indeed_scraper2022 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 0455b8d92dab8642f01c53ce43afc11eb24fbff801cce8df3306fd31eebeb223
+  data.tar.gz: eb8d24662b095fb0049209534ddfb1f0805454787105e091dc15eb7ca363d1b1
+SHA512:
+  metadata.gz: 342f6534ef851e12e78bf8e190ea98b437a6503c26510938dcb0021a30033c6be5ae819560a46ab0c11a4b42f833238517bbf9779c35c11e671d3a81ed8eedc9
+  data.tar.gz: 74be894713e2572bef1777cfafe099cc875d258fad6d741703a7cf00d22863c44cf02a877aa5106d73182d17408a16fc16f6b9538f4507c602e5588062a12334

data/lib/indeed_scraper2022.rb

@@ -0,0 +1,103 @@
+#!/usr/bin/env ruby
+
+# file: indeed_scraper2022.rb
+
+require 'mechanize'
+require 'nokorexi'
+
+# Given the nature of changes to jobsearch websites,
+# don't rely upon this gem working in the near future.
+
+
+class IndeedScraper2022
+
+  def initialize(url='https://uk.indeed.com/?r=us', q: '', location: '', debug: false)
+
+    @debug = debug
+    @url_base, @q, @location = url, q, location
+    @results = search
+
+  end
+
+  # returns an array containing the job search result
+  #
+  def results()
+    @results
+  end
+
+  def page()
+  end
+
+  # used for debugging
+  #
+  def a2()
+    @a2
+  end
+
+  private
+
+  def search(q='', location='')
+
+    a = Mechanize.new
+
+    page = a.get(@url_base)
+    form = page.forms.first
+    form.fields[0].value = @q
+    form.fields[1].value = @location
+    pg = form.submit
+
+    doc2 = Nokogiri::XML(pg.body)
+    a2 = doc2.xpath  "//a[div/div/div/div/table/tbody/tr/td/div]"
+    puts 'a2: ' + a2.length.inspect if @debug
+
+    @a2 = a2.map {|x| Rexle.new x.to_s }
+
+    @a2.map do |doc|
+
+      div = doc.element("a[@class='desktop']/div[@class='slider"  \
+          "_container']/div[@class='slider_list']/div[@class='sl"  \
+          "ider_item']/div[@class='job_seen_beacon']")
+      td = div.element("table[@class='jobCard_mainContent']/tbo"  \
+          "dy/tr/td[@class='resultContent']")
+
+      # job title (e.g. Software Developer)
+      jobtitle = td.element("div[@class='tapItem-gutter']/h2[@"  \
+          "class='jobTitle-color-purple']/span").text
+      puts 'jobtitle: ' + jobtitle.inspect if @debug
+
+      salary = td.element("div[@class='metadataContainer']/"  \
+          "div[@class='salary-snippet-container']/div[@class='sa"  \
+          "lary-snippet']/span")
+      salary = salary.text if salary
+      puts 'salary: ' + salary.inspect if @debug
+      div1 = td.element("div[@class='companyInfo']")
+
+      # company name (e.g. Coda Octopus Products Ltd)
+      company_name = div1.element("span[@class='companyName']").text
+
+      # company location (e.g. Edinburgh)
+      location = div1.element("div[@class='companyLocation']").text
+      tbody = div.element("table[@class='jobCardShelfContainer']/tbody")
+
+      div3 = tbody.element("tr[@class='underShelfFooter']/td/di"  \
+          "v[@class='result-footer']")
+
+      # job (e.g. Our products are primarily written in C#, using...)
+      jobsnippet = div3.element("div[@class='job-snippet']/ul/li").text
+
+      # visually (e.g. Posted 14 days ago)
+      dateposted =  div3.element("span[@class='date']").texts
+      date = (Date.today - dateposted.first.to_i).to_s
+
+      {
+        title: jobtitle,
+        salary: salary,
+        company: company_name,
+        location: location,
+        jobsnippet: jobsnippet,
+        date: date
+      }
+
+    end
+  end
+end

data.tar.gz.sig

@@ -0,0 +1,3 @@
+o����(-i}

+?*�i��b��S�='IS�x�$`�Wo��LBtM���sK�"����r]O�d��$Y�
7[HY(x[OLL&Rc:��惆>q�ڶ�����e4�9��N�v�5g��%��YD.��۷�@�m� &�IKB_i�ٽt�s��L����\�!���n�]vR�>�)��� MB�ML%�����<#8�_��=]�V���>~X$�<Lt����e�-
���SKU0d�)�%��&�8���
+(]�d��3=�@����I 3�(�dd��<�

metadata

@@ -0,0 +1,111 @@
+--- !ruby/object:Gem::Specification
+name: indeed_scraper2022
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- James Robertson
+autorequire: 
+bindir: bin
+cert_chain:
+- |
+  -----BEGIN CERTIFICATE-----
+  MIIEXjCCAsagAwIBAgIBATANBgkqhkiG9w0BAQsFADAsMSowKAYDVQQDDCFnZW1t
+  YXN0ZXIvREM9amFtZXNyb2JlcnRzb24vREM9ZXUwHhcNMjIwMTI0MTkxMjI1WhcN
+  MjMwMTI0MTkxMjI1WjAsMSowKAYDVQQDDCFnZW1tYXN0ZXIvREM9amFtZXNyb2Jl
+  cnRzb24vREM9ZXUwggGiMA0GCSqGSIb3DQEBAQUAA4IBjwAwggGKAoIBgQDExD/K
+  ccgeJeTOPpYSQXUzgyjX9z9wdK/Y/a6hZw8DIaYRac/4nWKEeCtFFOKJ1DK9hpba
+  jVnY5H4ZRtbp/S4ZBimUTgRwlUVbEz/93Nfnf0ZKn7yyCFyfJCedRW/lqU5ACVR7
+  DkdsGSXxQAeq87GAuW/ofCIWBB/RCrWsj7JYpDnIg+oLMOm1nBQxmiPPkW1H2zpg
+  OvWY6wItQuOxxdk5VBioecASXmQsanLyDyNSycvos24cS0msbbZfs1PuU42HxTS+
+  OjfOLRk1F02VA/lTVJCxdJx4u4jcDVJhqFP61suUN/oUFDob/rQhIT8EUITkPLY5
+  zhxDzCSZs1q5WlkTQ7MIGB1rgl8YwSfRgOhZtjkSPddK+iSRsME1pen0R6sNctRC
+  hsDNFaKAT41ZV/krNSD2lP0p7BUlKZQ6c2Whcw8NlqZkLO85CYrp5lov9EgAA+6M
+  bc4YR2Rg9lOwYFKxKVkMNGpymQANedbGJmOHDgA2O+nL9vi6uFD+HaylJx0CAwEA
+  AaOBijCBhzAJBgNVHRMEAjAAMAsGA1UdDwQEAwIEsDAdBgNVHQ4EFgQUC7f4K/R/
+  u34HepeoZ/BBZuOx2nQwJgYDVR0RBB8wHYEbZ2VtbWFzdGVyQGphbWVzcm9iZXJ0
+  c29uLmV1MCYGA1UdEgQfMB2BG2dlbW1hc3RlckBqYW1lc3JvYmVydHNvbi5ldTAN
+  BgkqhkiG9w0BAQsFAAOCAYEAweTlInWOWdP/OB0U1R5fdHeBOs9I2cn9OqH1I8AZ
+  3VGthsmza7nrrc94hCE1lQWSKwSahFPLY8AQoGibnPknCAuD5MvKSCRn3n8I7som
+  fCb2z05iv1ps49aUSLOp9nvQd2ghqIwZ2u7d7vnUvUh3FiG49KttLrnnJh4hrgpx
+  oMlT4cog/BtPs3pawgt3nbbTDamb3MTQyIMuo3uSjL867+J4I6vcDpPgCd0lPiS8
+  2ZZuRklW3qzm2Vhi7Uhxtm+POst/W6gix+3Acsx3jk+LQTSRLjvCaOCmY0ooi8D+
+  NiLKNdKO0oC8h9MyC+5EOq8Msfti3Vjr1kuFa7bcxzwSIQPFpoo+30V/uAEc0OYQ
+  qXLj/YPWi7B+J84HaWwSktqO8O0VYonAywpXTz0TcAv+6gsNF9HorTXPm/KyB6c6
+  YLGLkwmbiUHX5dRU3RwOwOrZiyvND5BIj7S6dZ6jYHe0I727apgQNc3swTz5mW6I
+  SW/2zInu2bkj/meWm5eBoWHT
+  -----END CERTIFICATE-----
+date: 2022-01-24 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: nokorexi
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.5'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.5.5
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.5'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.5.5
+- !ruby/object:Gem::Dependency
+  name: mechanize
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.8'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.8.4
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.8'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.8.4
+description: 
+email: james@jamesrobertson.eu
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/indeed_scraper2022.rb
+homepage: https://github.com/jrobertson/indeed_scraper2022
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.7.10
+signing_key: 
+specification_version: 4
+summary: Attempts to scrape the indeed.com jobsearch results (1 page).
+test_files: []