wikipedia-vandalism_detection 0.1.0-java

data/lib/wikipedia/vandalism_detection/wikitext_extractor.rb

@@ -0,0 +1,80 @@
+# The WikitextExtractor imports the WikitextExtractor class from the
+# sweble-wikitext-extractor.jar
+# The sweble-wikitext-extractor.jar is a custom Java project which uses the
+# Sweble wikitext parser to extract plaintext out of wikimarkup text.
+#
+# The Sweble WikitextExtractor currently depends on the swc-engine -v1.1.0 with
+# dependencies,
+# see: http://sweble.org/downloads/swc-devel/master-latest/ to download it.
+#
+# The Java source code can be found on:
+# webis.uni-weimar.de:/srv/cvsroot/code-in-progress/wikipedia-vandalism-detection/sweble-wikitext-extractor
+module Wikipedia
+  module VandalismDetection
+    require 'java'
+    require 'java/swc-engine-1.1.0-jar-with-dependencies.jar'
+    require 'java/sweble-wikitext-extractor.jar'
+
+    java_import 'de.webis.sweble.WikitextExtractor'
+
+    class WikitextExtractionError < StandardError; end
+
+    # This class wrapps the de.webis.sweble.WikitextExtractor Java class and
+    # provides methods to extract plaintext from wiki markup text both space
+    # preserving and cleaned without line breaks and whitespace.
+    class WikitextExtractor
+      REDIRECT = '#REDIRECT'.freeze
+
+      class << self
+        # Returns the extracted text from the given wiki markup preserving
+        # spacing with added section numbers.
+        def extract(wiki_text)
+          wiki_text = wiki_text.encode('UTF-8', 'binary', invalid: :replace, undef: :replace, replace: '')
+          wiki_text = wiki_text.gsub(REDIRECT, '')
+
+          WikitextExtractor.new.extract(wiki_text)
+        rescue => exception
+          message = "Wikitext extraction failed: \n#{exception.message}"
+          raise WikitextExtractionError, message, caller
+        end
+
+        # Returns the cleaned extracted text from the given wiki markup.
+        # Cleaned means a single string without breaks, multiple spaces and
+        # section numbers.
+        def extract_clean(wiki_text)
+          wiki_text = extract wiki_text
+
+          wiki_text = remove_section_numbering_from wiki_text
+          wiki_text = remove_line_breaks_from wiki_text
+          wiki_text = remove_uris_from wiki_text
+          wiki_text = remove_special_signes_from wiki_text
+          wiki_text = remove_multiple_spaces_from wiki_text
+          wiki_text.strip
+        end
+
+        private
+
+        # removes 1., 1.1., 2.3.4. etc. at the beginning of a line
+        def remove_section_numbering_from(text)
+          text.gsub(/^(\d\.)+/, '')
+        end
+
+        def remove_line_breaks_from(text)
+          text.gsub(/\n+/, ' ')
+        end
+
+        def remove_multiple_spaces_from(text)
+          text.gsub(/\s+/, ' ')
+        end
+
+        def remove_uris_from(text)
+          text.gsub(%r{(https?|ftp)\s?:\s?\/\/[^\s\/$.?#].[^\s]*}i, '')
+        end
+
+        def remove_special_signes_from(text)
+          text.gsub(/\[\]\{\}\|\=/, ' ')
+        end
+      end
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/bad.rb

@@ -0,0 +1,11 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      BAD = %i[
+        666 da dont dosent whatever guy hi nazi sup guise loser thats ugly wanna
+        whats wont gotta bloody fart pot prick stink smells smelly alot dunno
+        gotcha
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/biased.rb

@@ -0,0 +1,20 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      BIASED = %i[
+        acclaimed amazing astonishing authoritative beautiful best brilliant
+        canonical celebrated charismatic classic cutting-edge defining
+        definitive eminent enigma exciting extraordinary fabulous famous
+        infamous fantastic fully genius global great greatest iconic immensely
+        impactful incendiary indisputable influential innovative inspired
+        intriguing leader leading legendary major masterly mature memorable
+        notable outstanding pioneer popular prestigious really remarkable
+        renowned respected seminal significant skillful solution single-handedly
+        staunch talented most top transcendent undoubtedly unique visionary
+        virtually virtuoso well-known well-established world-class worst coolest
+        super probably hate ugly fat lame weird strange everyone cares boring
+        boreing ever huge like idiotic absolute total totally
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/emoticons.rb

@@ -0,0 +1,26 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      EMOTICONS = [
+        :':\)', :':p', :':\(', :';\)', :':D', :';D', :';P', :';p', :':\-\)',
+        :':\-\(', :';\-\)', :':\-D', :':\-p', :':\-P', :'8\-\)', :'8\)',
+        :'\^\^', :'\*_\*', :'\^_\^', :':\-I', :':\-X', :':\-x', :'X\-p',
+        :'X\-P', :':\-\]', :'\^\.\^', :':\*', :':\-\*', :XD, :'X\-D', :'8\-D',
+        :'8D', :':\-O', :':\-o', :':\-\|', :'X\-\(', :'X\(', :'\-_\-', :':o\)',
+        :':O\)', :'B\-\)', :':O', :':o', :':\-s', :':\-S', :':\-\/', :':\-\\',
+        :T_T, :':\*\(', :':\*\-\(', :':\(\(', :'\*\-\*', :':\-\[', :':\->',
+        :':\|', :':\-\|', :':\]', :':\[', :'\/:\(', :'\\:\(', :':\-$', :':$',
+        :':\-6', :':\-9', :'@_@', :'<3', :'\|\-D', :':0', :':\-0', :o_O,
+        :oO, :'\(\-:', :'\(\-;', :'\(:', :'\):', :'\)\-:', :'\(;', :'\(y\)',
+        :'\(\.\)\(\.\)', :O_O, :'0_0', :'8\-\[', :'8\-\]', :'8\[', :'8\]',
+        :'8\-\(', :'8\(', :':\-', :'%\)', :'%\-\)', :'8\|', :'8\-\|', :'=\)',
+        :':\]', :':>', :':c\)', :'\[:', :'<:', :'c:', :'\(x', :'\(o:', :'\(c:',
+        :'D:', :':\'\(', :':\'C', :';\(', :';o\)', :'\(o;', :':b', :':p', :'=P',
+        :':P', :dx, :xP, :'d\-:', :'d:', :'q:', :'d=', :'d;', :'c\(:', :'=D',
+        :'=\-D', :'=O', :'=o', :'=0', :'o=', :'O=', :'0=', :'\^_~', :'>_<',
+        :'~_~', :'>:', :':<', :'\(Y\)', :'\(=', :'\)=', :'=\(', :'=\[, :=\]',
+        :'\[=', :'\]='
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/markup.rb

@@ -0,0 +1,19 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      MARKUP = [
+        :'\{\{',
+        :'\[\[',
+        :infobox,
+        :category,
+        :defaultsort,
+        :'<ref>',
+        :cite,
+        :__toc__,
+        :__forcetoc__,
+        :defaultsort,
+        :reflist
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/pronouns.rb

@@ -0,0 +1,11 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      PRONOUNS = %i[
+        i me myself mine my we us ourselves ourself ours our you yourself yours
+        your thou thee thyself thine thy yourselves y'all youse you-uns y'all
+        youse yous yis yourselves y'all's selves yous's
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/sex.rb

@@ -0,0 +1,10 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      SEX = %i[
+        vagina sex anal penis breast breasts buttocks sodomy sodomized erect
+        nipple nipples vibrator vibrator dildo dildos
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists/vulgarism.rb

@@ -0,0 +1,96 @@
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      # This list is taken from https://github.com/snipe/banbuilder and can be
+      # downloaded from:
+      # https //:github.com/snipe/banbuilder/blob/master/word-dbs/wordlist.csv
+      VULGARISM = %i[
+        $#!+ $1ut $h1t $hit $lut 'ho 'hobag a$$ anus ass assmunch b1tch
+        ballsack bastard beaner beastiality biatch beeyotch bitchy
+        blow blowjob bollock bollocks bollok boner boob bugger buttplug
+        c-0-c-k c-o-c-k c-u-n-t c.0.c.k c.o.c.k. c.u.n. jerk jackoff
+        jackhole j3rk0ff homo hom0 hobag hell h0mo h0m0 goddamn goddammit
+        godamnit ghey ghay gfy gay fudgepacker fuckwad fucktard fuckoff
+        fucker fuck-tard fuck fellatio fellate felching felcher felch
+        fartknocker fart fannybandit fanny faggot fagg fag f.u.c.k f-u-c-k
+        dyke douchebag douche douch3 doosh dike dick damnit damn dammit d1ldo
+        d1ld0 d1ck d0uche d0uch3 cunt cumstain cum crap coon cock clitoris
+        clit cl1t cawk c0ck jerk0ff jerkoff jizz knobend labia lmfao moolie
+        muff nigga nigger p.u.s.s.y. piss piss-off pissoff prick pube pussy
+        queer retard retarded s-h-1-t s-h-i-t s.h.i.t. scrotum sh1t shit slut
+        smegma t1t tard terd tit tits titties turd twat vag wank wetback
+        whore whoreface 'f*ck' sh*t pu$$y p*ssy diligaf wtf stfu fu*ck fack
+        shite fxck sh!t @sshole assh0le assho!e a$$hole a$$h0le a$$h0!e
+        a$$h01e assho1e wh0re f@g f@gg0t f@ggot motherf*cker mofo cuntlicker
+        cuntface dickbag cockknocker beatch fucknut nucking futs mams cunny
+        quim clitty kike spic wop chink humper feltch feltcher fvck ahole
+        nads spick douchey bullturds gonads bitch butt fellatio lmao s-o-b
+        spunk he11 jizm jism bukkake shiz wigger gook ritard reetard
+        masterbate masturbate goatse masterbating masturbating hitler nazi
+        tubgirl gtfo foad r-tard rtard hoor g-spot gspot vulva assmaster
+        viagra phuck frack fuckwit assbang assbanged assbangs asshole
+        assholes asswipe asswipes b1tch bastards bitched bitches boners
+        bullshit bullshits bullshitted cameltoe chinc chincs chink chode
+        chodes clit clits cocks coons cumming cunts d1ck dickhead dickheads
+        doggie-style douchebags dumass dumbass dumbasses dykes faggit fags
+        fucked fucker fuckface fucks godamnit gooks humped humping jackass
+        jap japs jerk jizzed kikes knobend kooch kooches kootch fuckers
+        motherfucking niggah niggas niggers p.u.s.s.y. pussies queers rim s0b
+        shitface shithead shits shitted s.o.b. spik spiks twats whack whores
+        zoophile m-fucking mthrfucking muthrfucking mutherfucking
+        mutherfucker mtherfucker mthrfucker mthrf*cker whorehopper copulator
+        whoralicious whorealicious aeolus analprobe areola areole aryan arian
+        asses assfuck azazel baal babes bang banger barf bawdy beardedclam
+        beater beaver beer bigtits bimbo blew blow blowjobs blowup bod bodily
+        boink bone boned bong boobies boobs booby booger bookie booky bootee
+        bootie booty booze boozer boozy bosom bosomy bowel bowels bra
+        brassiere bung babe bush buttfuck cocaine kinky klan panties
+        pedophile pedophilia pedophiliac punkass queaf rape scantily essohbee
+        shithouse smut snatch toots doggie anorexia bulimia bulimiic burp
+        busty buttfucker caca cahone carnal carpetmuncher cervix climax
+        cocain cocksucker coital coke commie condom corpse coven crabs crack
+        crackwhore crappy cuervo cummin cumshot cumshots cunnilingus dago
+        dagos damned dick-ish dickish dickweed anorexic prostitute marijuana
+        lsd pcp diddle dawgie-style dimwit dingle doofus dopey douche drunk
+        dummy ejaculate enlargement erect erotic exotic extacy extasy faerie
+        faery fagged fagot fairy fisted fisting fisty floozy fondle foobar
+        foreskin frigg frigga fubar fucking fuckup ganja gays glans godamn
+        goddam goldenshower gonad gonads handjob hebe hemp heroin herpes
+        hijack hiv homey honky hooch hookah hooker hootch hooter hooters hump
+        hussy hymen inbred incest injun jerked jiz jizm horny junkie junky
+        kill kkk kraut kyke lech leper lesbians lesbos lez lezbian lezbians
+        lezbo lezbos lezzie lezzies lezzy loin loins lube lust lusty massa
+        masterbation masturbation maxi menses menstruate menstruation meth
+        molest moron motherfucka motherfucker murder muthafucker nad naked
+        napalm nappy nazism negro niggle nimrod ninny nooky nympho opiate
+        opium oral orally organ orgasm orgies orgy ovary ovum ovums paddy
+        pantie panty pastie pasty pecker pedo pee peepee penetrate
+        penetration penial penile perversion peyote phalli phallic
+        pillowbiter pimp pinko pissed pms polack porn porno pornography pot
+        potty prig prude pubic pubis punky puss queef queefing quife quicky
+        racist racy raped raper rapist raunch rectal rectum rectus reefer
+        reich revue risque rum rump sadism sadist satan scag schizo screw
+        screwed scrog scrot scrote scrud scum seaman seamen seduce semen
+        sex_story sexual shithole shitter shitty s*o*b sissy skag slave
+        sleaze sleazy sluts smutty sniper snuff sodom souse soused sperm
+        spooge stab steamy stiffy stoned strip stroke whacking suck sucked
+        sucking tampon tawdry teat teste testee testes testis thrust thug
+        tinkle titfuck titi titty whacked toke tramp trashy tush undies unwed
+        urinal urine uterus uzi valium virgin vixen vodka vomit voyeur vulgar
+        wad wazoo wedgie weed weenie weewee weiner weirdo wench whitey whiz
+        whored whorehouse whoring womb woody x-rated xxx b@lls yeasty yobbo
+        sumofabiatch doggy-style doggy wang dong d0ng w@ng wh0reface
+        wh0ref@ce wh0r3f@ce tittyfuck tittyfucker tittiefucker cockholster
+        cockblock gai gey faig faigt a55 a55hole gae corksucker rumprammer
+        slutdumper niggaz muthafuckaz gigolo pussypounder herp herpy
+        transsexual orgasmic cunilingus anilingus dickdipper dickwhipper
+        dicksipper dickripper dickflipper dickzipper homoey queero freex
+        cunthunter shamedame slutkiss shiteater fuckass fucka$$ clitorus
+        assfucker assfuckers dillweed cracker teabagging shitt azz fuk
+        fucknugget cuntlick g@y @ss beotch pussys 's***' paedophile
+        pedophiles pedophile sucks licker lickers bitchface idiot tosser
+        idiots tossers
+      ].freeze
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection/word_lists.rb

@@ -0,0 +1,17 @@
+require 'wikipedia/vandalism_detection/word_lists/bad'
+require 'wikipedia/vandalism_detection/word_lists/biased'
+require 'wikipedia/vandalism_detection/word_lists/pronouns'
+require 'wikipedia/vandalism_detection/word_lists/sex'
+require 'wikipedia/vandalism_detection/word_lists/vulgarism'
+require 'wikipedia/vandalism_detection/word_lists/markup'
+
+module Wikipedia
+  module VandalismDetection
+    module WordLists
+      # Returns an array of all wordlist words
+      def self.all
+        [*BAD, *BIASED, *PRONOUNS, *SEX, *VULGARISM].uniq!
+      end
+    end
+  end
+end

data/lib/wikipedia/vandalism_detection.rb

@@ -0,0 +1,29 @@
+require 'wikipedia'
+require 'wikipedia/vandalism_detection/version'
+require 'wikipedia/vandalism_detection/configuration'
+require 'wikipedia/vandalism_detection/exceptions'
+
+require 'wikipedia/vandalism_detection/text'
+require 'wikipedia/vandalism_detection/revision'
+require 'wikipedia/vandalism_detection/edit'
+require 'wikipedia/vandalism_detection/page'
+require 'wikipedia/vandalism_detection/page_parser'
+require 'wikipedia/vandalism_detection/revision_parser'
+
+require 'wikipedia/vandalism_detection/word_lists'
+require 'wikipedia/vandalism_detection/diff'
+require 'wikipedia/vandalism_detection/wikitext_extractor'
+require 'wikipedia/vandalism_detection/features'
+require 'wikipedia/vandalism_detection/feature_calculator'
+
+require 'wikipedia/vandalism_detection/instances'
+require 'wikipedia/vandalism_detection/training_dataset'
+require 'wikipedia/vandalism_detection/test_dataset'
+require 'wikipedia/vandalism_detection/classifier'
+require 'wikipedia/vandalism_detection/evaluator'
+
+require 'weka/classifiers/meta/one_class_classifier'
+require 'weka/classifiers/meta/real_ada_boost'
+require 'weka/classifiers/trees/balanced_random_forest'
+
+require 'weka/filters/supervised/instance/smote'

data/lib/wikipedia.rb

@@ -0,0 +1,41 @@
+require 'open-uri'
+require 'nokogiri'
+require 'timeout'
+
+module Wikipedia
+  def self.api_base_uri
+    'https://en.wikipedia.org/w/api.php?format=xml&action=query&'
+  end
+
+  def self.param_string(params)
+    params.map { |k, v| "#{k}=#{v}" }.join('&')
+  end
+
+  # Retries to call the request in the case of Timeout errors
+  def self.request_with_retry(uri, times = 1, timeout = 5)
+    content = ''
+
+    begin
+      Timeout.timeout(timeout) do
+        content = URI.parse(uri).read
+      end
+    rescue => error
+      if times > 0
+        times -= 1
+        retry
+      else
+        warn "Requesting '#{uri}' failed multiple times.\n#{error.message}"
+      end
+    end
+
+    content
+  end
+
+  def api_request(params = {})
+    uri = URI.encode(api_base_uri + param_string(params))
+    content = request_with_retry(uri, 3)
+    Nokogiri::XML(content)
+  end
+
+  module_function :api_request
+end

data/spec/factories/edit.rb

@@ -0,0 +1,19 @@
+FactoryBot.define do
+  factory :edit, class: Wikipedia::VandalismDetection::Edit do
+    old_revision { FactoryBot.build(:old_revision) }
+    new_revision { FactoryBot.build(:new_revision) }
+    page_id { nil }
+    page_title { nil }
+
+    initialize_with { new(old_revision, new_revision, page_id: page_id, page_title: page_title) }
+  end
+
+  factory :anonymous_edit, class: Wikipedia::VandalismDetection::Edit do
+    old_revision { FactoryBot.build(:old_revision) }
+    new_revision { FactoryBot.build(:anonymous_revision) }
+    page_id { nil }
+    page_title { nil }
+
+    initialize_with { new(old_revision, new_revision, page_id: page_id, page_title: page_title) }
+  end
+end

data/spec/factories/page.rb

@@ -0,0 +1,12 @@
+FactoryBot.define do
+  factory :page, class: Wikipedia::VandalismDetection::Page do
+    id { nil }
+    title { nil }
+
+    after :build do |obj|
+      obj.add_revision FactoryBot.build(:old_revision, contributor: 'User')
+      obj.add_revision FactoryBot.build(:new_revision, contributor: 'User')
+      obj.add_revision FactoryBot.build(:even_newer_revision, contributor: 'User')
+    end
+  end
+end

data/spec/factories/revision.rb

@@ -0,0 +1,51 @@
+FactoryBot.define do
+  factory :empty_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { nil }
+    f.parent_id { nil }
+    f.timestamp { nil }
+    f.text { Wikipedia::VandalismDetection::Text.new }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+  end
+
+  factory :old_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { '1' }
+    f.parent_id { nil }
+    f.timestamp { nil }
+    f.text { Wikipedia::VandalismDetection::Text.new('text 1') }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+  end
+
+  factory :new_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { '2' }
+    f.parent_id { '1' }
+    f.timestamp { '2014-11-27T18:00:00Z' }
+    f.text { Wikipedia::VandalismDetection::Text.new('text 2') }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+  end
+
+  factory :even_newer_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { '3' }
+    f.parent_id { '2' }
+    f.timestamp { '2014-11-28T18:00:00Z' }
+    f.text { Wikipedia::VandalismDetection::Text.new('text 3') }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+  end
+
+  factory :anonymous_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { '2' }
+    f.parent_id { '1' }
+    f.timestamp { '2014-11-27T18:00:00Z' }
+    f.text { Wikipedia::VandalismDetection::Text.new('text 2') }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+    f.contributor { '127.0.0.1' }
+  end
+
+  factory :registered_revision, class: Wikipedia::VandalismDetection::Revision do |f|
+    f.id { '2' }
+    f.parent_id { '1' }
+    f.timestamp { '2014-11-27T18:00:00Z' }
+    f.text { Wikipedia::VandalismDetection::Text.new('text 2') }
+    f.comment { Wikipedia::VandalismDetection::Text.new }
+    f.contributor { '12345' }
+  end
+end

data/spec/resources/config/wikipedia-vandalism-detection.yml

@@ -0,0 +1,35 @@
+corpora:
+  base_directory: ../../../../spec/resources/corpora
+
+  training:
+    base_directory: training
+    annotations_file: annotations.csv
+    edits_file: edits.csv
+    revisions_directory: revisions
+
+  test:
+    base_directory: test
+    edits_file: edits.csv
+    revisions_directory: revisions
+    ground_truth_file: ground-truth.txt
+
+output:
+  base_directory: ../../../../spec/resources/build
+  training:
+    index_file: training_index.yml
+    arff_file: training.arff
+  test:
+    index_file: test_index.yml
+    arff_file: test.arff
+    classification_file: classification.txt
+
+features:
+  - anonymity
+  - character sequence
+  - comment length
+
+classifier:
+  type: Trees::RandomForest
+  options: -I 10
+  cross-validation-fold: 2
+  training-data-options: unbalanced

data/spec/resources/corpora/test/edits.csv

@@ -0,0 +1,8 @@
+"editid","editor","oldrevisionid","newrevisionid","diffurl","edittime","editcomment","articleid","articletitle"
+1641,"137.163.16.199",328774088,328774188,"http://en.wikipedia.org/w/index.php?diff=328774188&oldid=328774088","2009-11-30T10:23:13Z","/* Location */",100935,"Not annotated article"
+1642,"J04n",307084144,326873205,"http://en.wikipedia.org/w/index.php?diff=326873205&oldid=307084144","2009-11-20T04:42:24Z","Repairing links to disambiguation pages - [[Wikipedia:Disambiguation pages with links|You can help!]]",19490449,"The Soundstage Sessions"
+1643,"64.186.73.198",326471754,326978767,"http://en.wikipedia.org/w/index.php?diff=326978767&oldid=326471754","2009-11-20T19:32:23Z","/* Non-electric telephones */",2193804,"Invention of the telephone"
+1644,"64.186.73.198",326471754,326978767,"http://en.wikipedia.org/w/index.php?diff=326978767&oldid=326471754","2009-11-20T19:32:23Z","/* Non-electric telephones */",2193804,"Invention of the telephone"
+1647,"137.163.16.199",328774035,328774110,"http://en.wikipedia.org/w/index.php?diff=328774110&oldid=328774035","2009-11-30T10:23:13Z","/* Location */",100935,"Chinatown, Manhattan"
+1648,"137.163.16.199",328774035,328774110,"http://en.wikipedia.org/w/index.php?diff=328774110&oldid=328774035","2009-11-30T10:23:13Z","/* Location */",100935,"Chinatown, Manhattan"
+1649,"137.163.16.199",328774035,328774110,"http://en.wikipedia.org/w/index.php?diff=328774110&oldid=328774035","2009-11-30T10:23:13Z","/* Location */",100935,"Chinatown, Manhattan"

data/spec/resources/corpora/test/ground-truth.txt

@@ -0,0 +1,3 @@
+307084144 326873205 R
+326471754 326978767 dunno
+328774035 328774110 V