youtube-captions 0.0.4 → 1.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8d883289c61993364fb20ad953798e0821ca349e
-  data.tar.gz: f755a93a93e8796e9db143604c271bb423e614e2
+  metadata.gz: b3a75ebb25b8d3a61d08a95226072c4f1dec17a7
+  data.tar.gz: c5256c04472b80bf77ac4e56e95c63358040b431
 SHA512:
-  metadata.gz: 06d90c1d253adaf1a2eb97d48fecf7e0c30ccbc4529973384ea160b4db8d48ff2645bc3f4108c712ab0200906fd7d6884c3aeecea720a8a3a9f57f883f57edc7
-  data.tar.gz: 33ec9654434a67ab7073e479fd987eb077c997ee95dc6cefe3a845f7f2b9116be5a38d58d798ed75bb36d3dbcaf69dad74b94e285b0eafbd6e3839b2b8e38d0f
+  metadata.gz: 75eb49ab43023fe2f0d4ff291a5a53fc853d0ff567df929be831406230867913d6232249ce01e8bd98540b99ac86e0d6fc9c8f60299e0505f52ade2d57adb486
+  data.tar.gz: a3652d02e585ff86b0baf97a8979b7f1b0ef594b4f88c96fcf2a09032b551d8711f161f85e6df288837aa1f8521b46da9a19ccb762cbba59e5b882d9a39639be

data/lib/youtube-captions/captions.rb

@@ -0,0 +1,49 @@
+require 'httparty'
+require 'cgi'
+
+module YoutubeCaptions
+  class Captions
+    include HTTParty
+
+    attr_reader :info, :lang
+    def initialize(info:, lang:)
+      @info = info
+      @lang = lang
+    end
+
+    def call
+      if lang.nil?
+        lang_info = default_lang_info
+      else
+        lang_info = search_lang_info
+      end
+
+      return raise LangNotAvailableError.new("Lang no available") unless lang_info_has_base_url?(lang_info)
+
+      response = self.class.get(lang_info["baseUrl"])
+      captions = response["transcript"]["text"]
+
+      clean_captions(captions)
+    end
+
+    private
+
+    def default_lang_info
+      info.find {|json| json["kind"] == "asr"} || info.first
+    end
+
+    def search_lang_info
+      info.find { |json| json["vssId"] == ".#{lang}"} || info.find {|json| json["vssId"] == "a.#{lang}" }
+    end
+
+    def lang_info_has_base_url?(lang_info)
+      lang_info && lang_info["baseUrl"]
+    end
+
+    def clean_captions(captions)
+      captions.map do |caption|
+        caption.tap { |caption_hash| caption_hash["__content__"] = CGI.unescapeHTML(caption_hash["__content__"]).split.join(" ") }
+      end
+    end
+  end
+end

data/lib/youtube-captions/constants.rb

@@ -0,0 +1,11 @@
+module YoutubeCaptions
+  YOUTUBE_VIDEO_URL   = "https://www.youtube.com/watch?v="
+  CAPTIONABLE_REGEX  = /(\[{"baseUrl":.*"trackName":"(.*?)"}\])/
+  YOUTUBE_URL_FORMATS = [
+    %r((?:https?://)?youtu\.be/(.+)),
+    %r((?:https?://)?(?:www\.)?youtube\.com/watch\?v=(.*?)(&|#|$)),
+    %r((?:https?://)?(?:www\.)?youtube\.com/embed/(.*?)(\?|$)),
+    %r((?:https?://)?(?:www\.)?youtube\.com/v/(.*?)(#|\?|$)),
+    %r((?:https?://)?(?:www\.)?youtube\.com/user/.*?#\w/\w/\w/\w/(.+)\b)
+  ]
+end

data/lib/youtube-captions/errors.rb

@@ -0,0 +1,5 @@
+module YoutubeCaptions
+  class Error < StandardError; end
+  class NoCaptionsAvailableError < Error; end
+  class LangNotAvailableError < Error; end
+end

data/lib/youtube-captions/info.rb

@@ -0,0 +1,20 @@
+require 'httparty'
+
+module YoutubeCaptions
+  class Info
+    include HTTParty
+
+    attr_reader :id
+    def initialize(id:)
+      @id = id
+    end
+
+    def call
+      youtube_html = self.class.get("#{YoutubeCaptions::YOUTUBE_VIDEO_URL}#{id}")
+      match_data = youtube_html.match(YoutubeCaptions::CAPTIONABLE_REGEX)
+      return raise NoCaptionsAvailableError.new("No captions available") unless match_data
+
+      JSON.parse(match_data[1])
+    end
+  end
+end

data/lib/youtube-captions/video.rb

@@ -0,0 +1,22 @@
+module YoutubeCaptions
+  class Video
+    attr_reader :info
+    def initialize(id:)
+      @info = YoutubeCaptions::Info.new(id: parse_youtube_id(id)).call
+    end
+
+    def captions(lang: nil)
+      YoutubeCaptions::Captions.new(info: info, lang: lang).call
+    end
+
+    def available_langs
+      info.map { |json| json["languageCode"] }
+    end
+
+    private
+
+    def parse_youtube_id(id)
+      YoutubeCaptions::YOUTUBE_URL_FORMATS.find { |format| id =~ format } && $1 || id
+    end
+  end
+end

data/lib/youtube-captions.rb

@@ -1,44 +1,8 @@
-require 'cgi'
-require 'httparty'
+require 'youtube-captions/captions'
+require 'youtube-captions/constants'
+require 'youtube-captions/errors'
+require 'youtube-captions/info'
+require 'youtube-captions/video'
 
-class YoutubeCaptions
-  include HTTParty
-
-  attr_reader :id, :lang
-  def initialize(id:, lang: nil)
-    @id = id
-    @lang = lang
-  end
-
-  TRANSLATABLE_REGEX = /({"captionTracks":.*"isTranslatable":(true|false)}\])/
-
-  def call
-    youtube_html = self.class.get("https://www.youtube.com/watch?v=#{id}")
-    match_data = youtube_html.match(TRANSLATABLE_REGEX)
-    raise StandardError.new("There are no captions") unless match_data
-
-    caption_tracks = JSON.parse("#{match_data[1]}}")["captionTracks"]
-    if lang.present?
-      subtitle = caption_tracks.find {|json| json["vssId"] == ".#{lang}"} || caption_tracks.find {|json| json["vssId"] == "a.#{lang}"}
-      if !subtitle || (subtitle && !subtitle["baseUrl"])
-        raise ArgumentError.new("Lang no available")
-      end
-    else
-      subtitle = caption_tracks.find {|json| json["vssId"] == ".en"} || caption_tracks.find {|json| json["vssId"] == "a.en"}
-      subtitle = caption_tracks.first unless subtitle
-    end
-
-    transcript_html = self.class.get(subtitle["baseUrl"])
-    transcript_tags = transcript_html.to_s.gsub('<?xml version="1.0" encoding="utf-8" ?><transcript>', '').gsub('</transcript>', '').split('</text>').select {|line| line && line.strip}
-
-    transcript_parts = transcript_tags.map do |transcript_tag|
-      encoded_transcript = transcript_tag.gsub(/<text.+>/, '')
-                                         .gsub("/&amp;/gi", '&')
-                                         .gsub("/<\/?[^>]+(>|$)/g", '')
-                                         .gsub(/&amp;#(\d+);/) { [$1.to_i].pack('U*') }
-      CGI.unescapeHTML(encoded_transcript)
-    end
-
-    transcript_parts.join(" ")
-  end
+module YoutubeCaptions
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: youtube-captions
 version: !ruby/object:Gem::Version
-  version: 0.0.4
+  version: 1.0.0
 platform: ruby
 authors:
 - Kevin S.
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2023-07-28 00:00:00.000000000 Z
+date: 2023-10-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: httparty
@@ -24,6 +24,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: 
 email: 
 executables: []
@@ -31,6 +45,11 @@ extensions: []
 extra_rdoc_files: []
 files:
 - lib/youtube-captions.rb
+- lib/youtube-captions/captions.rb
+- lib/youtube-captions/constants.rb
+- lib/youtube-captions/errors.rb
+- lib/youtube-captions/info.rb
+- lib/youtube-captions/video.rb
 homepage: https://github.com/sevinchek/youtube-captions
 licenses: []
 metadata: {}
@@ -42,7 +61,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 2.3.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="