rika 0.9.7-java → 0.9.8-java

data/lib/rika.rb

@@ -11,7 +11,7 @@ Dir[File.join(File.dirname(__FILE__), "../target/dependency/*.jar")].each do |ja
   require jar
 end
 
-# Heavily based on the Apache Tika API: http://tika.apache.org/1.2/api/org/apache/tika/Tika.html
+# Heavily based on the Apache Tika API: http://tika.apache.org/1.3/api/org/apache/tika/Tika.html
 module Rika
   import org.apache.tika.metadata.Metadata
   import org.apache.tika.Tika
@@ -24,14 +24,9 @@ module Rika
       @uri = file_location
       @tika = Tika.new
       @tika.set_max_string_length(max_content_length)
-      @metadata = Metadata.new
-
-      @is_file = File.exists?(@uri) && File.directory?(@uri) == false
-      is_http = URI(@uri).scheme == "http" && Net::HTTP.get_response(URI(@uri)).is_a?(Net::HTTPSuccess) if !@is_file
-      
-      if !@is_file && !is_http
-        raise IOError, "File does not exist or can't be reached."
-      end
+      @metadata_java = Metadata.new
+      @metadata_ruby = nil
+      @input_type = get_input_type
     end
 
     def content
@@ -40,14 +35,15 @@ module Rika
     end
 
     def metadata
-      self.parse
-      metadata_hash = {}
+      unless @metadata_ruby
+        self.parse
+        @metadata_ruby = {}
       
-      @metadata.names.each do |name|
-        metadata_hash[name] = @metadata.get(name) 
+        @metadata_java.names.each do |name|
+          @metadata_ruby[name] = @metadata_java.get(name)
+        end
       end
-
-      metadata_hash
+      @metadata_ruby
     end
 
     def media_type
@@ -55,25 +51,37 @@ module Rika
     end
 
     def available_metadata
-      self.parse
-      @metadata.names.to_a
+      metadata.keys
     end
 
     def metadata_exists?(name)
-      self.parse
-      @metadata.get(name) != nil
+      metadata[name] != nil
+    end
+
+    def file?
+      @input_type == :file
     end
 
     protected
     
     def parse
-      @content ||= @tika.parse_to_string(input_stream, @metadata).to_s.strip
+      @content ||= @tika.parse_to_string(input_stream, @metadata_java).to_s.strip
+    end
+
+    def get_input_type
+      if File.exists?(@uri) && File.directory?(@uri) == false
+        :file
+      elsif URI(@uri).scheme == "http" && Net::HTTP.get_response(URI(@uri)).is_a?(Net::HTTPSuccess)
+        :http
+      else
+        raise IOError, "Input (#{@uri}) is neither file nor http."
+      end
     end
 
     def input_stream
-      if @is_file
+      if file?
         FileInputStream.new(java.io.File.new(@uri))
-      else 
+      else # :http
         URL.new(@uri).open_stream
       end
     end

data/lib/rika/version.rb

@@ -1,3 +1,3 @@
 module Rika
-  VERSION = "0.9.7"
+  VERSION = "0.9.8"
 end

data/spec/rika_spec.rb

@@ -15,7 +15,7 @@ describe Rika::Parser do
     @dir = File.expand_path(File.join(File.dirname(__FILE__), 'fixtures'))  
     port = 50505
     @url = "http://#{Socket.gethostname}:#{port}"
-    @qoute = "First they ignore you, then they ridicule you, then they fight you, then you win."
+    @quote = "First they ignore you, then they ridicule you, then they fight you, then you win."
     @t1 = Thread.new do
       @server = HTTPServer.new(:Port => port, :DocumentRoot => @dir, 
       :AccessLog => [], :Logger => WEBrick::Log::new("/dev/null", 7))  
@@ -28,11 +28,11 @@ describe Rika::Parser do
   end
 
   it "should raise error if file does not exists" do
-    lambda { Rika::Parser.new(file_path("nonsense.txt")) }.should raise_error(IOError, "File does not exist or can't be reached.")
+    lambda { Rika::Parser.new(file_path("nonsense.txt")) }.should raise_error(IOError)
   end
 
   it "should raise error if URL does not exists" do
-    lambda { Rika::Parser.new("http://nonsense.com/whatever.pdf") }.should raise_error(IOError, "File does not exist or can't be reached.")
+    lambda { Rika::Parser.new("http://nonsense.com/whatever.pdf") }.should raise_error(IOError)
   end
 
   it "should detect file type without a file extension" do
@@ -41,20 +41,20 @@ describe Rika::Parser do
   end
 
   it "should not be possible to trick the parser to read a folder with an extension" do
-    lambda { Rika::Parser.new(file_path("folder.js")).content }.should raise_error(IOError, "File does not exist or can't be reached.")
+    lambda { Rika::Parser.new(file_path("folder.js")).content }.should raise_error(IOError)
   end
 
   describe '#content' do
     it "should return the content in a text file" do
-      @txt_parser.content.strip.should == @qoute
+      @txt_parser.content.strip.should == @quote
     end
 
     it "should return the content in a docx file" do
-      @docx_parser.content.should == @qoute
+      @docx_parser.content.should == @quote
     end
 
     it "should return the content in a pdf file" do 
-      @pdf_parser.content.should == @qoute
+      @pdf_parser.content.should == @quote
     end
 
     it "should return no content for an image" do
@@ -78,7 +78,7 @@ describe Rika::Parser do
 
     it "should return the content from a file over http" do
       parser = Rika::Parser.new(@url + "/document.pdf")
-      parser.content.should == @qoute   
+      parser.content.should == @quote
     end
 
     it "should return empty string for unknown file" do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: rika
 version: !ruby/object:Gem::Version
-  version: 0.9.7
+  version: 0.9.8
   prerelease:
 platform: java
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2013-01-22 00:00:00.000000000 Z
+date: 2013-01-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec