bplgeo 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 487f3110a2b6c018eeb692d229adc7fa6a282496
-  data.tar.gz: abc54d7696c6f51768300eaca230f1cff26eaee9
+  metadata.gz: 21c1271a3fb208898888438395ebafbfcf1bd167
+  data.tar.gz: 64bde835593a4f414441b03390dbe1327524b0e6
 SHA512:
-  metadata.gz: 0c0a2e5667f869bde5968845b6382ed5dfa0a1c1be086a38dee1ea705794cb52b69348dea05d87fde45a65ba8e2241b1ce1a497f7edb9730de4fd3a7d71d25ee
-  data.tar.gz: fe7d68b39d358e3abb32da18894810c1e0375e20aded599d6bb3eacc1fab5f3dddeecc45eb7b72590bbb8ab28f3eff639e073f9e2985e94ca7af5b9d26ed0b30
+  metadata.gz: 2a0254403d3d1a29c63ffbab106133083a3193d7c005716ff432385012a0eda2e0faef78eec927ace1bc931c396c8e434fd24bbfb878b397e128d4b79b78e723
+  data.tar.gz: 1a5382d42fb7f802c12569bdb983921d6335b7fc9b7b5b9f43f9a97c255e5bcfedd911250f191a9df44f983aee9b0750afa30b8b1ed6919c02c278a77e02d8e9

data/lib/bplgeo/standardizer.rb

@@ -80,7 +80,7 @@ module Bplgeo
     #Attempt to dedup a list of geographic areas.
     #FIXME: Horrendous first pass.
     #Aggresive flag removes less specific matches. IE. ['Hanoi, Vietnam' and 'Vietnam'] would return just ['Hanoi, Vietnam']
-    def self.dedup_geo(geo_list, aggresive=false)
+    def self.dedup_geo(geo_list, aggressive=false)
       geo_list = geo_list.clone
 
        base_word_geo_list = []
@@ -113,7 +113,9 @@ module Bplgeo
         }
 
         matched_words_count.each_with_index do |matched_count, matched_index|
-          if matched_count ==  base_word_geo_list[index].split(' ').size && ((base_word_geo_list[matched_index].split(' ').size < base_word_geo_list[index].split(' ').size && aggresive) || (base_word_geo_list[matched_index].split(' ').size == base_word_geo_list[index].split(' ').size))
+          matched_count ||= 0
+
+          if (matched_count ==  base_word_geo_list[matched_index].split(' ').size) && ((base_word_geo_list[matched_index].split(' ').size < base_word_geo_list[index].split(' ').size && aggressive) || (base_word_geo_list[matched_index].split(' ').size == base_word_geo_list[index].split(' ').size))
             if current_best_term.split(',').size < geo_list[matched_index].split(',').size || (current_best_term.size+1 < geo_list[matched_index].size && !geo_list[matched_index].include?('('))
               current_best_term =  geo_list[matched_index]
               indexes_to_remove << current_best_term_index

data/lib/bplgeo/town_lookup.rb

@@ -9,7 +9,7 @@ module Bplgeo
       matching_towns.each do |matching_town|
         if matching_town[:location_name].split(' ').length > matched_terms_count
           return_tgn_id = matching_town[:tgn_id]
-          matched_terms_count = matched_terms_count
+          matched_terms_count = matching_town[:location_name].split(' ').length
         end
       end
 

data/lib/bplgeo/version.rb

@@ -1,3 +1,3 @@
 module Bplgeo
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

data/lib/bplgeo.rb

@@ -10,6 +10,7 @@ module Bplgeo
   require "unidecoder"
   require "typhoeus"
   require "nokogiri"
+  require "htmlentities"
 
   def self.parse(term,parse_term=false)
     return {} if term.blank?

data/test/dummy/config/bplgeo.yml

@@ -1,23 +1,18 @@
 development:
   #scanned_image_drive: /home/bluewolf/mapped/scan_images/BPLDC/_
-  #getty_username: bplib
+  getty_username: bplib
   getty_password: 8{83N78kO;B)2
-  geonames_username: boston_library
   mapquest_key: Fmjtd%7Cluubn1utn0%2Ca2%3Do5-90b00a
   bing_key: Avmp8UMpfYiAJOYa2D-6_cykJoprZsvvN5YLv6SDalvN-BZnW9KMlCzjIV7Zrtmn
-  timeout: 100
 test: &TEST
   #scanned_image_drive: /home/bluewolf/mapped/scan_images/BPLDC/_
-  #getty_username: bplib
+  getty_username: bplib
   getty_password: 8{83N78kO;B)2
-  geonames_username: boston_library
   mapquest_key: Fmjtd%7Cluubn1utn0%2Ca2%3Do5-90b00a
   bing_key: Avmp8UMpfYiAJOYa2D-6_cykJoprZsvvN5YLv6SDalvN-BZnW9KMlCzjIV7Zrtmn
-  timeout: 100
 production:
+  #scanned_image_drive: /home/bluewolf/mapped/scan_images/BPLDC/_
   getty_username: bplib
   getty_password: 8{83N78kO;B)2
-  geonames_username: boston_library
   mapquest_key: Fmjtd%7Cluubn1utn0%2Ca2%3Do5-90b00a
   bing_key: Avmp8UMpfYiAJOYa2D-6_cykJoprZsvvN5YLv6SDalvN-BZnW9KMlCzjIV7Zrtmn
-  timeout: 7

data/test/dummy/log/test.log

@@ -0,0 +1,37 @@
+   (0.4ms)  begin transaction
+----------------------
+BplgeoTest: test_parse
+----------------------
+ETHON: Libcurl initialized
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=12.7465
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=0.660015
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=2.51674
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=0.5838760000000001
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=0.378902
+   (0.2ms)  rollback transaction
+   (0.1ms)  begin transaction
+------------------------------------------
+GeonamesTest: test_geonames_lookup_from_id
+------------------------------------------
+   (0.1ms)  rollback transaction
+   (0.1ms)  begin transaction
+--------------------------
+ParserTest: test_dedup_geo
+--------------------------
+   (0.1ms)  rollback transaction
+   (0.1ms)  begin transaction
+------------------------------
+ParserTest: test_google_parser
+------------------------------
+   (0.2ms)  rollback transaction
+   (0.2ms)  begin transaction
+--------------------------------
+TGNTest: test_tgn_lookup_from_id
+--------------------------------
+ETHON: performed EASY url= response_code=200 return_code=ok total_time=1.122996
+   (0.1ms)  rollback transaction
+   (0.1ms)  begin transaction
+------------------------------
+TownLookupTest: test_MA_lookup
+------------------------------
+   (0.1ms)  rollback transaction

data/test/standardizer_test.rb

@@ -0,0 +1,26 @@
+require 'test_helper'
+
+class ParserTest < ActiveSupport::TestCase
+  def test_dedup_geo
+    sample_array = ['Saigon, Vietnam', 'Saigon (Vietnam)', 'Vietnam', 'Vietnam, Party']
+    result = Bplgeo::Standardizer.dedup_geo(sample_array)
+    assert_equal ['Saigon, Vietnam', 'Vietnam', 'Vietnam, Party'], result
+
+    result = Bplgeo::Standardizer.dedup_geo(sample_array, true)
+    assert_equal ['Saigon, Vietnam', 'Vietnam, Party'], result
+
+    sample_array << 'Some Place, Vietnam'
+    result = Bplgeo::Standardizer.dedup_geo(sample_array)
+    assert_equal ['Saigon, Vietnam', 'Vietnam', 'Vietnam, Party', 'Some Place, Vietnam'], result
+
+    result = Bplgeo::Standardizer.dedup_geo(sample_array, true)
+    assert_equal ['Saigon, Vietnam', 'Vietnam, Party', 'Some Place, Vietnam'], result
+
+    #sample_array << 'Some Place, Vietnam, Saigon'
+    #result = Bplgeo::Standardizer.dedup_geo(sample_array, true)
+    #assert_equal ['Some Place, Vietnam, Saigon'], result
+
+  end
+
+
+end

data/test/town_lookup_test.rb

@@ -5,7 +5,7 @@ class TownLookupTest < ActiveSupport::TestCase
     result = Bplgeo::TownLookup.state_town_lookup('MA', "This test was written in Boston, MA.")
     assert_equal '7013445', result
 
-    #result = Bplgeo::TownLookup.state_town_lookup('MA', "This test was written in East Boston, MA.")
-    #assert_equal '7015009', result
+    result = Bplgeo::TownLookup.state_town_lookup('MA', "This test was written in East Boston, MA.")
+    assert_equal '7015009', result
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: bplgeo
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Boston Public Library
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-27 00:00:00.000000000 Z
+date: 2014-06-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rails
@@ -94,6 +94,20 @@ dependencies:
     - - '>='
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: htmlentities
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: sqlite3
   requirement: !ruby/object:Gem::Requirement
@@ -127,6 +141,7 @@ files:
 - lib/bplgeo/parser.rb
 - Rakefile
 - test/bplgeo_test.rb
+- test/standardizer_test.rb
 - test/dummy/Rakefile
 - test/dummy/config.ru
 - test/dummy/public/500.html
@@ -160,6 +175,7 @@ files:
 - test/dummy/bin/rails
 - test/dummy/bin/bundle
 - test/dummy/bin/rake
+- test/dummy/log/test.log
 - test/dummy/log/development.log
 - test/dummy/README.rdoc
 - test/test_helper.rb
@@ -193,6 +209,7 @@ summary: Parse string for potential geographic matches and return that data alon
   with the TGN ID and Geonames ID.
 test_files:
 - test/bplgeo_test.rb
+- test/standardizer_test.rb
 - test/dummy/Rakefile
 - test/dummy/config.ru
 - test/dummy/public/500.html
@@ -226,6 +243,7 @@ test_files:
 - test/dummy/bin/rails
 - test/dummy/bin/bundle
 - test/dummy/bin/rake
+- test/dummy/log/test.log
 - test/dummy/log/development.log
 - test/dummy/README.rdoc
 - test/test_helper.rb
@@ -233,4 +251,3 @@ test_files:
 - test/town_lookup_test.rb
 - test/parser_test.rb
 - test/geonames_test.rb
-has_rdoc: