scrub_db 2.22 → 2.23

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 92d54aa6daab25078012cec4499481f2800e2f39942923ee903a6157e5218c97
-  data.tar.gz: 57bb1f734437f6d60ae0fc1264667473b1917d55ced8a09ff623348998e3559c
+  metadata.gz: 71f435783cd9717b8a67d76e5866ed6a60ce25582e9df4e96ba8690346730b7e
+  data.tar.gz: 84d726d3d4d35ebfa89b71f943aa3c48063fe9fe5ac2bb72fcbf99d420cb8a17
 SHA512:
-  metadata.gz: 12a4e4b2aa0ee2ed5bb5d6286333f53bc20723c7e12c38fdc5505a9d6dab155fd21b42364eb89f0b1b300332b2c58f722b77d85d1b363a21c5bac74db839058b
-  data.tar.gz: d85c36b73d4affa6fe438ba271f117f558d540cfce4e7efa9d51aca2ee548c8e1618fbf2cf4ae45199722c68425122cef344a6aff961cf8cce20034d0085900f
+  metadata.gz: 5f032de540a7aa3521f2cb01d66d89339f354d4fedb4d221744b4a046a7f41942296aca6e21fdd16caa0c72c1deb1a7d1d55d610785b584edaac84f0c344ee60
+  data.tar.gz: e31b40b6a7f41be06368812e7796125d2f593f9feee5712fe854b4293117367724dd0ebdc5ecdc93a9c9c54cb7b5f7850a3755b84cb14a1764679601b2b50db0

data/Rakefile

@@ -17,15 +17,39 @@ task :console do
   require "active_support/all"
   ARGV.clear
 
-  scrubbed_webs = run_scrub_webs
+  # scrubbed_webs = run_scrub_webs
   # scrubbed_proper_strings = run_scrub_proper_strings
   # scrubbed_strings = run_scrub_strings
-  # binding.pry
+  scrubbed_hash = run_scrub_string
+  scrubbed_hash = run_scrub_proper_string
+  binding.pry
 
   IRB.start
 end
 
 
+def run_scrub_string
+  strings_criteria = {
+    pos_criteria: WebsCriteria.seed_pos_urls,
+    neg_criteria: WebsCriteria.seed_neg_urls
+  }
+  string = 'quick auto-approval gmc and bmw-world of AUSTIN tx, INC'
+  strings_obj = ScrubDb::Strings.new(strings_criteria)
+  scrubbed_hash = strings_obj.scrub_string(string)
+end
+
+
+def run_scrub_proper_string
+  strings_criteria = {
+    pos_criteria: WebsCriteria.seed_pos_urls,
+    neg_criteria: WebsCriteria.seed_neg_urls
+  }
+  string = 'quick auto-approval gmc and bmw-world of AUSTIN tx, INC'
+  strings_obj = ScrubDb::Strings.new(strings_criteria)
+  scrubbed_hash = strings_obj.scrub_proper_string(string)
+end
+
+
 def run_scrub_strings
   strings_criteria = {
     pos_criteria: WebsCriteria.seed_pos_urls,
@@ -82,11 +106,10 @@ def run_scrub_proper_strings
   ]
 
   strings_obj = ScrubDb::Strings.new(strings_criteria)
-  scrubbed_proper_strings = strings_obj.scrub_proper_strings(array_of_propers)
+  scrubbed_proper_strings = strings_obj.scrub_proper_strings(array_of_strings)
 end
 
 
-
 def run_scrub_webs
   urls = %w[
     austinchevrolet.not.real

data/lib/scrub_db/strings.rb

@@ -1,5 +1,4 @@
 
-
 module ScrubDb
   class Strings
     attr_accessor :filter, :empty_criteria
@@ -7,30 +6,40 @@ module ScrubDb
     def initialize(criteria={})
       @empty_criteria = criteria&.empty?
       @filter = ScrubDb::Filter.new(criteria) unless @empty_criteria
+      @crmf = CrmFormatter
     end
 
     def scrub_proper_strings(props=[])
-      prop_hashes = CrmFormatter.format_propers(props)
-      prop_hashes = merge_criteria(prop_hashes)
-      prop_hashes.map! { |prop_hsh| scrub_hash(prop_hsh) }
+      prop_hashes = props.map! { |str| scrub_proper_string(str) }
     end
 
-    def scrub_strings(strings=[])
-      str_hashes = strings_to_hashes(strings)
-      str_hashes = merge_criteria(str_hashes)
-      str_hashes.map! { |str_hsh| scrub_hash(str_hsh) }
+    def scrub_strings(strs=[])
+      str_hashes = strs.map! { |str| scrub_string(str) }
     end
 
-    def strings_to_hashes(strings)
-      str_hashes = strings.map { |str| { string: str } }
+
+    def scrub_proper_string(string)
+      hsh = @crmf.format_proper(string)
+      hsh = merge_criteria(hsh)
+      hsh = scrub_hash(hsh)
     end
 
-    def merge_criteria(hashes)
-      hashes.map do |hsh|
-        hsh.merge({ pos_criteria: [], neg_criteria: [] })
-      end
+    def scrub_string(string)
+      hsh = string_to_hash(string)
+      hsh = merge_criteria(hsh)
+      hsh = scrub_hash(hsh)
+    end
+
+
+    def string_to_hash(string)
+      hsh = { string: string, pos_criteria: [], neg_criteria: [] }
+    end
+
+    def merge_criteria(hsh)
+      hsh = hsh.merge({ pos_criteria: [], neg_criteria: [] })
     end
 
+
     def scrub_hash(hsh)
       str = hsh[:string]
       prop = hsh[:proper_f]
@@ -48,5 +57,4 @@ module ScrubDb
     end
 
   end
-
 end

data/lib/scrub_db/version.rb

@@ -1,3 +1,3 @@
 module ScrubDb
-  VERSION = "2.22"
+  VERSION = "2.23"
 end

data/lib/scrub_db/webs.rb

@@ -53,7 +53,7 @@ module ScrubDb
     def scrub_url_hash(url_hash)
       url = url_hash[:url_f]
       path = url_hash[:url_path]
-      href = url_hash[:href]
+      path = url_hash[:path]
       url_exts = url_hash[:url_exts]
 
       url_hash = @filter.scrub_oa(url_hash, url_exts, 'neg_exts', 'equal')

data/lib/webs_criteria.rb

@@ -37,11 +37,11 @@ class WebsCriteria
      %w(com net)
   end
 
-  # def self.seed_neg_hrefs
+  # def self.seed_neg_paths
   #   %w(? .com .jpg @ * afri after anounc apply approved blog book business buy call care career cash charit cheap check click collis commerc cont contrib deal distrib download employ event face feature feed financ find fleet form gas generat golf here holiday hospi hour info insta inventory join later light login mail mobile movie museu music news none now oil part pay phone policy priva pump quick quote rate regist review saving schedul service shop sign site speci ticket tire today transla travel truck tv twitter watch youth)
   # end
   #
-  # def self.seed_pos_hrefs
+  # def self.seed_pos_paths
   #   %w(team staff management)
   # end
 

data/scrub_db.gemspec

@@ -37,11 +37,10 @@ Gem::Specification.new do |spec|
 
   spec.required_ruby_version = '~> 2.5.1'
   spec.add_dependency 'activesupport', '~> 5.2'
+  spec.add_dependency 'crm_formatter', '~> 2.64'
   spec.add_dependency 'utf8_sanitizer', '~> 2.16'
-  spec.add_dependency 'crm_formatter', '~> 2.61'
 
-  # spec.add_development_dependency 'activesupport', '~> 5.2'
-  # spec.add_development_dependency 'utf8_sanitizer', '~> 2.15'
+
   # spec.add_dependency "activesupport-inflector", ['~> 0.1.0']
   spec.add_development_dependency 'bundler', '~> 1.16', '>= 1.16.2'
   spec.add_development_dependency 'pry', '~> 0.11.3'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: scrub_db
 version: !ruby/object:Gem::Version
-  version: '2.22'
+  version: '2.23'
 platform: ruby
 authors:
 - Adam Booth
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2018-07-01 00:00:00.000000000 Z
+date: 2018-07-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -25,33 +25,33 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '5.2'
 - !ruby/object:Gem::Dependency
-  name: utf8_sanitizer
+  name: crm_formatter
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.16'
+        version: '2.64'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.16'
+        version: '2.64'
 - !ruby/object:Gem::Dependency
-  name: crm_formatter
+  name: utf8_sanitizer
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.61'
+        version: '2.16'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.61'
+        version: '2.16'
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement