rflow-components-file 0.0.5 → 0.0.6

data/Rakefile

@@ -1,6 +1,6 @@
 require 'bundler'
 require 'rspec/core/rake_task'
-require 'rake/rdoctask'
+require 'rdoc/task'
 Bundler::GemHelper.install_tasks
 
 RSpec::Core::RakeTask.new(:spec) do |t|
@@ -13,4 +13,3 @@ Rake::RDocTask.new do |rd|
   rd.rdoc_files.include("README", "lib/**/*.rb")
   rd.rdoc_dir = File.join('doc', 'html')
 end
-

data/lib/rflow/components/file/directory_watcher.rb

@@ -8,7 +8,7 @@ class RFlow
       class DirectoryWatcher < RFlow::Component
         output_port :file_port
         output_port :raw_port
-        
+
         DEFAULT_CONFIG = {
           'directory_path'  => '/tmp/import',
           'file_name_glob'  => '*',
@@ -18,7 +18,7 @@ class RFlow
         }
 
         attr_accessor :config, :poll_interval, :directory_path, :file_name_glob, :remove_files
-        
+
         def configure!(config)
           @config = DEFAULT_CONFIG.merge config
           @directory_path  = ::File.expand_path(@config['directory_path'])
@@ -26,11 +26,11 @@ class RFlow
           @poll_interval   = @config['poll_interval'].to_i
           @files_per_poll  = @config['files_per_poll'].to_i
           @remove_files    = to_boolean(@config['remove_files'])
-          
+
           unless ::File.directory?(@directory_path)
             raise ArgumentError, "Invalid directory '#{@directory_path}'"
           end
-          
+
           unless ::File.readable?(@directory_path)
             raise ArgumentError, "Unable to read from directory '#{@directory_path}'"
           end
@@ -51,7 +51,7 @@ class RFlow
               RFlow.logger.debug "Importing #{file_path}"
               ::File.open(file_path, 'r:BINARY') do |file|
                 file_content = file.read
-                
+
                 RFlow.logger.debug "read #{file_content.bytesize} bytes of #{file.size} in #{file.path}, md5 #{Digest::MD5.hexdigest(file_content)}"
 
                 file_message = RFlow::Message.new('RFlow::Message::Data::File')
@@ -62,7 +62,7 @@ class RFlow
                 file_message.data.creation_timestamp = file.ctime
                 file_message.data.modification_timestamp = file.mtime
                 file_message.data.access_timestamp = file.atime
-                
+
                 file_port.send_message file_message
 
                 raw_message = RFlow::Message.new('RFlow::Message::Data::Raw')
@@ -88,7 +88,7 @@ class RFlow
             raise ArgumentError, "'#{string}' cannot be coerced to a boolean value"
           end
         end
-        
+
       end
     end
   end

data/lib/rflow/components/file/extensions.rb

@@ -1,10 +1,10 @@
 class RFlow
   module Components
     module File
-      
+
       # The set of extensions to add capability to File data types
       module Extensions
-        
+
         # Need to be careful when extending to not clobber data already in data_object
         module FileExtension
           def self.extended(base_data)
@@ -13,7 +13,7 @@ class RFlow
               'creation_timestamp' => nil, 'modification_timestamp' => nil, 'access_timestamp' => nil
             }
           end
-          
+
           # Default/string accessors
           ['path', 'content'].each do |name|
             define_method name do |*args|
@@ -23,7 +23,7 @@ class RFlow
               data_object[name] = args.first
             end
           end
-          
+
           # Integer Accessors
           ['size'].each do |name|
             define_method name do |*args|
@@ -33,7 +33,7 @@ class RFlow
               data_object[name] = args.first.to_i
             end
           end
-          
+
           # Timestamp Accessors.  Note, the precision of the
           # XMLTimestamp is set to 9 digits, meaning that the time you
           # put in might be slightly different from the time you read

data/lib/rflow/components/file/output_raw_to_files.rb

@@ -11,36 +11,36 @@ class RFlow
 
         DEFAULT_CONFIG = {
           'directory_path'  => '/tmp',
-          'file_name_prefix' => 'output.', 
-          'file_name_suffix' => '.out', 
+          'file_name_prefix' => 'output.',
+          'file_name_suffix' => '.out',
         }
-        
+
         attr_accessor :config, :directory_path, :file_name_prefix, :file_name_suffix
 
-        
+
         def configure!(config)
           @config = DEFAULT_CONFIG.merge config
           @directory_path  = ::File.expand_path(@config['directory_path'])
           @file_name_prefix = @config['file_name_prefix']
           @file_name_suffix = @config['file_name_suffix']
-          
+
           unless ::File.directory?(@directory_path)
             raise ArgumentError, "Invalid directory '#{@directory_path}'"
           end
-          
+
           unless ::File.writable?(@directory_path)
             raise ArgumentError, "Unable to read from directory '#{@directory_path}'"
           end
 
           @output_file_entropy = 0
-          
+
           # TODO: more error checking of input config
         end
 
-        
+
         def process_message(input_port, input_port_key, connection, message)
           return unless message.data_type_name == 'RFlow::Message::Data::Raw'
-          
+
           @output_file_entropy = 0
           begin
             final_output_file_name = output_file_name
@@ -49,8 +49,8 @@ class RFlow
             final_output_file_path = ::File.join(directory_path, "#{final_output_file_name}")
 
             RFlow.logger.debug "#{self.class.name}##{__method__}: Outputting raw message to #{final_output_file_path} (via #{temp_output_file_path}) with #{message.data.raw.bytesize} bytes and md5 #{Digest::MD5.hexdigest message.data.raw}"
-            
-            ::File.open(temp_output_file_path, ::File::CREAT|::File::EXCL|::File::RDWR, 0644) do |file|
+
+            ::File.open(temp_output_file_path, ::File::CREAT|::File::EXCL|::File::RDWR, 0644, :external_encoding => 'BINARY') do |file|
               file.flock(::File::LOCK_EX)
               file.write(message.data.raw)
             end
@@ -60,30 +60,30 @@ class RFlow
             RFlow.logger.debug("#{self.class.name}##{__method__}: File #{temp_output_file_path} exists, increasing entropy")
             retry
           end
-          
+
           final_output_file_path
         end
 
-        
+
         private
 
-        
-        def output_file_name 
+
+        def output_file_name
           "#{file_name_prefix}#{current_timestamp}-#{output_file_entropy}#{file_name_suffix}"
         end
 
-        
+
         def output_file_entropy
           @output_file_entropy += 1
           sprintf("%04d", @output_file_entropy)
         end
 
-        
+
         def current_timestamp
           time = Time.now
           time.utc.strftime("%Y%m%d_%H%M%S.") + "%06d" % time.utc.usec
         end
-        
+
       end
     end
   end

data/lib/rflow/components/file/version.rb

@@ -1,7 +1,7 @@
 class RFlow
   module Components
     module File
-      VERSION = "0.0.5"
+      VERSION = "0.0.6"
     end
   end
 end

data/lib/rflow/components/file.rb

@@ -7,11 +7,11 @@ class RFlow
     module File
       # Load the schemas
       SCHEMA_DIRECTORY = ::File.expand_path(::File.join(::File.dirname(__FILE__), '..', '..', '..', 'schema'))
-      
+
       SCHEMA_FILES = {
         'file.avsc' => 'RFlow::Message::Data::File',
       }
-      
+
       SCHEMA_FILES.each do |file_name, data_type_name|
         schema_string = ::File.read(::File.join(SCHEMA_DIRECTORY, file_name))
         RFlow::Configuration.add_available_data_type data_type_name, 'avro', schema_string

data/rflow-components-file.gemspec

@@ -9,7 +9,7 @@ Gem::Specification.new do |s|
   s.required_ruby_version = '~> 1.9'
   s.authors     = ["Michael L. Artz"]
   s.email       = ["michael.artz@redjack.com"]
-  s.homepage    = ""
+  s.homepage    = "https://github.com/redjack/rflow-components-file"
   s.summary     = %q{Components that operate on files for the RFlow FBP framework}
   s.description = %q{Components that operate on files for the RFlow FBP framework.  Also includes the File schema}
 
@@ -21,7 +21,6 @@ Gem::Specification.new do |s|
   s.require_paths = ["lib"]
 
   s.add_dependency 'rflow', '~> 0.0'
-  s.add_dependency 'eventmachine_httpserver', '~> 0.2'
 
   s.add_development_dependency 'rspec', '~> 2.6'
   s.add_development_dependency 'rake', '~> 0.8'

data/spec/extensions_spec.rb

@@ -10,7 +10,7 @@ describe RFlow::Components::File::Extensions::FileExtension do
   it "should add the extension to RFlow::Configuration" do
     RFlow::Configuration.available_data_extensions['RFlow::Message::Data::File'].should include(described_class)
   end
-  
+
   it "should set the defaults" do
     file = RFlow::Message.new('RFlow::Message::Data::File')
 
@@ -45,5 +45,5 @@ describe RFlow::Components::File::Extensions::FileExtension do
     file.data.creation_timestamp.should == Time.xmlschema(now.xmlschema)
   end
 
-  
+
 end

data/spec/output_raw_to_files_spec.rb

@@ -14,10 +14,10 @@ describe RFlow::Components::File::OutputRawToFiles do
 
     message = RFlow::Message.new('RFlow::Message::Data::Raw')
     message.data.raw = 'boomertown'
-    
+
     output_file_path = component.process_message nil, nil, nil, message
 
     File.exist?(output_file_path).should be_true
   end
-  
+
 end

data/spec/schema_spec.rb

@@ -4,7 +4,7 @@ describe 'RFlow::Message::Data::File Avro Schema' do
   before(:each) do
     @schema_string = RFlow::Configuration.available_data_types['RFlow::Message::Data::File']['avro']
   end
-  
+
   it "should encode and decode an object" do
     file = {
       'path' => '/full/file/path/filename',

data/spec/spec_helper.rb

@@ -3,7 +3,7 @@ require File.expand_path(File.join(File.dirname(__FILE__), '..', 'lib', 'rflow-c
 require 'logger'
 
 RFlow.logger = Logger.new STDOUT
-  
+
 def decode_avro(schema_string, serialized_object)
   schema = Avro::Schema.parse(schema_string)
   sio = StringIO.new(serialized_object)

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: rflow-components-file
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-16 00:00:00.000000000 Z
+date: 2014-03-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rflow
@@ -27,22 +27,6 @@ dependencies:
     - - ~>
       - !ruby/object:Gem::Version
         version: '0.0'
-- !ruby/object:Gem::Dependency
-  name: eventmachine_httpserver
-  requirement: !ruby/object:Gem::Requirement
-    none: false
-    requirements:
-    - - ~>
-      - !ruby/object:Gem::Version
-        version: '0.2'
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    none: false
-    requirements:
-    - - ~>
-      - !ruby/object:Gem::Version
-        version: '0.2'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -99,7 +83,7 @@ files:
 - spec/output_raw_to_files_spec.rb
 - spec/schema_spec.rb
 - spec/spec_helper.rb
-homepage: ''
+homepage: https://github.com/redjack/rflow-components-file
 licenses: []
 post_install_message: 
 rdoc_options: []