semaphore_classification 0.1.2 → 0.2.0

data/README.rdoc

@@ -14,6 +14,11 @@ To classify documents:
   
 Mostly likely you will specify a :document_uri when classifying documents, but if you do not you will need to specify an :alternate_body
 
+== Semaphore::Client.decode_term_id(<term_id>) Options
+
+This method is used when you have the ID of a term and would like to retrieve 
+its name from the Search Enhancement Server.
+
 == Semaphore::Client.classify() Options
 
 === :document_uri (optional)

data/VERSION

@@ -1 +1 @@
-0.1.2
+0.2.0

data/lib/semaphore_classification/client.rb

@@ -15,27 +15,57 @@ module Semaphore
                           :threshold => 48, :language => LANGUAGES[:english_marathon_stemmer], :generated_keys => true, :min_avg_article_page_size => 1.0, 
                           :character_cutoff => 500000, :document_score_limit => 0, :article_mode => :single
                           }
+    
+    TERM_QUERY = 'select.exe?TBDB=disp_taxonomy&TEMPLATE=service.xml&SERVICE=browse&ID='
+    
     @@connection = nil
+    @@decode_term_ids = nil
 
     class << self
       
       def set_realm(realm, proxy=nil)
         @@connection = Connection.new(realm, proxy)
       end
+      
+      def decode_term_ids=(value)
+        raise RealmNotSpecified if @@connection.nil?
+        @@decode_term_ids = value
+        @@connection.decode_term_ids = @@decode_term_ids
+      end
+      
+      def decode_term_ids
+        @@decode_term_ids
+      end
 
       def classify(*args)
         options = extract_options!(args)
-        raise InsufficientArgs if options[:alternate_body].empty? && options[:document_uri].empty?
+        raise InsufficientArgs if options[:alternate_body].nil? && options[:document_uri].nil?
         
         result = post @@default_options.merge(options)
       end
       
+      def decode_term_id(term_id)
+        raise RealmNotSpecified if @@connection.nil?
+        begin
+          raw_host = decode_host(@@connection.realm)
+          path = [raw_host, 'cgi-bin', "#{TERM_QUERY}#{term_id}"].join('/')
+          term_doc = Nokogiri::XML.parse(open(path))
+        rescue
+          raise SemaphoreError
+        end
+        term_doc.xpath('//BROWSE_TERM/TERM/NAME').inner_text
+      end
+      
     private
     
       def post(data)
         raise RealmNotSpecified if @@connection.nil?
         @@connection.post data
       end
+      
+      def decode_host(realm)
+        realm.split('/').delete_if {|i| i == 'index.html' }.join('/').gsub(':5058', '')
+      end
     
       def extract_options!(args)
         if args.last.is_a?(Hash)

data/lib/semaphore_classification/connection.rb

@@ -3,9 +3,11 @@ module Semaphore
   class Connection
 
     attr_reader :realm
+    attr_accessor :decode_term_ids
 
     def initialize(realm, proxy=nil)
       @realm = realm
+      @decode_term_ids = nil
       @proxy = proxy
     end
 
@@ -68,7 +70,11 @@ module Semaphore
         begin
           doc = Nokogiri::XML.parse(response)
           doc.xpath('//META').each do |node|
-            data << { :term => node['value'], :key => node['key'], :score => node['score'] } if node['name'] == "Generic"
+            if node['name'] == "Generic_ID"
+              term = { :term_id => node['value'].to_i, :score => node['score'].to_f }
+              term[:term] = Client.decode_term_id(term[:term_id]) if @decode_term_ids
+              data << term
+            end
           end
         rescue
           raise DecodeError, "content: <#{response.body}>"

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: semaphore_classification
 version: !ruby/object:Gem::Version 
-  hash: 31
+  hash: 23
   prerelease: false
   segments: 
   - 0
-  - 1
   - 2
-  version: 0.1.2
+  - 0
+  version: 0.2.0
 platform: ruby
 authors: 
 - Mauricio Gomes
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-08-27 00:00:00 -04:00
+date: 2010-10-12 00:00:00 -04:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency