thingfish 0.5.0.pre20161103181816 → 0.5.0

data/spec/thingfish/processor/{mp3_spec.rb → sha256_spec.rb}

@@ -8,14 +8,14 @@ require 'thingfish/processor'
 require 'strelka/httprequest/metadata'
 
 
-describe Thingfish::Processor, "MP3" do
+describe Thingfish::Processor, "SHA256" do
 
 	before( :all ) do
 		Strelka::HTTPRequest.class_eval { include Strelka::HTTPRequest::Metadata }
 	end
 
 
-	let( :processor ) { described_class.create(:mp3) }
+	let( :processor ) { described_class.create(:sha256) }
 
 	let( :factory ) do
 		Mongrel2::RequestFactory.new(
@@ -24,27 +24,26 @@ describe Thingfish::Processor, "MP3" do
 	end
 
 
-	it "extracts metadata from uploaded MP3 ID3 tags" do
+	it "generates a sha256 checksum from an uploaded file" do
 		req = factory.post( '/tf', fixture_data('APIC-1-image.mp3'), 'Content-type' => 'audio/mp3' )
-
 		processor.process_request( req )
 
-		expect( req.metadata ).to include( 'mp3:artist', 'mp3:bitrate', 'mp3:comments' )
+		expect( req.metadata['checksum'] ).to eq( 'e6b7070cbec90cdc2d8206819d86d100f076f480c9ae19d3eb8f878b3b86f2d6' )
 	end
 
 
-	it "attaches album art as a related resource" do
+	it "generates a sha256 checksum for related resources" do
+		content = "data data data data data"
 		req = factory.post( '/tf', fixture_data('APIC-1-image.mp3'), 'Content-type' => 'audio/mp3' )
+		req.add_related_resource( StringIO.new( content ), {} )
 
 		processor.process_request( req )
 
 		related = req.related_resources
-		expect( related.size ).to eq( 1 )
-		expect( related.values.first ).
-			to include( 'format' => 'image/jpeg', 'extent' => 7369, 'relationship' => 'album-art' )
-		expect( related.keys.first ).to respond_to( :read )
-	end
+		related_metadata = related.first.last
 
+		expect( related_metadata[:checksum] ).to eq( Digest::SHA256.hexdigest(content) )
+	end
 end
 
 # vim: set nosta noet ts=4 sw=4 ft=rspec:

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: thingfish
 version: !ruby/object:Gem::Version
-  version: 0.5.0.pre20161103181816
+  version: 0.5.0
 platform: ruby
 authors:
 - Michael Granger
@@ -36,7 +36,7 @@ cert_chain:
   w8aNA5re5+Rt/Vvjxj5AcEnZnZiz5x959NaddQocX32Z1unHw44pzRNUur1GInfW
   p4vpx2kUSFSAGjtCbDGTNV2AH8w9OU4xEmNz8c5lyoA=
   -----END CERTIFICATE-----
-date: 2016-11-04 00:00:00.000000000 Z
+date: 2016-11-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: strelka
@@ -58,14 +58,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.43'
+        version: '0.46'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.43'
+        version: '0.46'
 - !ruby/object:Gem::Dependency
   name: hoe-mercurial
   requirement: !ruby/object:Gem::Requirement
@@ -122,20 +122,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '0.7'
-- !ruby/object:Gem::Dependency
-  name: ruby-mp3info
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '0.8'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '0.8'
 - !ruby/object:Gem::Dependency
   name: rdoc
   requirement: !ruby/object:Gem::Requirement
@@ -182,6 +168,8 @@ extra_rdoc_files:
 - README.rdoc
 files:
 - ".simplecov"
+- ChangeLog
+- Gemfile
 - History.rdoc
 - LICENSE
 - Manifest.txt
@@ -190,6 +178,7 @@ files:
 - Rakefile
 - bin/tfprocessord
 - bin/thingfish
+- etc/mongrel2-config.rb
 - etc/thingfish.conf.example
 - lib/strelka/app/metadata.rb
 - lib/strelka/httprequest/metadata.rb
@@ -202,7 +191,7 @@ files:
 - lib/thingfish/metastore/memory.rb
 - lib/thingfish/mixins.rb
 - lib/thingfish/processor.rb
-- lib/thingfish/processor/mp3.rb
+- lib/thingfish/processor/sha256.rb
 - lib/thingfish/processordaemon.rb
 - lib/thingfish/spechelpers.rb
 - spec/data/APIC-1-image.mp3
@@ -217,7 +206,7 @@ files:
 - spec/thingfish/metastore/memory_spec.rb
 - spec/thingfish/metastore_spec.rb
 - spec/thingfish/mixins_spec.rb
-- spec/thingfish/processor/mp3_spec.rb
+- spec/thingfish/processor/sha256_spec.rb
 - spec/thingfish/processor_spec.rb
 - spec/thingfish_spec.rb
 homepage: http://bitbucket.org/ged/thingfish
@@ -238,9 +227,9 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 2.0.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">"
+  - - ">="
     - !ruby/object:Gem::Version
-      version: 1.3.1
+      version: '0'
 requirements: []
 rubyforge_project: 
 rubygems_version: 2.5.1

data/lib/thingfish/processor/mp3.rb

@@ -1,167 +0,0 @@
-# -*- ruby -*-
-#encoding: utf-8
-
-require 'mp3info'
-
-require 'thingfish' unless defined?( Thingfish )
-require 'thingfish/processor' unless defined?( Thingfish::Processor )
-
-
-# Attach ID3 data to an mp3, along with any embedded album art as a related resource.
-class Thingfish::Processor::MP3 < Thingfish::Processor
-	extend Loggability
-
-	# Loggability API -- log to the :thingfish logger
-	log_to :thingfish
-
-	# The list of handled types
-	handled_types 'audio/mpeg', 'audio/mpg', 'audio/mp3'
-
-
-	# Null character
-	NULL = "\x0"
-
-	# Attached picture   "PIC"
-	#    Frame size         $xx xx xx
-	#    ---- mp3info is 'helpfully' cropping out frame size.
-	#    Text encoding      $xx
-	#    Image format       $xx xx xx
-	#    Picture type       $xx
-	#    Description        <textstring> $00 (00)
-	#    Picture data       <binary data>
-	PIC_FORMAT = 'h a3 h Z* a*'
-
-	# Attached picture "APIC"
-	#    Text encoding   $xx
-	#    MIME type       <text string> $00
-	#    Picture type    $xx
-	#    Description     <text string according to encoding> $00 (00)
-	#    Picture data    <binary data>
-	APIC_FORMAT = 'h Z* h Z* a*'
-
-
-	### Synchronous processor API -- extract metadata from uploaded MP3s
-	def on_request( request )
-		mp3info = Mp3Info.new( request.body )
-
-		mp3_metadata = self.extract_id3_metadata( mp3info )
-		request.add_metadata( mp3_metadata )
-
-		self.extract_images( mp3info ) do |imageio, metadata|
-			metadata[:title] = "Album art for %s - %s" % mp3_metadata.values_at( 'mp3:artist', 'mp3:title' )
-			request.add_related_resource( imageio, metadata )
-		end
-	end
-
-
-	### Normalize metadata from the MP3Info object and return it as a hash.
-	def extract_id3_metadata( mp3info )
-		self.log.debug "Extracting MP3 metadata"
-
-		mp3_metadata = {
-			'mp3:frequency' => mp3info.samplerate,
-			'mp3:bitrate'   => mp3info.bitrate,
-			'mp3:vbr'       => mp3info.vbr,
-			'mp3:title'     => mp3info.tag.title,
-			'mp3:artist'    => mp3info.tag.artist,
-			'mp3:album'     => mp3info.tag.album,
-			'mp3:year'      => mp3info.tag.year,
-			'mp3:genre'     => mp3info.tag.genre,
-			'mp3:tracknum'  => mp3info.tag.tracknum,
-			'mp3:comments'  => mp3info.tag.comments,
-		}
-
-		# ID3V2 2.2.0 has three-letter tags, so map those if the artist info isn't set
-		if mp3info.hastag2?
-			if mp3_metadata['mp3:artist'].nil?
-				self.log.debug "   extracting old-style ID3v2 info" % [mp3info.tag2.version]
-
-				mp3_metadata.merge!({
-					'mp3:title'     => mp3info.tag2.TT2,
-					'mp3:artist'    => mp3info.tag2.TP1,
-					'mp3:album'     => mp3info.tag2.TAL,
-					'mp3:year'      => mp3info.tag2.TYE,
-					'mp3:tracknum'  => mp3info.tag2.TRK,
-					'mp3:comments'  => mp3info.tag2.COM,
-					'mp3:genre'     => mp3info.tag2.TCO,
-				})
-			end
-		end
-
-		self.log.debug "  raw metadata: %p" % [ mp3_metadata ]
-		return sanitize_values( mp3_metadata )
-	end
-
-
-	### Extract image data from ID3 information, supports both APIC (2.3) and the older style
-	### PIC (2.2).  Return value is a hash with IO keys and mimetype values.
-	### {
-	###    io  => { format => 'image/jpeg' }
-	###    io2 => { format => 'image/jpeg' }
-	### }
-	def extract_images( mp3info )
-		self.log.debug "Extracting embedded images"
-		raise LocalJumpError, "no block given" unless block_given?
-
-		unless mp3info.hastag2?
-			self.log.debug "...no id3v2 tag, so no embedded images possible."
-			return
-		end
-
-		self.log.debug "...id3v2 tag present..."
-
-		if mp3info.tag2.APIC
-			self.log.debug "...extracting APIC (id3v2.3+) image data."
-
-			images = [ mp3info.tag2.APIC ].flatten
-			images.each do |img|
-				blob, mime = img.unpack( APIC_FORMAT ).values_at( 4, 1 )
-				yield( StringIO.new(blob),
-					:format       => mime,
-					:extent       => blob.length,
-					:relationship => 'album-art' )
-			end
-
-		elsif mp3info.tag2.PIC
-			self.log.debug "...extracting PIC (id3v2.2) image data."
-
-			images = [ mp3info.tag2.PIC ].flatten
-			images.each do |img|
-				blob, type = img.unpack( PIC_FORMAT ).values_at( 4, 1 )
-				mime = Mongrel2::Config.mimetypes[ ".#{type.downcase}" ] or next
-				yield( StringIO.new(blob),
-					:format       => mime,
-					:extent       => blob.length,
-					:relationship => 'album-art' )
-			end
-
-		else
-			self.log.debug "...no known image tag types in tags: %p" % [ mp3info.tag2.keys.sort ]
-		end
-	end
-
-
-
-	#######
-	private
-	#######
-
-	### Strip NULLs from the values of the given +metadata_hash+ and return it.
-	def sanitize_values( metadata_hash )
-		metadata_hash.each do |k,v|
-			case v
-			when String
-				metadata_hash[k] = v.chomp(NULL).strip
-			when Array
-				metadata_hash[k] = v.collect {|vv| vv.chomp(NULL).strip }
-			when Numeric, TrueClass, FalseClass
-				# No-op
-			end
-		end
-
-		return metadata_hash.delete_if {|_,v| v.nil? }
-	end
-
-
-end # class Thingfish::Processor::MP3
-