cambridge_dict 0.2.2 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8c7205facd81e402d36a810dc1b595547207b0f6e3168193fd4bcb56b5c2f9ee
-  data.tar.gz: 29adecbe8c3d0f96e59f8b43ff1b34fc093291af9a2649568ad716104aa78792
+  metadata.gz: 7b1a1cd7771bcdca2f9784cbcb7d68bb65d685e06b5387a9ac1c1090529339f0
+  data.tar.gz: fa35b9a0d9c7640b1a4f71c0e01ec0ab913a9781026e9ed4ecaecd2341aebefe
 SHA512:
-  metadata.gz: 7dfa104d86a45b96b7dac3b7a97beb99c14f662ddc25b58b2e58e2e262d82f8730458b39e5e6ffb2aca9824188bb531e54bbcd38f0841e4777c561777269c13c
-  data.tar.gz: 2e8f2f7f5ceac4f88568b7752d950add874c99472671b5ff04e86c3330042de34774462eb6a42a0833b1d252ed808619c455a5dc24d80dca82b42e324c629673
+  metadata.gz: dd1fac189c809b80bb3938a29119efba3e834c4c9149caa4a8377ddee9eb1d6fbda8003bbd8c90434a85f7f632144113f7540774758374ff0a555a1925a19bf5
+  data.tar.gz: c338c5002025478c39951420698eee13cd2f4db41d28dd1cc9d90fafbb81c2cc04001935302c4d14033c0e1c3e56512c3b50a4695e15afdcb69eadc5557f157b

data/README.md

@@ -12,13 +12,13 @@ gem install cambridge_dict
 
 ```
 # throws error
-client = CambridgeDictionary::Client.lookup! "hello", "en"
+result = CambridgeDict::Client.lookup! "hello", language: "en"
 
 # returns nil if not found
-client = CambridgeDictionary::Client.lookup "hello", "en"
+result = CambridgeDict::Client.lookup "hello", language: "en"
 
 # can also create instance
-client = CambridgeDictionary::Client.new
+client = CambridgeDict::Client.new
 result = client.lookup("back", language: "en")
 
 return if result.nil?
@@ -64,6 +64,10 @@ return if result.nil?
     :source=>"cald4-us",                                                                        
     :text=>"in return: ",
     :translation=>nil,                                                                                             
+
+# sometimes words is returned
+result = client.lookup("a", language: "en")
+result[:words] => ["A", "a"]
 ```
 
 ## Webserver Setup

data/lib/cambridge_dict/version.rb

@@ -1,3 +1,3 @@
 module CambridgeDict
-  VERSION = "0.2.2"
+  VERSION = "0.3.0"
 end

data/lib/cambridge_dict.rb

@@ -37,17 +37,23 @@ module CambridgeDict
       encoded_word = PARSER.escape(word)
       wiki_verbs = fetch_verbs("#{@wiki_base}/#{encoded_word}")
 
-      word = extract_word(doc)
+      words = extract_word(doc)
 
-      return nil if word.nil?
+      return nil if words.nil?
 
-      {
-        word: word,
+      resp = {
+        word: words&.last,
         pos: extract_pos(doc),
         verbs: wiki_verbs,
         pronunciations: extract_pronunciation(doc),
         definitions: extract_definitions(doc)
       }
+
+      if words.size > 1
+        resp[:words] = words
+      end
+
+      resp
     end
 
     def lookup!(word, language: 'en')
@@ -86,7 +92,7 @@ module CambridgeDict
       doc = Nokogiri::HTML(response.body)
       verbs = []
 
-      rows = doc.css('tr > td > p')
+      rows = doc.css('table:not(.toccolours) tr > td > p')
 
       lines = rows.map {|i| i.children.collect {|c| c.text } }.flatten.map(&:strip).reject(&:empty?)
 
@@ -99,7 +105,7 @@ module CambridgeDict
     end
 
     def extract_word(doc)
-      doc.at_css('.superentry .tb')&.text
+      doc.at_css('.superentry .tb')&.text&.split ", "
     end
 
     def extract_pos(doc)
@@ -147,13 +153,14 @@ module CambridgeDict
 
         pos = entry&.at_css('.pos.dpos')&.text
         headword = entry&.at_css('.headword .hw')&.text
+        text = clean_string block.at_css('.def.ddef_d.db')&.text&.strip
 
         definition = {
           id: index,
           headword: headword,
           pos: pos,
           source: source || nil,
-          text: block.at_css('.def.ddef_d.db')&.text&.strip,
+          text: text,
           translation: block.at_css('.def-body.ddef_b > span.trans.dtrans')&.text&.strip,
           example: extract_examples(block)
         }
@@ -177,6 +184,10 @@ module CambridgeDict
 
       examples
     end
+
+    def clean_string(input)
+      input&.gsub(/\s+/, ' ')&.strip
+    end
   end
 
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: cambridge_dict
 version: !ruby/object:Gem::Version
-  version: 0.2.2
+  version: 0.3.0
 platform: ruby
 authors:
 - Ben D'Angelo
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-10-23 00:00:00.000000000 Z
+date: 2024-12-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri