toname 0.1.0

data/README.md

@@ -0,0 +1,52 @@
+# ToName
+
+Overview
+--------
+ToName is a Ruby library which parses media filenames and extracts metadata (e.g. name and year)
+
+e.g.
+ 
+ * Primer.2004.DVDRip.x264.AC3.avi become Name => 'Primer', year => 2004
+ * ManMen.S01E01.720p.HDTV.x264-CTU.mkv becomes Name => 'ManMen', series => 1, episode => 1
+
+Most media file names sourced from P2P sites are not immediately machine readable, however, they generally have a fairly consistent naming schema.  ToName applies some simple parsing rules to extract enough information so that the movie/tv show can be looked up in a metadata database (e.g. [IMDB](http://www.imdb.com)).  ToName does not claim a 100% success rate, however, it's generally good enough to find a match on IMDB.
+
+Features
+--------
+ * Extract movie name and year
+ * Extract TV show series and episode
+ * Will look at the parent folder name if media name looks like it has been abbreviated
+ * Works with both media and torrent filenames
+
+Installation
+------------
+
+	gem install toname
+
+Examples
+--------
+
+	ruby-1.9.2-preview3 > require 'to_name'
+	 => true 
+	ruby-1.9.2-preview3 > info = ToName.to_name('Primer.2004.DVDRip.x264.AC3.avi')
+	 => Primer (2004) 
+	ruby-1.9.2-preview3 > info.name
+	 => "Primer" 
+	ruby-1.9.2-preview3 > info.year
+	 => 2004 
+	ruby-1.9.2-preview3 > info = ToName.to_name('ManMen.S01E01.720p.HDTV.x264-CTU.mkv')
+	 => ManMen S: 1 E: 1 
+	ruby-1.9.2-preview3 > info.name
+	 => "ManMen"
+	ruby-1.9.2-preview3 > info.series
+	 => 1 
+	ruby-1.9.2-preview3 > info.episode
+	 => 1 
+	
+Licence
+-------
+MIT
+
+Contact
+-------
+Sam Cavenagh [(cavenaghweb@hotmail.com)](mailto:cavenaghweb@hotmail.com)

data/lib/file_name_info.rb

@@ -0,0 +1,41 @@
+class FileNameInfo
+  attr_accessor :raw_name, :location, :name, :year, :series, :episode
+
+  def initialize(params = {})
+    @name = params[:name]
+    @year = params[:year]
+    @raw_name = params[:raw_name]
+    @location = params[:location]
+    @series = params[:series]
+    @episode = params[:episode]
+  end
+
+  def to_s
+    s = ''
+    if @name
+      s += @name
+      s += " (#{@year})" if @year
+      s += " S: #{@series} E: #{@episode}" if @series || @episode
+    elsif @raw_name
+      s += @raw_name
+    else
+      s += @location
+    end
+    s
+  end
+  
+  def <=>(other)
+    if series && other.series && series != other.series
+      return (series <=> other.series) 
+    elsif episode && other.episode && episode != other.episode
+      return (episode <=> other.episode)
+    elsif name != other.name
+      return (name <=> other.name)
+    elsif year && other.year
+      #Note: with year we want newest first
+      return (other.year <=> year)
+    else
+      return 0
+    end
+  end
+end

data/lib/to_name.rb

@@ -0,0 +1,103 @@
+require 'file_name_info'
+
+class ToName
+
+  FILE_SEP_REGEX = /\//
+  FILE_EXT_SEP_REGEX = /\./
+  CD_FOLDER_REGEX = /\/CD(\d)\//
+  #Chars used in file names as a subsitude for spaces
+  SPACE_SUB_REGEX = /(\.|_|\-)/
+  VIDEO_TYPE_NAMES = ['DVDRIP', '1080p', '720p','R5', 'DVDSCR', 'BDRip', 'CAM', 'TS', 'PPV', 'Xvid', 'divx', 'DVDSCREENER']
+  CONTENT_SOURCE_FOLDER_TEST_REGEX = /#{VIDEO_TYPE_NAMES.join('|')}/i
+  CONTENT_SOURCE_REGEX = /(\(|\[|\s)+(#{VIDEO_TYPE_NAMES.join('|')})(\)|\]|\s|$)+/i
+  YEAR_REGEX = /(\(|\[|\s)+\d{4}(,|\)|\]|\s|$)+/
+  SESSION_ESP_REGEX_1 = /S(\d{2})\s?E(\d{2})/i
+  SESSION_ESP_REGEX_2 = /\s+(\d+)x(\d+)(\s|$)+/i
+  SESSION_ESP_REGEX_3 = /Season (\d+) Episode (\d+)/i
+  SESSION_ESP_REGEX_OF = /(\d+)\s?of\s?(\d+)/i
+  SESSION_REGEXS = [SESSION_ESP_REGEX_1, SESSION_ESP_REGEX_2, SESSION_ESP_REGEX_3]
+
+  def self.to_name(location)
+    raw_name = self.get_file_name(location)
+  
+    #Check to see if we are better off looking at the folder name
+    check_extention = true
+    unless raw_name =~ CONTENT_SOURCE_REGEX || raw_name =~ SESSION_ESP_REGEX_1
+      parent_folder = self.parent_folder_name(location)
+      if parent_folder && parent_folder =~ CONTENT_SOURCE_FOLDER_TEST_REGEX
+        raw_name = parent_folder
+        check_extention = false
+      end  
+    end  
+  
+    #Remove file extention
+    raw_name = raw_name[0, raw_name.rindex(FILE_EXT_SEP_REGEX)] if check_extention && raw_name =~ FILE_EXT_SEP_REGEX
+    #Remove space sub chars  
+    raw_name = raw_name.gsub(SPACE_SUB_REGEX, ' ')
+
+    name = raw_name.dup
+    #Chop off any info about the movie format or source
+    name = $` if name =~ CONTENT_SOURCE_REGEX
+  
+    #Extract year if it's in the filename
+    if name =~ YEAR_REGEX && name.index(YEAR_REGEX) > 0
+      name = $`
+      #Strip any surronding brackets and convert to int
+      year = $&.gsub(/\(|\)|\[|\]/, '').to_i
+    end
+
+    #Strip LIMITED off the end.  Note: This is NOT case sensitive
+    name = $` if name =~ /LIMITED$/
+
+    #Try to extract the session and episode
+    session = nil
+    episode = nil
+    SESSION_REGEXS.each do |session_regex|
+      if name =~ session_regex
+        name = $`
+        session = $1.to_i
+        episode = $2.to_i
+        break
+      end
+    end  
+
+    if session.nil? && name =~ SESSION_ESP_REGEX_OF
+      name = $`
+      session = 1
+      episode = $1.to_i  
+    end
+
+    # Sometimes there can be multiple media files for a single movie, we want to remove the version number if this is the case
+    if location =~ CD_FOLDER_REGEX
+      cd_number = $1.to_i
+      if name =~ /#{cd_number}$/
+        name = $`
+      elsif name =~ /part\s?#{cd_number}/i
+        name = $`
+      end  
+    end  
+
+    name.strip!
+    return FileNameInfo.new(:raw_name => raw_name, :name => name, :year => year, 
+                            :series => session, :episode => episode, :location => location)  
+  end
+
+  def self.get_file_name(location)
+    file_name = location.dup
+    #Change to just the filename
+    file_name = file_name[file_name.rindex(FILE_SEP_REGEX) + 1, file_name.length] if file_name =~ FILE_SEP_REGEX  
+  
+    return file_name
+  end
+
+  def self.parent_folder_name(location)
+    # Remove first / and break by folder name
+    folders = location.sub(/^\//, '').split('/')
+    return nil if folders.empty? || folders.size < 2
+    parent_folder = folders[folders.size - 2]
+    # If the folder is a CD folder e.g. CD1 go up 1 more
+    return folders[folders.size - 3] if folders.size > 2 && parent_folder  =~ /CD\d/i
+    return parent_folder
+  end
+
+end

metadata

@@ -0,0 +1,81 @@
+--- !ruby/object:Gem::Specification 
+name: toname
+version: !ruby/object:Gem::Version 
+  prerelease: false
+  segments: 
+  - 0
+  - 1
+  - 0
+  version: 0.1.0
+platform: ruby
+authors: 
+- Sam Cavenagh
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2010-08-08 00:00:00 -07:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        segments: 
+        - 2
+        - 5
+        - 8
+        version: 2.5.8
+  type: :development
+  version_requirements: *id001
+description: Convert video/torrent filename into movie/tv series name and year
+email: cavenaghweb@hotmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- README.md
+files: 
+- README.md
+- lib/file_name_info.rb
+- lib/to_name.rb
+has_rdoc: true
+homepage: http://github.com/o-sam-o/toname
+licenses: []
+
+post_install_message: 
+rdoc_options: 
+- --main
+- README.md
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.3.7
+signing_key: 
+specification_version: 3
+summary: To Name
+test_files: []
+