diarize-ruby 0.3.6 → 0.3.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 36e00b3c5b5a169cc746f0cf140e3d1fbb5e2055
-  data.tar.gz: d082d84e13a307504f5fac7916549d3d201a5415
+  metadata.gz: acabfb7160ffeb0ee59815441c7108aebd3dbc57
+  data.tar.gz: 1e1ac96bdeb06c3546ef652f1b8324f38c42316f
 SHA512:
-  metadata.gz: 0d44decde02e7d5f74e4dac4247c9e2ff1f5d95f169112f1b34108c3e510be74f68d393705135e76d93719da6ea09539ff9ee46aed235d82c5fe6e3e2909acaf
-  data.tar.gz: f15c9b16585f52566e532d8417b23cb946629fda36db1d0b6de189cb6be0004a5893c0816a34b1acd636c63bde375e237e9871697001116cae3694e08ef207e0
+  metadata.gz: f453a319e2d7e4084fa95fa33240a78773bcf6d7dddf8551b02337a5a9ed294ebc48317c52ba2d845fb2f095f86397038e0759124aa4409721a1f3d5cf2f7786
+  data.tar.gz: 6be77004bd7ba178fba11f5b3e8510357f7c2708b5fc49bfc302a570af8b6e0727603cd776373e157c1227e7e354b1abf3bec63810d87f5df7b72a7148e99dbc

data/CHANGELOG.md

@@ -1,3 +1,10 @@
+## [v0.3.7] - 2016-11-18
+
+- Sort segments by start time by default
+- Allow to pass audio file names to `Diarize::Audio.new("~/foo.wav")`
+- Download https URLs without certs checking
+- Fix `diarize` druby uri for remote audio
+
 ## [v0.3.6] - 2016-11-16
 
 - Add pidfile option to diarize server command (--pidfile, -P)

data/README.md

@@ -71,8 +71,8 @@ Some Java implementations (i.e. OpenJDK on Linux) are causing trouble running [R
 Start the diarizer in a separate process as a server:
 
     $ diarize server -P 9999 -H localhost
-    Drb server
-    diarize-ruby 0.3.4
+    DRb server
+    diarize-ruby x.y.z
     Listening on druby://localhost:9999, CTRL+C to stop
 
 ### Client
@@ -91,8 +91,7 @@ server_uri = "druby://localhost:9999"
 DRb.start_service
 client = DRbObject.new_with_uri(server_uri)
 
-audio_uri = URI.join('file:///', File.join(File.expand_path(File.dirname(__FILE__)), "test", "data", "will-and-juergen.wav"))
-audio = client.build_audio(audio_uri)
+audio = client.build_audio(File.join("test", "data", "will-and-juergen.wav"))
 audio.analyze!
 audio.segments
 ...

data/bin/diarize

@@ -3,39 +3,21 @@ require "diarize"
 require "gli"
 require "uri"
 require "drb/drb"
-require "byebug"
 
 include GLI::App
 
-def uri_from_args(args)
-  url_or_file_name = args.first
-  uri = URI.parse(url_or_file_name)
-  if uri.scheme && uri.scheme.match(/^(http|https|file)$/)
-    uri = url_or_file_name
-  else
-    uri = if url_or_file_name[0] == "/"
-      URI.join('file:///', url_or_file_name)
-    else
-      URI.join('file:///', File.join(File.expand_path(Dir.pwd), url_or_file_name))
-    end
-  end
-  uri
-end
-
 def build_audio_from_args(args, options = {})
-  uri = uri_from_args(args)
-  $stdout.puts uri.to_s if options[:verbose]
-  audio = Diarize::Audio.new(uri)
+  $stdout.puts args.first.to_s if options[:verbose]
+  audio = Diarize::Audio.new(args.first)
   audio.analyze!
   audio
 end
 
 def build_remote_audio_from_args(args, options = {})
   remote = client(options)
-  audio_uri = uri_from_args(args)
 
-  $stdout.puts uri.to_s if options[:verbose]
-  audio = remote.build_audio(audio_uri)
+  $stdout.puts args.first.to_s if options[:verbose]
+  audio = remote.build_audio(args.first)
   audio.analyze!
   audio
 rescue DRb::DRbConnError => ex
@@ -76,7 +58,7 @@ command [:remote, :r] do |remote|
     audio.arg_name 'FILENAME', :multiple
     audio.command [:se, :segment, :segments] do |segments|
       segments.action do |global_options, options, args|
-        audio = build_remote_audio_from_args(args, options.merge(global_options))
+        audio = build_remote_audio_from_args(args, options.first.last.first.last)
         audio.segments.each do |segment|
           $stdout.puts segment.uri
         end if audio
@@ -87,13 +69,25 @@ command [:remote, :r] do |remote|
     audio.arg_name 'FILENAME', :multiple
     audio.command [:sp, :speaker, :speakers] do |speakers|
       speakers.action do |global_options, options, args|
-        audio = build_remote_audio_from_args(args, options.merge(global_options))
+        audio = build_remote_audio_from_args(args, options.first.last.first.last)
         audio.speakers.each do |speaker|
           $stdout.puts speaker.uri
         end if audio
       end
     end
   end
+
+  #-h, --host HOSTNAME
+  remote.desc 'Host, e.g. "localhost"'
+  remote.default_value 'localhost'
+  remote.arg_name 'HOSTNAME'
+  remote.flag [:h, :host]
+
+  #-p, --port PORT
+  remote.desc 'Port number'
+  remote.default_value 9999
+  remote.arg_name 'PORT'
+  remote.flag [:p, :port]
 end
 
 desc 'audio file'

data/lib/diarize/audio.rb

@@ -2,22 +2,30 @@ module Diarize
   class Audio
     attr_reader :path, :file, :uri
 
-    def initialize(url_or_uri)
-      @uri = url_or_uri.is_a?(String) ? URI(url_or_uri) : url_or_uri
-      if uri.scheme == 'file'
-        # Local file
+    def initialize(uri_url_or_file_name)
+      if uri_url_or_file_name.is_a?(URI)
+        @uri = uri_url_or_file_name
+      elsif uri_url_or_file_name.is_a?(String)
+        # url or file name
+        @uri = URI.parse(uri_url_or_file_name)
+        if @uri.scheme && @uri.scheme.match(/^(http|https|file)$/)
+          # url or file:/// uri, do nothing
+        else
+          @uri = URI.join('file:///', File.join(File.expand_path(uri_url_or_file_name)))
+        end
+      end
+
+      if @uri.scheme == 'file'
         @path = uri.path
       else
-        # Remote file, we get it locally
-        @path = '/tmp/' + Digest::MD5.hexdigest(uri.to_s)
-        File.open(@path, "wb") {|f| f << open(uri).read }
+        # remote file, we download it locally
+        @path = '/tmp/' + Digest::MD5.hexdigest(@uri.to_s)
+        File.open(@path, "wb") {|f| f << open(@uri, {ssl_verify_mode: OpenSSL::SSL::VERIFY_NONE}).read}
       end
 
-      if !File.exist?(@path)
-        raise "Unable to locate: #{@path}.  Check that the file is available at #{uri.inspect}."
-      end
+      raise "Unable to locate '#{@path}' from '#{@uri.inspect}'." unless File.exist?(@path)
 
-      @file = File.new @path
+      @file = File.new(@path)
     end
 
     def analyze!(train_speaker_models = true)
@@ -37,7 +45,7 @@ module Diarize
       #parameter.parameterDiarization.cEClustering = true # We use CE clustering by default
       parameter.parameterInputFeature.setFeatureMask(@path)
       @clusters = ester2(parameter)
-      @segments = Segmentation.from_clusters(self, @clusters)
+      @segments = Segmentation.from_clusters(self, @clusters).sort_by(&:start)
       train_speaker_gmms if train_speaker_models
     end
 
@@ -47,7 +55,7 @@ module Diarize
     end
 
     def segments
-      raise Exception.new('You need to run analyze! before being able to access the analysis results') unless @segments
+      raise RuntimeError, "You need to run analyze! before being able to access the analysis results" unless @segments
       @segments
     end
 

data/lib/diarize/segment.rb

@@ -5,11 +5,11 @@ module Diarize
     attr_reader :start, :duration, :gender, :bandwidth
 
     def initialize(audio, start, duration, gender, bandwidth, speaker_id)
-      @audio = audio
-      @start = start
-      @duration = duration
-      @bandwidth = bandwidth
-      @speaker_id = speaker_id
+      @audio          = audio
+      @start          = start
+      @duration       = duration
+      @bandwidth      = bandwidth
+      @speaker_id     = speaker_id
       @speaker_gender = gender
     end
 

data/lib/diarize/segmentation.rb

@@ -5,12 +5,12 @@ module Diarize
       segmentation = []
       File.open(seg_file).each_line do |line|
         next if line.start_with? ';;'
-        parts      = line.split(' ')
-        start      = parts[2].to_i / 100.0
-        duration   = parts[3].to_i / 100.0
-        gender     = parts[4]
-        bandwidth  = parts[6]
-        speaker_id = parts[7]
+        parts        = line.split(' ')
+        start        = parts[2].to_i / 100.0
+        duration     = parts[3].to_i / 100.0
+        gender       = parts[4]
+        bandwidth    = parts[6]
+        speaker_id   = parts[7]
         segmentation << Segment.new(audio, start, duration, gender, bandwidth, speaker_id)
       end
       segmentation
@@ -19,9 +19,9 @@ module Diarize
     def self.from_clusters(audio, clusters)
       segmentation = []
       clusters.map(&:to_s).each do |speaker_id|
-        cluster = clusters.getCluster(speaker_id)
-        gender = cluster.gender
-        bandwidth = cluster.bandwidth
+        cluster        = clusters.getCluster(speaker_id)
+        gender         = cluster.gender
+        bandwidth      = cluster.bandwidth
         cluster.each do |segment|
           start        = segment.start_in_second
           duration     = segment.length_in_second

data/lib/diarize/version.rb

@@ -1,3 +1,3 @@
 module Diarize
-  VERSION = "0.3.6"
+  VERSION = "0.3.7"
 end

data/lib/diarize.rb

@@ -7,10 +7,10 @@ Rjb::load(RJB_LOAD_PATH, RJB_OPTIONS)
 
 require "matrix"
 
-require "to_rdf"
 require "uri"
 require "open-uri"
 require "digest/md5"
+require "to_rdf"
 
 require "diarize/version"
 require "diarize/audio"

data/test/audio_test.rb

@@ -43,8 +43,8 @@ class AudioTest < Test::Unit::TestCase
     audio.clean!
   end
 
-  def test_segments_raises_exception_when_audio_is_not_analysed
-    assert_raise Exception do
+  def test_segments_raises_exception_when_audio_is_not_analyzed
+    assert_raise RuntimeError do
       @audio.segments
     end
   end

data/test/version_test.rb

@@ -3,7 +3,7 @@ require 'test_helper'
 class VersionTest < Test::Unit::TestCase
 
   def test_current_version
-    assert_equal "0.3.6", Diarize::VERSION
+    assert_equal "0.3.7", Diarize::VERSION
   end
 
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: diarize-ruby
 version: !ruby/object:Gem::Version
-  version: 0.3.6
+  version: 0.3.7
 platform: ruby
 authors:
 - Yves Raimond