qa 0.5.0 → 0.6.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 6edd488c2f33caee0d368664c08a2f4ea70f3b9e
-  data.tar.gz: 0bc9d51827a5ece5ce2bf71d8b184d232d859c6b
+  metadata.gz: 16b9b7d34d5db8f520bb0a0c3330f75b836d0d01
+  data.tar.gz: fcea780be638d13a57f4ccd539f3ce90c5fcca48
 SHA512:
-  metadata.gz: 592b389388e0fde9fd85996694ecb44a5bd477cbf78c4b10d066f757989a0d4a6d842d3832f8475bb4b77040afec66131db8ed8df52a4ae81de3789d7c7fffb1
-  data.tar.gz: 406889f087a33c4089b0f72e43e91f4a542e3569c29116f3657155f30a8591e6a8140c8785ec1fc70e9ac061b9dcb43c371a8b526ee15b5543dbe9fbcbe7b4e8
+  metadata.gz: 85bdd4ca0d32bb58b07326aad0a19721fc3efc8139a16af66d8ab54f2c132458a4205e185bf0f5f9038f33f6a5a72d6f964bec722f9a711f8a93c370899dd012
+  data.tar.gz: 1a7a76fdc4e0c1b8cb843f95d122f9c21785023779bf19b2ea078845c1aed829a8c9b9709f2947e26b5c81c5d5333f225a6a82995a483544cd39a6ed2bf8c4c9

data/README.md

@@ -111,6 +111,43 @@ QA then presents this data to you in JSON format:
         etc...
     ]
 
+### FAST
+
+In the same manner, QA provides a wrapper around OCLC's FAST autocomplete
+service. The following subauthorities are available:
+
+* all
+* personal
+* corporate
+* event
+* uniform
+* topical
+* geographic
+* form_genre
+
+Example qa URL: /qa/search/assign_fast/all?q=periodic+table
+
+The result includes both 'label' and 'value' to help users select the correct
+heading, e.g.:
+
+    [
+        {
+            "id":"fst01789629",
+            "label":"Periodic table",
+            "value":"Periodic table"
+        },
+        {
+            "id":"fst01789629",
+            "label":"Periodic table (Saunders, N.) USE Periodic table",
+            "value":"Periodic table"
+        },...
+    ]
+
+Make sure you handle these correctly in your form.
+
+For more details on this OCLC API, see
+http://www.oclc.org/developer/develop/web-services/fast-api/assign-fast.en.html
+
 ### Local Authorities
 
 For simple use cases when you have a few terms that don't change very often.

data/app/controllers/qa/terms_controller.rb

@@ -30,7 +30,7 @@ class Qa::TermsController < ApplicationController
 
   def init_authority
     begin
-      mod = authority_class.constantize
+      mod = authority_class.camelize.constantize
     rescue NameError => e
       logger.warn "Unable to initialize authority #{authority_class}"
       head :not_found

data/lib/qa/authorities.rb

@@ -13,4 +13,6 @@ module Qa::Authorities
   autoload :Oclcts
   autoload :Tgnlang
   autoload :WebServiceBase
+  autoload :AssignFast
+  autoload :AssignFastSubauthority
 end

data/lib/qa/authorities/assign_fast.rb

@@ -0,0 +1,17 @@
+module Qa::Authorities
+  # Provide authority namespace
+  module AssignFast
+    extend AuthorityWithSubAuthority
+    extend AssignFastSubauthority
+
+    require 'qa/authorities/assign_fast/generic_authority'
+    # Create an authority object for given subauthority
+    #
+    # @param [String] subauthority to use
+    # @return [GenericAuthority]
+    def self.subauthority_for(subauthority)
+      validate_subauthority!(subauthority)
+      GenericAuthority.new(subauthority)
+    end
+  end
+end

data/lib/qa/authorities/assign_fast/generic_authority.rb

@@ -0,0 +1,61 @@
+module Qa::Authorities
+  # A wrapper around the FAST api for use with questioning_authority
+  # API documentation:
+  # http://www.oclc.org/developer/develop/web-services/fast-api/assign-fast.en.html
+  class AssignFast::GenericAuthority < Base
+    attr_reader :subauthority
+    def initialize(subauthority)
+      @subauthority = subauthority
+    end
+
+    include WebServiceBase
+
+    # Search the FAST api
+    #
+    # @param [String] the query
+    # @return json results
+    def search q
+      url = build_query_url q
+      begin
+        raw_response = get_json(url)
+      rescue JSON::ParserError => error
+        Rails.logger.info "Could not parse response as JSON. Request url: #{url}"
+        return []
+      end
+      parse_authority_response(raw_response)
+    end
+
+    # Build a FAST API url
+    #
+    # @param [String] the query
+    # @return [String] the url
+    def build_query_url q
+      escaped_query = clean_query_string q
+      index = AssignFast.index_for_authority(subauthority)
+      return_data = "#{index}%2Cidroot%2Cauth%2Ctype"
+      num_rows = 20 # max allowed by the API
+      url = "http://fast.oclc.org/searchfast/fastsuggest?&query=#{escaped_query}&queryIndex=#{index}&queryReturn=#{return_data}&suggest=autoSubject&rows=#{num_rows}"
+    end
+
+    private
+
+    # Removes characters from the query string that are not tolerated by the API
+    #   See oclc sample code at
+    #   http://experimental.worldcat.org/fast/assignfast/js/assignFASTComplete.js
+    def clean_query_string q
+      URI.escape(q.gsub(/-|\(|\)|:/, ""))
+    end
+
+    def parse_authority_response(raw_response)
+      results = raw_response['response']['docs'].map do |doc|
+        index = AssignFast.index_for_authority(subauthority)
+        term = doc[index].first
+        if doc['type'] == 'alt'
+          term += ' USE ' + doc['auth']
+        end
+        { id: doc['idroot'], label: term, value: doc['auth'] }
+      end
+    end
+
+  end
+end

data/lib/qa/authorities/assign_fast_subauthority.rb

@@ -0,0 +1,31 @@
+# Encapsulate information about assignFAST subauthorities
+module Qa::Authorities::AssignFastSubauthority
+
+  # Hash of subauthority names used in qa routes => 'index' used by API
+  Subauthorities = {
+    'all'        => 'suggestall',
+    'personal'   => 'suggest00',
+    'corporate'  => 'suggest10',
+    'event'      => 'suggest11',
+    'uniform'    => 'suggest30',
+    'topical'    => 'suggest50',
+    'geographic' => 'suggest51',
+    'form_genre' => 'suggest55',
+  }
+
+  # Get a list of subauthorities by name
+  #
+  # @return [Array<String>]
+  def subauthorities
+    Subauthorities.keys
+  end
+
+  # Get an API index name from an English name
+  #
+  # @param [String] English name
+  # @return [String] index name
+  def index_for_authority authority
+    Subauthorities[authority]
+  end
+
+end

data/lib/qa/authorities/getty.rb

@@ -3,14 +3,23 @@ require 'uri'
 module Qa::Authorities
   module Getty
     require 'qa/authorities/getty/aat'
+    require 'qa/authorities/getty/tgn'
+    require 'qa/authorities/getty/ulan'
     extend AuthorityWithSubAuthority
 
     def self.subauthorities
-      [ "aat" ]
+      [ "aat" , "tgn", "ulan" ]
     end
 
-    def self.subauthority_class(_)
-      AAT
+    def self.subauthority_class(subauthority)
+      case subauthority
+        when 'aat'
+          AAT
+        when 'tgn'
+          TGN
+        when 'ulan'
+          Ulan
+      end
     end
   end
 end

data/lib/qa/authorities/getty/aat.rb

@@ -1,6 +1,7 @@
 module Qa::Authorities
   class Getty::AAT < Base
     include WebServiceBase
+
     def search q
       parse_authority_response(json(build_query_url(q)))
     end
@@ -23,7 +24,7 @@ module Qa::Authorities
                  skos:inScheme <http://vocab.getty.edu/aat/> ;
                  gvp:prefLabelGVP [skosxl:literalForm ?name].
               FILTER regex(?name, \"#{search}\", \"i\") .
-            } LIMIT 10"
+            } ORDER BY ?name"
     end
 
     def untaint(q)
@@ -44,7 +45,7 @@ module Qa::Authorities
 
     private
 
-    # Reformats the data received from the LOC service
+    # Reformats the data received from the Getty service
     def parse_authority_response(response)
       response['results']['bindings'].map do |result|
         { 'id' => result['s']['value'], 'label' => result['name']['value'] }

data/lib/qa/authorities/getty/tgn.rb

@@ -0,0 +1,80 @@
+module Qa::Authorities
+  class Getty::TGN < Base
+    include WebServiceBase
+
+    def search q
+      parse_authority_response(json(build_query_url(q)))
+    end
+
+    # get_json is not ideomatic, so we'll make an alias
+    def json(*args)
+      get_json(*args)
+    end
+
+    def build_query_url q
+      query = URI.escape(sparql(untaint(q)))
+      # Replace ampersands, otherwise the query will fail
+      # Gsub hack to convert the encoded regex in the REPLACE into a form Getty understands
+      "http://vocab.getty.edu/sparql.json?query=#{query.gsub('&','%26').gsub(',[%5E,]+,[%5E,]+$','%2C[^%2C]%2B%2C[^%2C]%2B%24')}&_implicit=false&implicit=true&_equivalent=false&_form=%2Fsparql"
+    end
+
+    # Use a regex to exclude the continent and 'world' from the query
+    # If only one word is entered only search the name (not the parent string)
+    # If more than one word is entered, one word must appear in the name, and all words in either parent or name
+    def sparql(q)
+      search = untaint(q)
+      if search.include?(' ')
+        ex = "(("
+        search.split(' ').each do | i |
+          ex += "regex(CONCAT(?name, ', ', REPLACE(str(?par), \",[^,]+,[^,]+$\", \"\")), \"#{i}\",\"i\" ) && "
+        end
+        ex = ex[0..ex.length - 4]
+        ex += ') && ('
+        search.split(' ').each do | i |
+          ex += "regex(?name, \"#{i}\",\"i\" ) || "
+        end
+        ex = ex[0..ex.length - 4]
+        ex += ") )"
+
+      else
+        ex = "regex(?name, \"#{search}\", \"i\")"
+      end
+
+      # The full text index matches on fields besides the term, so we filter to ensure the match is in the term.
+      sparql = "SELECT DISTINCT ?s ?name ?par {
+              ?s a skos:Concept; luc:term \"#{search}\";
+                 skos:inScheme <http://vocab.getty.edu/tgn/> ;
+                 gvp:prefLabelGVP [skosxl:literalForm ?name] ;
+                  gvp:parentString ?par .
+              FILTER #{ex} .
+            } ORDER BY ?name ASC(?par)"
+    end
+
+    def untaint(q)
+      q.gsub(/[^\w\s-]/, '')
+    end
+
+    def find id
+      json(find_url(id))
+    end
+
+    def find_url id
+      "http://vocab.getty.edu/tgn/#{id}.json"
+    end
+
+    def request_options
+      { accept: 'application/sparql-results+json'}
+    end
+
+    private
+
+    # Reformats the data received from the service
+    # Adds the parentString, minus the contintent and 'World' for disambiguation
+    def parse_authority_response(response)
+      response['results']['bindings'].map do |result|
+        { 'id' => result['s']['value'], 'label' => result['name']['value']  + ' (' + result['par']['value'].gsub(/\,[^\,]+\,[^\,]+$/, '') + ')' }
+      end
+    end
+
+  end
+end

data/lib/qa/authorities/getty/ulan.rb

@@ -0,0 +1,71 @@
+module Qa::Authorities
+  class Getty::Ulan < Base
+    include WebServiceBase
+
+    def search q
+      parse_authority_response(json(build_query_url(q)))
+    end
+
+    # get_json is not ideomatic, so we'll make an alias
+    def json(*args)
+      get_json(*args)
+    end
+
+    def build_query_url q
+      query = URI.escape(sparql(untaint(q)))
+      # Replace ampersands, otherwise the query will fail
+      "http://vocab.getty.edu/sparql.json?query=#{URI.escape(sparql(q)).gsub('&','%26')}&_implicit=false&implicit=true&_equivalent=false&_form=%2Fsparql"
+    end
+
+    def sparql(q)
+      search = untaint(q)
+      # if more than one term is supplied, check both preferred and alt labels
+      if search.include?(' ')
+        ex = "("
+        search.split(' ').each do | i |
+          ex += "regex(CONCAT(?name, ' ', ?alt), \"#{i}\",\"i\" ) && "
+        end
+        ex = ex[0..ex.length - 4]
+        ex += ")"
+      else
+        ex = "regex(?name, \"#{search}\", \"i\")"
+      end
+      # The full text index matches on fields besides the term, so we filter to ensure the match is in the term.
+      sparql = "SELECT DISTINCT ?s ?name ?bio {
+              ?s a skos:Concept; luc:term \"#{search}\";
+                 skos:inScheme <http://vocab.getty.edu/ulan/> ;
+                 gvp:prefLabelGVP [skosxl:literalForm ?name] ;
+                 foaf:focus/gvp:biographyPreferred [schema:description ?bio] ;
+                 skos:altLabel ?alt .
+              FILTER #{ex} .
+            } ORDER BY ?name"
+    end
+
+    def untaint(q)
+      q.gsub(/[^\w\s-]/, '')
+    end
+
+    def find id
+      json(find_url(id))
+    end
+
+    def find_url id
+      "http://vocab.getty.edu/ulan/#{id}.json"
+    end
+
+    def request_options
+      { accept: 'application/sparql-results+json'}
+    end
+
+    private
+
+    # Reformats the data received from the Getty service
+    # Add the bio for disambiguation
+    def parse_authority_response(response)
+      response['results']['bindings'].map do |result|
+        { 'id' => result['s']['value'], 'label' => result['name']['value'] + ' (' + result['bio']['value'] + ')' }
+      end
+    end
+
+  end
+end

data/lib/qa/authorities/web_service_base.rb

@@ -1,4 +1,4 @@
-require 'rest_client'
+require 'faraday'
 
 module Qa::Authorities
   module WebServiceBase
@@ -6,12 +6,14 @@ module Qa::Authorities
 
     # mix-in to retreive and parse JSON content from the web
     def get_json(url)
-      r = RestClient.get url, request_options
+      r = response(url).body
       JSON.parse(r)
     end
 
-    def request_options
-      { accept: :json }
+    def response(url)
+      Faraday.get(url) do |req|
+        req.headers['Accept'] = 'application/json'
+      end
     end
   end
 end

data/lib/qa/version.rb

@@ -1,3 +1,3 @@
 module Qa
-  VERSION = "0.5.0"
+  VERSION = "0.6.0"
 end

data/spec/controllers/terms_controller_spec.rb

@@ -49,20 +49,34 @@ describe Qa::TermsController, :type => :controller do
 
   describe "#search" do
 
-    before :each do
-      stub_request(:get, "http://id.loc.gov/search/?format=json&q=Berry&q=cs:http://id.loc.gov/authorities/names").
-        with(:headers => {'Accept'=>'application/json'}).
-        to_return(:body => webmock_fixture("loc-names-response.txt"), :status => 200)
-    end
+    context "loc" do
+      before :each do
+        stub_request(:get, "http://id.loc.gov/search/?format=json&q=Berry&q=cs:http://id.loc.gov/authorities/names").
+          with(:headers => {'Accept'=>'application/json'}).
+          to_return(:body => webmock_fixture("loc-names-response.txt"), :status => 200)
+      end
+
+      it "should return a set of terms for a tgnlang query" do
+        get :search, {:q => "Tibetan", :vocab => "tgnlang" }
+        expect(response).to be_success
+      end
 
-    it "should return a set of terms for a tgnlang query" do
-      get :search, {:q => "Tibetan", :vocab => "tgnlang" }
-      expect(response).to be_success
+      it "should not return 404 if subauthority is valid" do
+        get :search, { :q => "Berry", :vocab => "loc", :subauthority => "names" }
+        expect(response).to be_success
+      end
     end
 
-    it "should not return 404 if subauthority is valid" do
-      get :search, { :q => "Berry", :vocab => "loc", :subauthority => "names" }
-      expect(response).to be_success
+    context "assign_fast" do
+      before do
+        stub_request(:get, "http://fast.oclc.org/searchfast/fastsuggest?query=word&queryIndex=suggest50&queryReturn=suggest50,idroot,auth,type&rows=20&suggest=autoSubject").
+          with(:headers => {'Accept'=>'application/json'}).
+          to_return(:body => webmock_fixture("assign-fast-topical-result.json"), :status => 200, :headers => {})
+      end
+      it "succeeds if authority class is camelcase" do
+        get :search, { :q => "word", :vocab => "assign_fast", :subauthority => "topical" }
+        expect(response).to be_success
+      end
     end
 
   end
@@ -103,7 +117,7 @@ describe Qa::TermsController, :type => :controller do
 
       before do
         stub_request(:get, "http://id.loc.gov/authorities/subjects/sh85077565.json").
-          with(:headers => {'Accept'=>'application/json', 'Accept-Encoding'=>'gzip, deflate', 'User-Agent'=>'Ruby'}).
+          with(:headers => {'Accept'=>'application/json'}).
           to_return(:status => 200, :body => webmock_fixture("loc-names-response.txt"), :headers => {})
       end