pdf_info 0.4.0 → 0.5.0

checksums.yaml

@@ -0,0 +1,15 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    OGYzNzE5ZTY5NDdmOWNkZmRkZTE5MTgyODk1YjM4ODFjZjY5M2E2Nw==
+  data.tar.gz: !binary |-
+    NTFiYjlhNGUzYjM0OGM3YTNhZTdmNTg0N2I4N2VlYTMwMDNiZjM2Mg==
+!binary "U0hBNTEy":
+  metadata.gz: !binary |-
+    YmEyMjM1NzFjYTRjNjQxNmUzYjk3OTQyYjZjMTc5MTI0ZDU0MmY1M2M1Yjhh
+    YzU3OWRiYzQ3MDIyMzdmMTI1YmNmNjE1MmFjNGMwMGU2NGYyZDFlNmVlN2Rk
+    Y2FiMGVjNWU3MTExYjczNmIyZTViMWUzNDUyYjc3ZmJkYmExMzk=
+  data.tar.gz: !binary |-
+    NGI0ZGViZGFjNmM5NmEyMjYzZWQ4ZmVlNGQ2NjA4MjdkNzRlYzcwODVhMjFm
+    MTk0ZjQzNzAxMmE0N2U2NjQxZjU1MjFlNjE5M2QxNDY1ZDE4NWQ4MDkzYWVm
+    MDIyYjRjMjAwOTUyZjY4NjI2NzI4MGNkYzlkNGMzMGEwNjEzZGE=

data/.travis.yml

@@ -1,17 +1,12 @@
 language: ruby
 rvm:
  - 1.8.7
- - 1.9.2
  - 1.9.3
+ - 2.0.0
  - ruby-head
  - jruby
- - rbx
  - ree
 
 before_install:
  - sudo apt-get update -qq
  - sudo apt-get install -qq -y poppler-utils poppler-data
-
-matrix:
-  allow_failures:
-    - rvm: ruby-head

data/lib/pdf/info.rb

@@ -1,3 +1,4 @@
+require 'date' unless defined? DateTime
 require 'pdf/info/exceptions'
 
 module PDF
@@ -51,6 +52,8 @@ module PDF
       metadata = {}
       rows.each do |row|
         pair = row.split(':', 2)
+        pair.map!(&:strip)
+
         case pair.first
         when "Pages"
           metadata[:page_count] = pair.last.to_i
@@ -63,9 +66,11 @@ module PDF
         when "PDF version"
           metadata[:version] = pair.last.to_f
         when "CreationDate"
-          metadata[:creation_date] = DateTime.parse(pair.last)
+          creation_date = parse_datetime(pair.last)
+          metadata[:creation_date] = creation_date if creation_date
         when "ModDate"
-          metadata[:modification_date] = DateTime.parse(pair.last)
+          modification_date = parse_datetime(pair.last)
+          metadata[:modification_date] = modification_date if modification_date
         when /^Page.*size$/
           metadata[:pages] ||= []
           metadata[:pages] << pair.last.scan(/[\d.]+/).map(&:to_f)
@@ -78,5 +83,17 @@ module PDF
       metadata
     end
 
+    private
+
+    def parse_datetime(value)
+      DateTime.parse(value)
+    rescue
+      begin
+        DateTime.strptime(value, '%m/%d/%Y %k:%M:%S')
+      rescue
+        nil
+      end
+    end
+
   end
 end

data/lib/pdf/info/version.rb

@@ -1,6 +1,6 @@
 module PDF
   class Info
-    VERSION = "0.4.0"
+    VERSION = "0.5.0"
   end
 end
     

data/spec/pdf_info_spec.rb

@@ -87,6 +87,58 @@ describe PDF::Info do
     end
   end
 
+  describe ".process_output" do
+    subject do
+      PDF::Info.new('test.pdf')
+    end
+
+    it "symbolizes all keys" do
+      output = "a:foo\nb:bar\nc:baz"
+      [:a, :b, :c].each do |key|
+        expect(subject.process_output(output)).to have_key key
+        expect(subject.process_output(output)).to_not have_key key.to_s
+      end
+    end
+
+    it "downcases key" do
+      output = "I AM ALL CAPITAL:I STAY ALL CAPITAL"
+      expected = {:'i_am_all_capital' => 'I STAY ALL CAPITAL'}
+      expect(subject.process_output(output)).to include expected
+    end
+
+    it "replaces whitespace in key with underscore" do
+      output = "key with space:value without underscore"
+      expected = {:'key_with_space' => 'value without underscore'}
+      expect(subject.process_output(output)).to include expected
+    end
+
+    it "strips whitespace from metadata pair" do
+      output = "  key with space  :value without space\nkey without space:  value with space  "
+      expected = {:'key_with_space' => 'value without space', :'key_without_space' => 'value with space'}
+      expect(subject.process_output(output)).to include expected
+    end
+  end
+
+  describe ".parse_datetime" do
+    subject do
+      pdf_info = PDF::Info.new('test.pdf')
+      pdf_info.stub!(:command).and_return(output('successful.txt'))
+      pdf_info
+    end
+
+    it 'parse standard datetime format' do
+      expect(subject.send(:parse_datetime, '2001-02-03T04:05:06+07:00')).to be_kind_of DateTime
+    end
+
+    it 'parse american datetime format' do
+      expect(subject.send(:parse_datetime, '4/23/2004 18:37:34')).to be_kind_of DateTime
+    end
+
+    it 'return nil if string can not be parsed' do
+      expect(subject.send(:parse_datetime, 'asdf')).to be_nil
+    end
+  end
+
   describe "running on sample.pdf" do
     subject do
       PDF::Info.command_path = "pdfinfo"

metadata

@@ -1,20 +1,18 @@
 --- !ruby/object:Gem::Specification
 name: pdf_info
 version: !ruby/object:Gem::Version
-  version: 0.4.0
-  prerelease: 
+  version: 0.5.0
 platform: ruby
 authors:
 - Tom Taylor
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-01-20 00:00:00.000000000 Z
+date: 2013-05-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
@@ -22,7 +20,6 @@ dependencies:
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
@@ -30,7 +27,6 @@ dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
@@ -38,7 +34,6 @@ dependencies:
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
@@ -70,33 +65,26 @@ files:
 homepage: https://github.com/newspaperclub/pdf_info
 licenses:
 - MIT
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
   - - ! '>='
     - !ruby/object:Gem::Version
       version: '0'
-      segments:
-      - 0
-      hash: -1371290085031323010
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
   - - ! '>='
     - !ruby/object:Gem::Version
       version: '0'
-      segments:
-      - 0
-      hash: -1371290085031323010
 requirements: []
 rubyforge_project: pdf_info
-rubygems_version: 1.8.23
+rubygems_version: 2.0.3
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Wraps the pdfinfo command line tool to provide a hash of metadata
 test_files:
 - spec/assets/sample.pdf