gscholar 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 1fec59a3b6b44a6050886d596cd8f3b3a72c8863
-  data.tar.gz: f393420400d2963d3650e7934410e7939688cc60
+  metadata.gz: 5b3e7afc6e9b27fae3181575aa7476d31aa6fd99
+  data.tar.gz: 8dd9aae6331535de9c6495cfb6155393598629c5
 SHA512:
-  metadata.gz: f8278c5ed8e3d5313b72d5728cc71d7de56f2b8549fe32d8419bfdcd73788b1bad98bd8b3d3a40ec006ce7b84e82e6bbb9e9bd3a5a0cc36a8a42fcbbc1e55f9a
-  data.tar.gz: 3d605ecff90282092385acaa6d4f5e283b15fe3e150c9f3d7b6c4d28aec17ef99bf129d7097f24842dae1c83170cb85e2db6ccbda06c5351bcacf9fec35951c5
+  metadata.gz: bfd9f8bc016098917df22c1436aba774f2770883eed14f329d9aa413af78cd4e8650fc994fe53eac76c315a4786cd36375acd48f1ee5fcded76c7a93838992a7
+  data.tar.gz: 7ff03988617a7b508e72895af066257fd5fb6f95d7919772819ea5942086805c2cf3b85a7f461c59de92a2c5c97b912ddd6a2c2f756cc8d2ab90f8f0c9316f89

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.1.1

data/gscholar.gemspec

@@ -5,7 +5,7 @@
 
 Gem::Specification.new do |s|
   s.name = "gscholar"
-  s.version = "0.1.0"
+  s.version = "0.1.1"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Zifei Tong"]
@@ -31,6 +31,7 @@ Gem::Specification.new do |s|
     "lib/gscholar/utils.rb",
     "lib/gscholar/utils/fetcher.rb",
     "lib/gscholar/utils/lazy_proxy.rb",
+    "lib/gscholar/utils/text_plain_parser.rb",
     "test/helper.rb",
     "test/test_fetcher.rb",
     "test/test_lazy_proxy.rb",

data/lib/gscholar/utils/fetcher.rb

@@ -7,6 +7,9 @@ module GScholar
       def initialize
         @cache = {}
         @agent = Mechanize.new
+        @agent.pluggable_parser['text/plain'] = TextPlainParser
+        # Without changing user-agent, google will not return utf-8 page
+        @agent.user_agent_alias = 'Mac Safari'
 
         # hack to enable BibTeX download
         scisig = @agent.get('http://scholar.google.com/scholar_settings').
@@ -16,6 +19,8 @@ module GScholar
 
       def fetch(url)
         @cache[url] ||= @agent.get(url)
+        p @cache[url].response.to_hash
+        @cache[url]
       end
     end
 

data/lib/gscholar/utils/text_plain_parser.rb

@@ -0,0 +1,12 @@
+module GScholar
+  module Utils
+    class TextPlainParser < Mechanize::File
+      def initialize uri = nil, response = nil, body = nil, code = nil
+        super uri, response, body, code
+        encoding = response['content-type'][/;(?:\s*,)?\s*charset\s*=\s*([^()<>@,;:\\\"\/\[\]?={}\s]+)/i, 1]
+        @body.force_encoding encoding if encoding && encoding != 'none'
+        @body.encode! 'UTF-8'
+      end
+    end
+  end
+end

data/lib/gscholar/utils.rb

@@ -1,5 +1,6 @@
 require 'gscholar/utils/fetcher'
 require 'gscholar/utils/lazy_proxy'
+require 'gscholar/utils/text_plain_parser'
 
 module GScholar
   module Utils

data/test/test_paper.rb

@@ -1,9 +1,13 @@
+# encoding: utf-8
+
 require 'helper'
 
 # Helmer, Scott, et al. Curious george: The ubc semantic robot vision system. Vol. 2. Technical Report AAAI-WS-08-XX, AAAI Technical Report Series, 2007.
 PAPER_0 = '8322442812667615406'
 # Maxwell, Bruce A., Brian M. Leighton, and Leah R. Perlmutter. "A Responsive Vision System to Support Human-Robot Interaction."
 PAPER_1 = '778988983414467064'
+# 何祚庥. "量子力学的建立与科技创新的评价体系——纪念普朗克创立量子论 100 周年." 昆明理工大学学报 (社会科学版) 1.1 (2001).
+PAPER_UTF8 = '11787577865338439018'
 
 class TestPaper < Test::Unit::TestCase
   def setup
@@ -35,4 +39,9 @@ class TestPaper < Test::Unit::TestCase
     c1 = @p1.citations
     assert_equal(c1.size, 0)
   end
+
+  should "successfully get paper information with utf-8 encoding" do
+    @p = GScholar::Paper.new PAPER_UTF8
+    assert_equal(@p.author.first, '何祚庥')
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: gscholar
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Zifei Tong
@@ -159,6 +159,7 @@ files:
 - lib/gscholar/utils.rb
 - lib/gscholar/utils/fetcher.rb
 - lib/gscholar/utils/lazy_proxy.rb
+- lib/gscholar/utils/text_plain_parser.rb
 - test/helper.rb
 - test/test_fetcher.rb
 - test/test_lazy_proxy.rb