indeed_scraper2022 0.5.1 → 0.5.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e9d1593bc7531fd77592f8765ccac9011f47980f0fb1f0d7e9c89513a319b85e
-  data.tar.gz: ac88d94ce2271fdc4561f43ff7218baed73db60867506c20bc34522fd357cb89
+  metadata.gz: 1cf994e7ea8c7fd89ad96e55d2d3f2d3908b8c565e3f07bad5056e4d63b90b65
+  data.tar.gz: eff54d2290a32a6a78cd31646ceef2736d329b40ea327f72f7a5eb4594dd3fe2
 SHA512:
-  metadata.gz: 133e26d46ca8bae7b08aff57a7451c4cf23f507f4cb4e3875f3b3ca320c60e14f747952f1b893a8016cfd3dfb907440597864158e0cf9a5dc92a833d98855aff
-  data.tar.gz: 0c5d883b0b9a1582a2b83c8a64d47e420b7c0f79f2e6f8b1e409439fd171539adca26f10c66b57db19f9d5238c3017649c527d0e2f8fe405b3aa73c4cf276e8f
+  metadata.gz: 4da62ce0099ee02985981f8f166e371a4b05388998c94b5b5d3e33637e7dc649cc7aefbbeb19e002d777634b0bef9222563aad552c89d575757633c7764b5eb9
+  data.tar.gz: dc4a0bcdb86bef4925880c95c5b558de5e1c4cff0756e134f8502be97e0dc2a16e4e38fd7217db6b9994a02c2384956c90ec7a66913b8baaf5fbdfe92d46b322

data/lib/indeed_scraper2022.rb

@@ -262,7 +262,7 @@ class IS22Plus < IndeedScraper2022
 
   def initialize(q: '', location: '', headless: true, cookies: nil, debug: false)
     super(q: q, location: location, headless: headless, cookies: cookies,
-          debug: true)
+          debug: debug)
   end
 
   # note: The most efficient method to accumulate vacancy articles is to
@@ -342,16 +342,17 @@ end
 
 
 class IS22Archive
+  include RXFReadWriteModule
 
   attr_reader :index
 
   def initialize(filepath='/tmp/indeed', debug: false)
 
-    FileUtils.mkdir_p filepath
+    FileX.mkdir_p filepath
     @idxfile = File.join(filepath, 'index.yml')
 
-    @index = if File.exists? @idxfile then
-      YAML.load(File.read(@idxfile))
+    @index = if FileX.exists? @idxfile then
+      YAML.load(FileX.read(@idxfile))
     else
       {}
     end
@@ -372,4 +373,67 @@ class IS22Archive
 
   end
 
+  def to_html()
+
+    rows = latest().map do |h|
+
+      puts 'h: ' + h.inspect if @debug
+      co = h[:company].length > 1 ? " (%s)" % h[:company] : ''
+      "* %s: [%s](%s)%s" % [h[:added].strftime("%d %b"), h[:title], h[:link], co]
+
+    end.join("\n")
+
+
+    md = '# Indeed.com: Latest jobs
+
+' + rows
+
+    RDiscount.new(md).to_html
+
+  end
+
+  def to_form(action: '')
+
+    rows = latest().map.with_index do |h, i|
+
+      co = h[:company].length > 1 ? " (%s)" % h[:company] : ''
+
+      "<input type='checkbox' id='#{h[:jobid]}' name='#{h[:jobid]}' value='#{h[:title]}'/>
+      <label for='j#{i}'>#{h[:added].strftime("%d %b")}: #{h[:title] + ' ' + co}</label><br/>
+      "
+
+    end.join("\n")
+
+
+    return "<form action='#{action}'>#{rows}" +
+        "<input type='submit' value='submit'/></form>"
+
+  end
+
+  def filter(a)
+
+    dx = Dynarex.new
+    a2 = latest().select {|h| a.include? h[:jobid] }
+    dx.import a2
+
+    return dx
+  end
+
+  private
+
+  def latest()
+
+    a = @index.to_a.map do |id, h|
+      h[:jobid] = id
+      h[:added] = Date.parse(h[:added]) if h[:added].is_a? String
+      h
+    end
+
+    a.select do |x|
+      x[:added] >= (Date.today - 7)
+    end.reverse
+
+  end
+
 end
+

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: indeed_scraper2022
 version: !ruby/object:Gem::Version
-  version: 0.5.1
+  version: 0.5.2
 platform: ruby
 authors:
 - James Robertson
@@ -35,7 +35,7 @@ cert_chain:
   YLGLkwmbiUHX5dRU3RwOwOrZiyvND5BIj7S6dZ6jYHe0I727apgQNc3swTz5mW6I
   SW/2zInu2bkj/meWm5eBoWHT
   -----END CERTIFICATE-----
-date: 2022-05-12 00:00:00.000000000 Z
+date: 2022-05-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokorexi
@@ -78,7 +78,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: 0.3.1
 - !ruby/object:Gem::Dependency
-  name: reveal_url22
+  name: url_reveal22
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
@@ -97,6 +97,26 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: 0.1.0
+- !ruby/object:Gem::Dependency
+  name: dynarex
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.9'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.9.11
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.9'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.9.11
 description: 
 email: digital.robertson@gmail.com
 executables: []