anystyle-parser 0.6.3 → 0.6.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 70f2754e977f718991ed7b0cba97ebd7edc81482
-  data.tar.gz: 848af0659993730ee5f8fd7feb18c75b216c0cdf
+  metadata.gz: bc9078e1caff3c730c0db06e5ccdf4785b86b7a6
+  data.tar.gz: 9a6aae1bfde176b574c49ceab02b78254fdf3251
 SHA512:
-  metadata.gz: 29aac5f181336dd316c7d403c2f7315cbbd5364bce69cc4b8dd3946a24bdaba52cc9b6e8de74cccecca359b023566fac96e945a38de509d6dbb41805208881d8
-  data.tar.gz: 95602ac50a24d1858fe7c0bc09b1369d8ebc2c7400b98fbdc6f6f86a5f3ab2c07542cac158072b1cf8b42908ab3c5d060f4ded0627be6c85c3f95736be0ea274
+  metadata.gz: 3eff36b014282cb5147fc3224c9c86564390ea125a819c62745a13ebfb35d7cd7ca62bddbefbb00bf0597cdf23b596e858c08a2b759e9234decbb8515acbcf46
+  data.tar.gz: 7316bef01b00867ebe5f06a7d6fdba4827bb2b6ea3b7429ec16fdfa6c0fa2492ccd40cb4945b7ce16f90b61cd5bb80e3e80f723a1d7a7a843a1702a203babfde

data/README.md

@@ -14,6 +14,11 @@ train the model with data that is relevant to your parsing needs), and
 compatibility (Anystyle-Parser exports to Ruby Hashes, BibTeX, or the
 CSL/CiteProc JSON format).
 
+Web Application and Web Service
+-------------------------------
+Anystyle-Parser is avaialble as a web application and a web service at
+[http://anystyle.io](http://anystyle.io).
+
 Installation
 ------------
 

data/anystyle-parser.gemspec

@@ -10,7 +10,7 @@ Gem::Specification.new do |s|
   s.platform    = Gem::Platform::RUBY
   s.authors     = ['Sylvester Keil']
   s.email       = ['http://sylvester.keil.or.at']
-  s.homepage    = 'http://github.com/inukshuk/anystyle-parser'
+  s.homepage    = 'http://anystyle.io'
   s.summary     = 'Smart and fast academic bibliography parser.'
   s.description = 'A sophisticated parser for academic reference lists and bibliographies based on machine learning algorithms using conditional random fields.'
   s.license     = 'FreeBSD'
@@ -20,7 +20,7 @@ Gem::Specification.new do |s|
   s.add_runtime_dependency('bibtex-ruby', '~>3.0')
   s.add_runtime_dependency('builder', '>=3.0', '<4.0')
   s.add_runtime_dependency('wapiti', '~>0.1')
-  s.add_runtime_dependency('namae', '~>0.8', '>=0.8.6')
+  s.add_runtime_dependency('namae', '~>0.8', '>=0.8.7')
 
   s.files        = `git ls-files`.split("\n").reject { |path|
     path.start_with?('.')

data/lib/anystyle/parser/normalizer.rb

@@ -53,10 +53,30 @@ module Anystyle
         unmatched(key, hash, dangling) unless dangling.empty?
 
         token.gsub!(/^[^[:alnum:]]+|[^[:alnum:]]+$/, '')
+
         hash[key] = token
         hash
       end
 
+      def normalize_key(hash)
+        token, *dangling =  hash[:key]
+        unmatched(:key, hash, dangling) unless dangling.empty?
+
+        token.gsub!(/^[^[:alnum:]]+|[^[:alnum:]]+$/, '')
+        token.gsub!(/^bibitem\{/i, '')
+
+        hash[:key] = token
+        hash
+      end
+
+      def normalize_citation_number(hash)
+        token, *dangling =  hash[:citation_number]
+        unmatched(:citation_number, hash, dangling) unless dangling.empty?
+
+        hash[:citation_number] = token[/\d+/] || token
+        hash
+      end
+
       def normalize_author(hash)
         authors, *dangling = hash[:author]
         unmatched(:author, hash, dangling) unless dangling.empty?
@@ -253,7 +273,7 @@ module Anystyle
         if date =~ /(\d{4})/
           hash[:year] = $1.to_i
 
-          if hash.key?(:month) && date =~ /(\d{1,2})\b/
+          if hash.key?(:month) && date =~ /\b(\d{1,2})\b/
             hash[:day] = $1.to_i
           end
 

data/lib/anystyle/parser/version.rb

@@ -1,5 +1,5 @@
 module Anystyle
   module Parser
-    VERSION = '0.6.3'.freeze
+    VERSION = '0.6.4'.freeze
   end
 end

data/spec/anystyle/parser/normalizer_spec.rb

@@ -99,6 +99,7 @@ module Anystyle
           h[:year].should == 2009
           h[:month].should == 7
           h.should_not have_key(:date)
+          h.should_not have_key(:day)
         end
 
         it 'extracts month and year from a string like "(1997 Sept.)"' do
@@ -106,10 +107,12 @@ module Anystyle
           h[:year].should == 1997
           h[:month].should == 9
           h.should_not have_key(:date)
+          h.should_not have_key(:day)
 
           h = Normalizer.instance.normalize_date(:date => '(1997 Okt.)')
           h[:year].should == 1997
           h[:month].should == 10
+          h.should_not have_key(:day)
         end
 
         it 'extracts days if month and year are present' do

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: anystyle-parser
 version: !ruby/object:Gem::Version
-  version: 0.6.3
+  version: 0.6.4
 platform: ruby
 authors:
 - Sylvester Keil
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-05-12 00:00:00.000000000 Z
+date: 2014-05-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bibtex-ruby
@@ -67,7 +67,7 @@ dependencies:
         version: '0.8'
     - - '>='
       - !ruby/object:Gem::Version
-        version: 0.8.6
+        version: 0.8.7
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
@@ -77,7 +77,7 @@ dependencies:
         version: '0.8'
     - - '>='
       - !ruby/object:Gem::Version
-        version: 0.8.6
+        version: 0.8.7
 description: A sophisticated parser for academic reference lists and bibliographies
   based on machine learning algorithms using conditional random fields.
 email:
@@ -116,7 +116,7 @@ files:
 - spec/fixtures/train_dps.txt
 - spec/profile.rb
 - spec/spec_helper.rb
-homepage: http://github.com/inukshuk/anystyle-parser
+homepage: http://anystyle.io
 licenses:
 - FreeBSD
 metadata: {}