speech2text 0.01 → 0.3.0

data/README.rdoc

@@ -1,10 +1,23 @@
-== Speech2Text
+= Speech2Text
 
 Using the power of ffmpeg/flac/Google and ruby here is a simple interface to play with to convert speech to text.
 
-At this point the API from Google is not documented and seemly free.
-The Google API will frequently return 500 errors without providing much reason as to why.
+Using a new undocumentd speech API from Google with the help of this article: http://mikepultz.com/2011/03/accessing-google-speech-api-chrome-11/
 
-It's possible that Google will decide to not open this API up and this effort my completely be for not...
+We're able to provide a very simple API in Ruby to decode simple audio to text.
 
-This was all made possible in short order all thanks to Chrome 11 and http://mikepultz.com/2011/03/accessing-google-speech-api-chrome-11/
+The API from Google is not yet public and so may change. It also seems to be very fragile as more times than not it will return a 500, so the library has retry code built in - for larger audio files 10+ failures may return before a successful result is retrieved...
+
+It also appears that the API only likes smaller audio files so there is a built in chunker that allows us to split the audio up into smaller chunks.
+
+== Example
+
+ audio = Speech::AudioToText.new("i-like-pickles.wav")
+ puts audio.to_text.inspect
+ => {"captured_json"=>[["I like pickles", 0.92731786]], "confidence"=>0.92731786}
+
+== Command Line
+
+ speech2text i-like-pickles.wav
+ cat i-like-pickles.json
+ {"captured_json"=>[["I like pickles", 0.92731786]], "confidence"=>0.92731786}

data/bin/speech2text

@@ -1,2 +1,7 @@
 #!/this/will/be/replaced/by/rubygems
 # -*- encoding: binary -*-
+
+require 'speech'
+
+captured_json = Speech::AudioToText.new(ARGV[0]).to_text
+puts captured_json.inspect

data/lib/speech.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 require 'curb'
 require 'json'
 

data/lib/speech/audio_inspector.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 module Speech
 
   class AudioInspector

data/lib/speech/audio_splitter.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 module Speech
 
   class AudioSplitter
@@ -8,7 +9,7 @@ module Speech
 
       def initialize(splitter, offset, duration)
         self.offset = offset
-        self.chunk = "chunk-" + splitter.original_file.gsub(/\.(.*)$/, "-#{offset}" + '.\1') 
+        self.chunk = File.join(File.dirname(splitter.original_file), "chunk-" + File.basename(splitter.original_file).gsub(/\.(.*)$/, "-#{offset}" + '.\1'))
         self.duration = duration
         self.splitter = splitter
       end

data/lib/speech/audio_to_text.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 module Speech
 
   class AudioToText
@@ -21,6 +22,10 @@ module Speech
       JSON.parse(File.read(self.captured_file))
     end
 
+    def clean
+      File.unlink self.captured_file if self.captured_file && File.exist?(self.captured_file)
+    end
+
   protected
 
     def convert_chunk(easy, chunk, options={})
@@ -29,6 +34,7 @@ module Speech
       while retrying
         #easy.verbose = true
         easy.headers['Content-Type'] = "audio/x-flac; rate=#{chunk.flac_rate}"
+        easy.headers['User-Agent'] = "https://github.com/taf2/speech2text"
         easy.post_body = "Content=#{chunk.to_flac_bytes}"
         easy.on_progress {|dl_total, dl_now, ul_total, ul_now| printf("%.2f/%.2f\r", ul_now, ul_total); true }
         easy.on_complete {|easy| puts }

data/lib/speech/version.rb

@@ -1,5 +1,6 @@
+# -*- encoding: binary -*-
 module Speech
   class Info
-    VERSION='0.01'
+    VERSION='0.3.0'
   end
 end

data/speech2text.gemspec

@@ -2,14 +2,15 @@ $:.unshift File.expand_path(File.dirname(__FILE__) + "/lib")
 require "speech/version"
 
 Gem::Specification.new do |s|
-  s.name    = "speech2text"
-  s.authors = ["Todd A. Fisher"]
-  s.email   = "todd.fisher@gmail.com"
-  s.version = Speech::Info::VERSION
-  s.homepage = "https://github.com/taf2/speech2text"
-  s.summary  = "Speech to Text Library"
-  s.description = "Super powers of Google wrapped in a nice Ruby interface"
-  s.files = Dir["{lib,bin,test}/**/*", "Rakefile", "README.rdoc", "*.gemspec"]
+  s.name           = "speech2text"
+  s.authors        = ["Todd A. Fisher"]
+  s.email          = "todd.fisher@gmail.com"
+  s.version        = Speech::Info::VERSION
+  s.homepage       = "https://github.com/taf2/speech2text"
+  s.summary        = "Speech to Text Library"
+  s.description    = "Super powers of Google wrapped in a nice Ruby interface"
+  s.files          = Dir["{lib,bin,test}/**/*", "Rakefile", "README.rdoc", "*.gemspec"]
+  s.executables    = %w(speech2text)
 
   s.add_dependency "curb"
   s.add_dependency "json"

data/test/audio_inspector_test.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 require 'test/unit'
 $:.unshift File.expand_path(File.dirname(__FILE__) + '/../lib')
 require 'speech'

data/test/audio_splitter_test.rb

@@ -1,3 +1,4 @@
+# -*- encoding: binary -*-
 require 'test/unit'
 $:.unshift File.expand_path(File.dirname(__FILE__) + '/../lib')
 require 'speech'
@@ -5,7 +6,7 @@ require 'speech'
 class SpeechAudioSplitterTest < Test::Unit::TestCase
 
   def test_audio_splitter
-    splitter = Speech::AudioSplitter.new("i-like-pickles.wav", 1)
+    splitter = Speech::AudioSplitter.new("samples/i-like-pickles.wav", 1)
 
     assert_equal '00:00:03:52', splitter.duration.to_s
     assert_equal 3.52, splitter.duration.to_f

data/test/audio_to_text_test.rb

@@ -0,0 +1,21 @@
+# -*- encoding: binary -*-
+require 'test/unit'
+$:.unshift File.expand_path(File.dirname(__FILE__) + '/../lib')
+require 'speech'
+
+class SpeechAudioToTextTest < Test::Unit::TestCase
+  def test_audio_to_text
+    audio = Speech::AudioToText.new("samples/i-like-pickles.wav")
+    captured_json = audio.to_text
+    assert captured_json
+    assert captured_json.key?("captured_json")
+    assert !captured_json['captured_json'].empty?
+    assert_equal ['captured_json', 'confidence'], captured_json.keys.sort
+    assert_equal "I like pickles", captured_json['captured_json'].flatten.first
+    assert captured_json['confidence'] > 0.9
+#    {"captured_json"=>[["I like pickles", 0.92731786]], "confidence"=>0.92731786}
+#    puts captured_json.inspect
+  ensure
+    audio.clean
+  end
+end

data/test/samples/i-like-pickles.json

@@ -0,0 +1 @@
+{"captured_json":[["I like pickles",0.92731786]],"confidence":0.92731786}

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: speech2text
 version: !ruby/object:Gem::Version
-  version: '0.01'
+  version: 0.3.0
   prerelease: 
 platform: ruby
 authors:
@@ -9,12 +9,12 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-03-24 00:00:00.000000000 -04:00
+date: 2011-03-25 00:00:00.000000000 -04:00
 default_executable: 
 dependencies:
 - !ruby/object:Gem::Dependency
   name: curb
-  requirement: &2157005460 !ruby/object:Gem::Requirement
+  requirement: &2157005180 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -22,10 +22,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *2157005460
+  version_requirements: *2157005180
 - !ruby/object:Gem::Dependency
   name: json
-  requirement: &2157005040 !ruby/object:Gem::Requirement
+  requirement: &2157004740 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -33,24 +33,25 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *2157005040
+  version_requirements: *2157004740
 description: Super powers of Google wrapped in a nice Ruby interface
 email: todd.fisher@gmail.com
-executables: []
+executables:
+- speech2text
 extensions: []
 extra_rdoc_files: []
 files:
 - lib/speech/audio_inspector.rb
 - lib/speech/audio_splitter.rb
 - lib/speech/audio_to_text.rb
-- lib/speech/text.rb
 - lib/speech/version.rb
 - lib/speech.rb
 - bin/speech2text
 - test/audio_inspector_test.rb
 - test/audio_splitter_test.rb
-- test/i-like-pickles.wav
+- test/audio_to_text_test.rb
 - test/SampleAudio.wav
+- test/samples/i-like-pickles.json
 - test/samples/i-like-pickles.wav
 - Rakefile
 - README.rdoc

data/lib/speech/text.rb

@@ -1,11 +0,0 @@
-module Speech
-  class Text
-    
-    def initialize(audio_file, options={})
-    end
-
-    def decode_audio(flac16k_audio)
-    end
-
-  end
-end