get_location(offer_text)
-          region   = nil
           keywords = Support::OfferParser.get_keywords(offer_text)
 
-          csv << [job_url, location, region, keywords]
+          csv << [job_url, location, keywords]
         end
       end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: remote_job_scraper
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - Rafał Trojanowski