rwdeliza 0.03

data/lib/linguistics/en/linkparser.rb

@@ -0,0 +1,142 @@
+#!/usr/bin/ruby
+# 
+# This file contains the extensions to the Linguistics::EN module which provide
+# support for the Ruby LinkParser module. LinkParser enables grammatic queries
+# of English language sentences.
+#
+# == Synopsis
+#
+#   # Test to see whether or not the link parser is loaded.
+#   Linguistics::EN.has_link_parser?
+#   # => true
+# 
+#   # Diagram the first linkage for a test sentence
+#   puts "he is a big dog".sentence.linkages.first.to_s
+# 	  +---O*---+ 
+# 	  | +--Ds--+ 
+#    +Ss+ |  +-A-+ 
+#    |  | |  |   | 
+#   he is a big dog
+# 
+#   # Find the verb in the sentence
+#   "he is a big dog".en.sentence.verb.to_s      
+#   # => "is"
+# 
+#   # Combined infinitive + LinkParser: Find the infinitive form of the verb of the
+#   given sentence.
+#   "he is a big dog".en.sentence.verb.infinitive
+#   # => "be"
+# 
+#   # Find the direct object of the sentence
+#   "he is a big dog".en.sentence.object.to_s
+#   # => "dog"
+# 
+#   # Look at the raw LinkParser::Word for the direct object of the sentence.
+#   "he is a big dog".en.sentence.object     
+#   # => #<LinkParser::Word:0x403da0a0 @definition=[[{@A-}, Ds-, {@M+}, J-], [{@A-},
+#   Ds-, {@M+}, Os-], [{@A-}, Ds-, {@M+}, Ss+, {@CO-}, {C-}], [{@A-}, Ds-, {@M+},
+#   Ss+, R-], [{@A-}, Ds-, {@M+}, SIs-], [{@A-}, Ds-, {R+}, {Bs+}, J-], [{@A-}, Ds-,
+#   {R+}, {Bs+}, Os-], [{@A-}, Ds-, {R+}, {Bs+}, Ss+, {@CO-}, {C-}], [{@A-}, Ds-,
+#   {R+}, {Bs+}, Ss+, R-], [{@A-}, Ds-, {R+}, {Bs+}, SIs-]], @right=[], @suffix="",
+#   @left=[#<LinkParser::Connection:0x403da028 @rword=#<LinkParser::Word:0x403da0a0
+#   ...>, @lword=#<LinkParser::Word:0x403da0b4 @definition=[[Ss-, O+, {@MV+}], [Ss-,
+#   B-, {@MV+}], [Ss-, P+], [Ss-, AF-], [RS-, Bs-, O+, {@MV+}], [RS-, Bs-, B-,
+#   {@MV+}], [RS-, Bs-, P+], [RS-, Bs-, AF-], [{Q-}, SIs+, O+, {@MV+}], [{Q-}, SIs+,
+#   B-, {@MV+}], [{Q-}, SIs+, P+], [{Q-}, SIs+, AF-]],
+#   @right=[#<LinkParser::Connection:0x403da028 ...>], @suffix="", @left=[],
+#   @name="is", @position=1>, @subName="*", @name="O", @length=3>], @name="dog",
+#   @position=4>
+# 
+#   # Combine WordNet + LinkParser to find the definition of the direct object of
+#   # the sentence
+#   "he is a big dog".en.sentence.object.gloss
+#   # => "a member of the genus Canis (probably descended from the common wolf) that
+#   has been domesticated by man since prehistoric times; occurs in many breeds;
+#   \"the dog barked all night\""
+# 
+# == Authors
+# 
+# * Martin Chase <stillflame@FaerieMUD.org>
+# * Michael Granger <ged@FaerieMUD.org>
+# 
+# == Copyright
+#
+# Copyright (c) 2003 The FaerieMUD Consortium. All rights reserved.
+# 
+# This module is free software. You may use, modify, and/or redistribute this
+# software under the terms of the Perl Artistic License. (See
+# http://language.perl.com/misc/Artistic.html)
+# 
+#  # == Version
+#
+#  $Id: linkparser.rb,v 1.4 2003/09/14 11:15:33 deveiant Exp $
+# 
+
+require 'linguistics/en'
+
+module Linguistics::EN
+
+	@hasLinkParser	= true
+	@lpParser		= nil
+	@lpError		= nil
+
+	begin
+		require "linkparser"
+		@hasLinkParser = true
+	rescue LoadError => err
+		@lpError = err
+	end
+
+
+	#################################################################
+	###	M O D U L E   M E T H O D S
+	#################################################################
+	class << self
+
+		### Returns +true+ if LinkParser was loaded okay
+		def hasLinkParser? ; @hasLinkParser ; end
+
+		### If #hasLinkParser? returns +false+, this can be called to fetch the
+		### exception which was raised when trying to load LinkParser.
+		def lpError ; @lpError ; end
+
+		### The instance of LinkParser used for all Linguistics LinkParser
+		### functions.
+		def linkParser
+			if @lpError
+				raise NotImplementedError, 
+					"LinkParser functions are not loaded: %s" %
+					@lpError.message
+			end
+
+			return @lpParser if ! @lpParser.nil?
+
+			LinkParser::Word::extend( Linguistics )
+			Linguistics::installDelegatorProxy( LinkParser::Word, :en )
+
+			dictOpts = Hash.new('')
+			dictOpts['datadir'] = 'ElizaData'
+		dictOpts['dict'] = 'tiny.dict'
+			parseOpts = Hash.new
+
+			@lpParser = LinkParser.new( dictOpts, parseOpts )
+		end
+	end
+
+
+	#################################################################
+	###	L I N K P A R S E R   I N T E R F A C E
+	#################################################################
+
+	###############
+	module_function
+	###############
+
+	### Return a LinkParser::Sentence, with or without a sentence in it.
+	def linkParse( sent )
+		return Linguistics::EN::linkParser.parse( sent.to_s )
+	end
+	alias_method :sentence, :linkParse
+	module_function :sentence
+
+end

data/lib/linguistics/en/wordnet.rb

@@ -0,0 +1,253 @@
+#!/usr/bin/ruby
+# 
+# This file contains functions for finding relations for English words. It
+# requires the Ruby-WordNet module to be installed; if it is not installed,
+# calling the functions defined by this file will raise NotImplemented
+# exceptions if called. Requiring this file adds functions and constants to the
+# Linguistics::EN module.
+#
+# == Synopsis
+#
+#   # Test to be sure the WordNet module loaded okay.
+#   Linguistics::EN.has_wordnet?
+#   # => true
+#  
+#   # Fetch the default synset for the word "balance"
+#   "balance".synset
+#   # => #<WordNet::Synset:0x40376844 balance (noun): "a state of equilibrium"
+#    (derivations: 3, antonyms: 1, hypernyms: 1, hyponyms: 3)>
+#  
+#   # Fetch the synset for the first verb sense of "balance"
+#   "balance".en.synset( :verb )
+#   # => #<WordNet::Synset:0x4033f448 balance, equilibrate, equilibrize, equilibrise
+#   (verb): "bring into balance or equilibrium; "She has to balance work and her
+#   domestic duties"; "balance the two weights"" (derivations: 7, antonyms: 1,
+#   verbGroups: 2, hypernyms: 1, hyponyms: 5)>
+#  
+#   # Fetch the second noun sense
+#   "balance".en.synset( 2, :noun )
+#   # => #<WordNet::Synset:0x404ebb24 balance (noun): "a scale for weighing; depends
+#   on pull of gravity" (hypernyms: 1, hyponyms: 5)>
+#  
+#   # Fetch the second noun sense's hypernyms (more-general words, like a superclass)
+#   "balance".en.synset( 2, :noun ).hypernyms
+#   # => [#<WordNet::Synset:0x404e5620 scale, weighing machine (noun): "a measuring
+#   instrument for weighing; shows amount of mass" (derivations: 2, hypernyms: 1,
+#   hyponyms: 2)>]
+#  
+#   # A simpler way of doing the same thing:
+#   "balance".en.hypernyms( 2, :noun )
+#   # => [#<WordNet::Synset:0x404e5620 scale, weighing machine (noun): "a measuring
+#   instrument for weighing; shows amount of mass" (derivations: 2, hypernyms: 1,
+#   hyponyms: 2)>]
+#  
+#   # Fetch the first hypernym's hypernyms
+#   "balance".en.synset( 2, :noun ).hypernyms.first.hypernyms
+#   # => [#<WordNet::Synset:0x404c60b8 measuring instrument, measuring system,
+#   measuring device (noun): "instrument that shows the extent or amount or quantity
+#   or degree of something" (hypernyms: 1, hyponyms: 83)>]
+#  
+#   # Find the synset to which both the second noun sense of "balance" and the
+#   # default sense of "shovel" belong.
+#   ("balance".en.synset( 2, :noun ) | "shovel".en.synset)
+#   # => #<WordNet::Synset:0x40473da4 instrumentality, instrumentation (noun): "an
+#   artifact (or system of artifacts) that is instrumental in accomplishing some
+#   end" (derivations: 1, hypernyms: 1, hyponyms: 13)>
+#  
+#   # Fetch just the words for the other kinds of "instruments"
+#   "instrument".en.hyponyms.collect {|synset| synset.words}.flatten
+#   # => ["analyzer", "analyser", "cautery", "cauterant", "drafting instrument",
+#   "extractor", "instrument of execution", "instrument of punishment", "measuring
+#   instrument", "measuring system", "measuring device", "medical instrument",
+#   "navigational instrument", "optical instrument", "plotter", "scientific
+#   instrument", "sonograph", "surveying instrument", "surveyor's instrument",
+#   "tracer", "weapon", "arm", "weapon system", "whip"]
+# 
+# 
+# == Authors
+# 
+# * Michael Granger <ged@FaerieMUD.org>
+# 
+# == Copyright
+#
+# Copyright (c) 2003 The FaerieMUD Consortium. All rights reserved.
+# 
+# This module is free software. You may use, modify, and/or redistribute this
+# software under the terms of the Perl Artistic License. (See
+# http://language.perl.com/misc/Artistic.html)
+# 
+# == Version
+#
+#  $Id: wordnet.rb,v 1.3 2003/09/14 11:28:02 deveiant Exp $
+# 
+
+module Linguistics
+module EN
+
+	@hasWordnet		= false
+	@wnError		= nil
+	@wnLexicon		= nil
+
+	# Load WordNet and open the lexicon if possible, saving the error that
+	# occurs if anything goes wrong.
+	begin
+		require 'wordnet'
+		@hasWordnet = true
+	rescue LoadError => err
+		@wnError = err
+	end
+
+
+	#################################################################
+	###	M O D U L E   M E T H O D S
+	#################################################################
+	class << self
+
+		### Returns +true+ if WordNet was loaded okay
+		def hasWordnet? ; @hasWordnet; end
+
+		### If #haveWordnet? returns +false+, this can be called to fetch the
+		### exception which was raised when WordNet was loaded.
+		def wnError ; @wnError; end
+
+		### The instance of the WordNet::Lexicon used for all Linguistics WordNet
+		### functions.
+		def wnLexicon
+			if @wnError
+				raise NotImplementedError,
+					"WordNet functions are not loaded: %s" %
+					@wnError.message
+			end
+
+			@wnLexicon ||= WordNet::Lexicon::new
+		end
+
+		### Make a function that calls the method +meth+ on the synset of an input
+		### word.
+		def def_synset_function( meth )
+			(class << self; self; end).instance_eval do
+				define_method( meth ) {|*args|
+					word, pos, sense = *args
+					raise ArgumentError,
+						"wrong number of arguments (0 for 1)" unless word
+					sense ||= 1
+
+					syn = synset( word.to_s, pos, sense )
+					return syn.nil? ? nil : syn.send( meth )
+				}
+			end
+		end
+	end
+
+
+
+	#################################################################
+	###	W O R D N E T   I N T E R F A C E
+	#################################################################
+
+	###############
+	module_function
+	###############
+
+	### Look up the synset associated with the given word or collocation in the
+	### WordNet lexicon and return a WordNet::Synset object.
+	def synset( word, pos=nil, sense=1 )
+		lex = Linguistics::EN::wnLexicon
+		if pos.is_a?( Fixnum)
+			sense = pos
+			pos = nil
+		end
+		postries = pos ? [pos] : [:noun, :verb, :adjective, :adverb, :other]
+		syn = nil
+
+		postries.each {|pos|
+			break if syn = lex.lookupSynsets( word.to_s, pos, sense )
+		}
+
+		return syn
+	end
+
+
+	### Look up all the synsets associated with the given word or collocation in
+	### the WordNet lexicon and return an Array of WordNet::Synset objects. If
+	### +pos+ is +nil+, return synsets for all parts of speech.
+	def synsets( word, pos=nil )
+		lex = Linguistics::EN::wnLexicon
+		postries = pos ? [pos] : [:noun, :verb, :adjective, :adverb, :other]
+		syns = []
+
+		postries.each {|pos|
+			syns << lex.lookupSynsets( word.to_s, pos )
+		}
+
+		return syns.flatten.compact
+	end
+
+
+	# Returns definitions and/or example sentences as a String.
+	def_synset_function :gloss
+
+	# Returns definitions and/or example sentences as an Array.
+	def_synset_function :glosses
+
+	# Return nouns or verbs that have the same hypernym as the receiver.
+	def_synset_function :coordinates
+
+	# Returns the Array of synonyms contained in the synset for the receiver.
+	def_synset_function :words
+	def_synset_function :synonyms
+
+	# Returns the name of the lexicographer file that contains the raw data for
+	# the receiver.
+	def_synset_function :lexInfo
+
+	# :TODO: Finish these comments, and figure out how the hell to get the
+	# methods to show up in RDoc.
+	def_synset_function :frames
+
+	def_synset_function :antonyms
+	def_synset_function :hypernyms
+	def_synset_function :entailment
+	def_synset_function :hyponyms
+	def_synset_function :causes
+	def_synset_function :verbgroups
+	def_synset_function :similarTo
+	def_synset_function :participles
+	def_synset_function :pertainyms
+	def_synset_function :attributes
+	def_synset_function :derivedFrom
+	def_synset_function :seeAlso
+	def_synset_function :functions
+
+	def_synset_function :meronyms
+	def_synset_function :memberMeronyms
+	def_synset_function :stuffMeronyms
+	def_synset_function :portionMeronyms
+	def_synset_function :componentMeronyms
+	def_synset_function :featureMeronyms
+	def_synset_function :phaseMeronyms
+	def_synset_function :placeMeronyms
+
+	def_synset_function :holonyms
+	def_synset_function :memberHolonyms
+	def_synset_function :stuffHolonyms
+	def_synset_function :portionHolonyms
+	def_synset_function :componentHolonyms
+	def_synset_function :featureHolonyms
+	def_synset_function :phaseHolonyms
+	def_synset_function :placeHolonyms
+
+	def_synset_function :domains
+	def_synset_function :categoryDomains
+	def_synset_function :regionDomains
+	def_synset_function :usageDomains
+
+	def_synset_function :members
+	def_synset_function :categoryMembers
+	def_synset_function :regionMembers
+	def_synset_function :usageMembers
+
+
+end # module EN
+end # module Linguistics
+

data/lib/linguistics/iso639.rb

@@ -0,0 +1,456 @@
+#!/usr/bin/ruby
+#
+# linguistics/iso639.rb - A hash of International 2- and 3-letter 
+# ISO639-1 and ISO639-2 language codes. Each entry has two keys:
+#
+# [<tt>:codes</tt>]
+#   All of the codes known for this language
+# [<tt>:desc</tt>]
+#   The English-language description of the language.
+#
+
+### A language-independent framework for adding linguistics functions to Ruby
+### classes.
+module Linguistics
+
+	# Hash of ISO639 2- and 3-letter language codes
+	LanguageCodes = {}
+
+	# Read through the source for this file, capturing everything
+	# between __END__ and __END_DATA__ tokens.
+	inDataSection = false
+	File::readlines( __FILE__ ).each {|line|
+		case line
+		when /^__END_DATA__$/
+			inDataSection = false
+			false
+			
+		when /^__END__$/
+			inDataSection = true
+			false
+			
+		else
+			if inDataSection
+				codes, desc = line[0,15].split(%r{/|\s+}), line[15...-1]
+				codes.delete_if {|code| code.empty?}
+				entry = {
+					:desc	=> desc.strip,
+					:codes	=> codes.dup,
+				}
+				codes.each {|code|
+					raise "Duplicate language code #{code}:"\
+						"(#{LanguageCodes[code][:desc]}})}" \
+						if LanguageCodes.key?( code )
+					LanguageCodes[ code.strip ] = entry
+				}
+			end
+		end
+	}
+end
+
+__END__
+abk      ab    Abkhazian
+ace            Achinese
+ach            Acoli
+ada            Adangme
+aar      aa    Afar
+afh            Afrihili
+afr      af    Afrikaans
+afa            Afro-Asiatic (Other)
+aka            Akan
+akk            Akkadian
+alb/sqi  sq    Albanian
+ale            Aleut
+alg            Algonquian languages
+tut            Altaic (Other)
+amh      am    Amharic
+apa            Apache languages
+ara      ar    Arabic
+arc            Aramaic
+arp            Arapaho
+arn            Araucanian
+arw            Arawak
+arm/hye  hy    Armenian
+art            Artificial (Other)
+asm      as    Assamese
+ath            Athapascan languages
+map            Austronesian (Other)
+ava            Avaric
+ave            Avestan
+awa            Awadhi
+aym      ay    Aymara
+aze      az    Azerbaijani
+nah            Aztec
+ban            Balinese
+bat            Baltic (Other)
+bal            Baluchi
+bam            Bambara
+bai            Bamileke languages
+bad            Banda
+bnt            Bantu (Other)
+bas            Basa
+bak      ba    Bashkir
+baq/eus  eu    Basque
+bej            Beja
+bem            Bemba
+ben      bn    Bengali
+ber            Berber (Other)
+bho            Bhojpuri
+bih      bh    Bihari
+bik            Bikol
+bin            Bini
+bis      bi    Bislama
+bra            Braj
+bre      br    Breton
+bug            Buginese
+bul      bg    Bulgarian
+bua            Buriat
+bur/mya  my    Burmese
+bel      be    Byelorussian
+cad            Caddo
+car            Carib
+cat      ca    Catalan
+cau            Caucasian (Other)
+ceb            Cebuano
+cel            Celtic (Other)
+cai            Central American Indian (Other)
+chg            Chagatai
+cha            Chamorro
+che            Chechen
+chr            Cherokee
+chy            Cheyenne
+chb            Chibcha
+chi/zho  zh    Chinese
+chn            Chinook jargon
+cho            Choctaw
+chu            Church Slavic
+chv            Chuvash
+cop            Coptic
+cor            Cornish
+cos      co    Corsican
+cre            Cree
+mus            Creek
+crp            Creoles and Pidgins (Other)
+cpe            Creoles and Pidgins, English-based (Other)
+cpf            Creoles and Pidgins, French-based (Other)
+cpp            Creoles and Pidgins, Portuguese-based (Other)
+cus            Cushitic (Other)
+         hr    Croatian
+ces/cze  cs    Czech
+dak            Dakota
+dan      da    Danish
+del            Delaware
+din            Dinka
+div            Divehi
+doi            Dogri
+dra            Dravidian (Other)
+dua            Duala
+dut/nla  nl    Dutch
+dum            Dutch, Middle (ca. 1050-1350)
+dyu            Dyula
+dzo      dz    Dzongkha
+efi            Efik
+egy            Egyptian (Ancient)
+eka            Ekajuk
+elx            Elamite
+eng      en    English
+enm            English, Middle (ca. 1100-1500)
+ang            English, Old (ca. 450-1100)
+esk            Eskimo (Other)
+epo      eo    Esperanto
+est      et    Estonian
+ewe            Ewe
+ewo            Ewondo
+fan            Fang
+fat            Fanti
+fao      fo    Faroese
+fij      fj    Fijian
+fin      fi    Finnish
+fiu            Finno-Ugrian (Other)
+fon            Fon
+fra/fre  fr    French
+frm            French, Middle (ca. 1400-1600)
+fro            French, Old (842- ca. 1400)
+fry      fy    Frisian
+ful            Fulah
+gaa            Ga
+gae/gdh        Gaelic (Scots)
+glg      gl    Gallegan
+lug            Ganda
+gay            Gayo
+gez            Geez
+geo/kat  ka    Georgian
+deu/ger  de    German
+gmh            German, Middle High (ca. 1050-1500)
+goh            German, Old High (ca. 750-1050)
+gem            Germanic (Other)
+gil            Gilbertese
+gon            Gondi
+got            Gothic
+grb            Grebo
+grc            Greek, Ancient (to 1453)
+ell/gre  el    Greek, Modern (1453-)
+kal      kl    Greenlandic
+grn      gn    Guarani
+guj      gu    Gujarati
+hai            Haida
+hau      ha    Hausa
+haw            Hawaiian
+heb      he    Hebrew
+her            Herero
+hil            Hiligaynon
+him            Himachali
+hin      hi    Hindi
+hmo            Hiri Motu
+hun      hu    Hungarian
+hup            Hupa
+iba            Iban
+ice/isl  is    Icelandic
+ibo            Igbo
+ijo            Ijo
+ilo            Iloko
+inc            Indic (Other)
+ine            Indo-European (Other)
+ind      id    Indonesian
+ina      ia    Interlingua (International Auxiliary language Association)
+ile            Interlingue
+iku      iu    Inuktitut
+ipk      ik    Inupiak
+ira            Iranian (Other)
+gai/iri  ga    Irish
+sga            Irish, Old (to 900)
+mga            Irish, Middle (900 - 1200)
+iro            Iroquoian languages
+ita      it    Italian
+jpn      ja    Japanese
+jav/jaw  jv/jw Javanese
+jrb            Judeo-Arabic
+jpr            Judeo-Persian
+kab            Kabyle
+kac            Kachin
+kam            Kamba
+kan      kn    Kannada
+kau            Kanuri
+kaa            Kara-Kalpak
+kar            Karen
+kas      ks    Kashmiri
+kaw            Kawi
+kaz      kk    Kazakh
+kha            Khasi
+khm      km    Khmer
+khi            Khoisan (Other)
+kho            Khotanese
+kik            Kikuyu
+kin      rw    Kinyarwanda
+kir      ky    Kirghiz
+kom            Komi
+kon            Kongo
+kok            Konkani
+kor      ko    Korean
+kpe            Kpelle
+kro            Kru
+kua            Kuanyama
+kum            Kumyk
+kur      ku    Kurdish
+kru            Kurukh
+kus            Kusaie
+kut            Kutenai
+lad            Ladino
+lah            Lahnda
+lam            Lamba
+oci      oc    Langue d'Oc (post 1500)
+lao      lo    Lao
+lat      la    Latin
+lav      lv    Latvian
+ltz            Letzeburgesch
+lez            Lezghian
+lin      ln    Lingala
+lit      lt    Lithuanian
+loz            Lozi
+lub            Luba-Katanga
+lui            Luiseno
+lun            Lunda
+luo            Luo (Kenya and Tanzania)
+mac/mke  mk    Macedonian
+mad            Madurese
+mag            Magahi
+mai            Maithili
+mak            Makasar
+mlg      mg    Malagasy
+may/msa  ms    Malay
+mal            Malayalam
+mlt      ml    Maltese
+man            Mandingo
+mni            Manipuri
+mno            Manobo languages
+max            Manx
+mao/mri  mi    Maori
+mar      mr    Marathi
+chm            Mari
+mah            Marshall
+mwr            Marwari
+mas            Masai
+myn            Mayan languages
+men            Mende
+mic            Micmac
+min            Minangkabau
+mis            Miscellaneous (Other)
+moh            Mohawk
+mol      mo    Moldavian
+mkh            Mon-Kmer (Other)
+lol            Mongo
+mon      mn    Mongolian
+mos            Mossi
+mul            Multiple languages
+mun            Munda languages
+nau      na    Nauru
+nav            Navajo
+nde            Ndebele, North
+nbl            Ndebele, South
+ndo            Ndongo
+nep      ne    Nepali
+new            Newari
+nic            Niger-Kordofanian (Other)
+ssa            Nilo-Saharan (Other)
+niu            Niuean
+non            Norse, Old
+nai            North American Indian (Other)
+nor      no    Norwegian
+nno            Norwegian (Nynorsk)
+nub            Nubian languages
+nym            Nyamwezi
+nya            Nyanja
+nyn            Nyankole
+nyo            Nyoro
+nzi            Nzima
+oji            Ojibwa
+ori      or    Oriya
+orm      om    Oromo
+osa            Osage
+oss            Ossetic
+oto            Otomian languages
+pal            Pahlavi
+pau            Palauan
+pli            Pali
+pam            Pampanga
+pag            Pangasinan
+pan      pa    Panjabi
+pap            Papiamento
+paa            Papuan-Australian (Other)
+fas/per  fa    Persian
+peo            Persian, Old (ca 600 - 400 B.C.)
+phn            Phoenician
+pol      pl    Polish
+pon            Ponape
+por      pt    Portuguese
+pra            Prakrit languages
+pro            Provencal, Old (to 1500)
+pus      ps    Pushto
+que      qu    Quechua
+roh      rm    Rhaeto-Romance
+raj            Rajasthani
+rar            Rarotongan
+roa            Romance (Other)
+ron/rum  ro    Romanian
+rom            Romany
+run      rn    Rundi
+rus      ru    Russian
+sal            Salishan languages
+sam            Samaritan Aramaic
+smi            Sami languages
+smo      sm    Samoan
+sad            Sandawe
+sag      sg    Sango
+san      sa    Sanskrit
+srd            Sardinian
+sco            Scots
+sel            Selkup
+sem            Semitic (Other)
+         sr    Serbian
+scr      sh    Serbo-Croatian
+srr            Serer
+shn            Shan
+sna      sn    Shona
+sid            Sidamo
+bla            Siksika
+snd      sd    Sindhi
+sin      si    Singhalese
+sit            Sino-Tibetan (Other)
+sio            Siouan languages
+sla            Slavic (Other)
+         ss    Siswati
+slk/slo  sk    Slovak
+slv      sl    Slovenian
+sog            Sogdian
+som      so    Somali
+son            Songhai
+wen            Sorbian languages
+nso            Sotho, Northern
+sot      st    Sotho, Southern
+sai            South American Indian (Other)
+esl/spa  es    Spanish
+suk            Sukuma
+sux            Sumerian
+sun      su    Sudanese
+sus            Susu
+swa      sw    Swahili
+ssw            Swazi
+sve/swe  sv    Swedish
+syr            Syriac
+tgl      tl    Tagalog
+tah            Tahitian
+tgk      tg    Tajik
+tmh            Tamashek
+tam      ta    Tamil
+tat      tt    Tatar
+tel      te    Telugu
+ter            Tereno
+tha      th    Thai
+bod/tib  bo    Tibetan
+tig            Tigre
+tir      ti    Tigrinya
+tem            Timne
+tiv            Tivi
+tli            Tlingit
+tog      to    Tonga (Nyasa)
+ton            Tonga (Tonga Islands)
+tru            Truk
+tsi            Tsimshian
+tso      ts    Tsonga
+tsn      tn    Tswana
+tum            Tumbuka
+tur      tr    Turkish
+ota            Turkish, Ottoman (1500 - 1928)
+tuk      tk    Turkmen
+tyv            Tuvinian
+twi      tw    Twi
+uga            Ugaritic
+uig      ug    Uighur
+ukr      uk    Ukrainian
+umb            Umbundu
+und            Undetermined
+urd      ur    Urdu
+uzb      uz    Uzbek
+vai            Vai
+ven            Venda
+vie      vi    Vietnamese
+vol      vo    Volap�k
+vot            Votic
+wak            Wakashan languages
+wal            Walamo
+war            Waray
+was            Washo
+cym/wel  cy    Welsh
+wol      wo    Wolof
+xho      xh    Xhosa
+sah            Yakut
+yao            Yao
+yap            Yap
+yid      yi    Yiddish
+yor      yo    Yoruba
+zap            Zapotec
+zen            Zenaga
+zha      za    Zhuang
+zul      zu    Zulu
+zun            Zuni
+__END_DATA__