package org.wikimedia.search.extra.analysis.turkish;
   
   import static org.junit.Assert.assertEquals;
   
   import java.util.Arrays;
   import java.util.Collection;
   
   import org.junit.Test;
   import org.junit.runner.RunWith;
  import org.junit.runners.Parameterized;
  
  @RunWith(Parameterized.class)
  public class BetterApostropheTest {
      private String input;
      private String expected;
      private BetterApostrophe apostrophe = new BetterApostrophe();
  
      public BetterApostropheTest(String input, String expected) {
          this.input = input;
          this.expected = expected;
      }
  
      @Parameterized.Parameters
      public static Collection<Object[]> apostropheTestCases() {
          return Arrays.asList(new Object[][]{
              // simple single and mult-apostrophe examples
              {"türkiye'den", "türkiye"},
              {"xi'an'in", "xian"},
              {"yefâ’î’nin", "yefâî"},
              {"vak‘ası’nın", "vakası"},
              {"k'at'aph'ia", "kataph"},
              {"hawai'i'o'o", "hawaiio"},
              {"k'oyitl'ots'ina", "koyitlots"},
              {"isnâ‘aşer’îyye'nin", "isnâaşerîyye"},
  
              // apostrophe-like characters
              // note that other tests also include apostrophe-like characters other than '
              {"türkiye'den", "türkiye"}, // apostrophe
              {"türkiyeʼden", "türkiye"}, // modifier apostrophe
              {"türkiye＇den", "türkiye"}, // fullwidth apostrophe
              {"türkiye‘den", "türkiye"}, // left curly quote
              {"türkiye’den", "türkiye"}, // right curly quote
              {"türkiye`den", "türkiye"}, // grave accent
              {"türkiye´den", "türkiye"}, // acute accent
              {"türkiyeˋden", "türkiye"}, // modifier grave accent
              {"türkiyeˊden", "türkiye"}, // modifier acute accent
  
              // testWholeWordExceptions
              // whole word exceptions: things that are easier to process as whole words
              {"qu'il", "il"},
              {"s'il", "il"},
              {"d'un", "un"},
              {"l'un", "un"},
              {"qu'un", "un"},
  
              // special case—kuran/quran/etc.
              {"kur'ân", "kurân"},
              {"kur'an'daki", "kuran"},
              {"kurʼân'da", "kurân"},
              {"kur'an'dır", "kuran"},
              {"qur'anic", "quranic"},
              {"qur’ān", "qurān"},
  
              // special case—English -n't
              {"needn't", "neednt"},
              {"shan't", "shant"},
              {"shouldn’t", "shouldnt"},
              {"wasn't", "wasnt"},
  
              // special case—English -'n'-
              {"drum'n'bass", "drumnbass"},
              {"nice’n’easy", "niceneasy"},
              {"r'n'b", "rnb"},
              {"rock'n'roll", "rocknroll"},
  
              // special case—English -'s + ' (+ tr suffix)
              {"dalek's'de", "dalek"},
              {"mcdonald's'ında", "mcdonald"},
              {"mcvitie's'nin", "mcvitie"},
              {"scott's’da", "scott"},
  
              // very French/Italian elision
              {"j't'aime", "aime"},
              {"j'n'attends", "attends"},
              {"j'étais", "étais"},
              {"l'écologie", "écologie"},
              {"qu’aucun", "aucun"},
              {"sull'uscio", "uscio"},
  
              // looks like elision and Turkish suffixation at the same time--suffixation wins!
              {"j'den", "j"},
              {"d'deki", "d"},
              {"un'un", "un"},
              {"all'daki", "all"},
              {"l'ı", "l"},
  
              // strip multiple common Turkish suffixes used at once
              {"alp'lerindeki", "alp"},
              {"ceo'luklarını", "ceo"},
             {"lacan'ınkilerden", "lacan"},
             {"ankara'dakilerin", "ankara"},
             {"amerika'dakileri", "amerika"},
             {"plüton'unkindense", "plüton"},
             {"profesör'lerindendir", "profesör"},
             {"archaeopteryx'inkilere", "archaeopteryx"},
 
             // Turkish suffix stripping should work on non-Latin words, too
             {"πάπυρος'tan", "πάπυρος"},
             {"ребро'dan", "ребро"},
             {"Աշտարակ'in", "Աշտարակ"},
             {"قاعدة‎'nin", "قاعدة‎"},
 
             // generally disfavor one-letter "stems"
             {"a'oulhalak", "aoulhalak"},
             {"b'day", "bday"},
             {"c'hwennenn", "chwennenn"},
             {"e'cole", "ecole"},
             {"g'day", "gday"},
             {"g‘azalkent", "gazalkent"},
             {"y'all", "yall"},
 
             // non-Turkish letters can't be in Turkish suffixes, so don't treat as suffixes
             // numbers
             {"00'19", "0019"},
 
             // non-letters
             {"albʊ'raːq", "albʊraːq"},
             {"a·lü·mi'n·yum", "a·lü·min·yum"},
 
             // non-Turkish Latin
             {"awa’uq", "awauq"},
             {"arc’teryx", "arcteryx"},
 
             // non-Turkish diacritics
             {"ba'aṯ", "baaṯ"},
             {"bābā’ī", "bābāī"},
             {"abdülkerim'ê", "abdülkerimê"},
 
             // non-Latin
             {"В’в", "Вв"},
             {"Х’агәы́шь", "Хагәы́шь"},
             {"прем'єра", "премєра"},
             {"ג'אלה", "גאלה"},
             {"여성들'에", "여성들에"},
             {"կ’ընթրենք", "կընթրենք"},
             {"επ'ευκαιρία", "επευκαιρία"},
 
             // a few two-letter prefixes that are almost never stems
             {"ch'alla", "challa"},
             {"ch'ang", "chang"},
             {"ch'ing", "ching"},
             {"ma'arif", "maarif"},
             {"ma'rifette", "marifette"},
             {"ma'ruflardır", "maruflardır"},
             {"ta'izz", "taizz"},
             {"ta'rikh", "tarikh"},
             {"ta'us", "taus"},
             {"te'lif", "telif"},
             {"te'vîl", "tevîl"},
             {"te'mine", "temine"},
 
             // various tests of rule ordering
 
             // remove common Fr/It elision before removing words that look like multiple
             // (admittedly possibly nonsensical) Turkish suffixes
             {"d'adieu", "adieu"}, // a + di + e + u
             {"l'ındiana", "ındiana"}, // ın + di + a + na
             {"dell'ıtalia", "ıtalia"}, // ı + ta + li + a
 
             // remove common Fr/It elision before removing multiple apostrophes
             {"dell'arte'nin", "arte"}, // not dellarte
             {"nell'emilia'da", "emilia"}, // not nellemilia
             {"d'artagnan'ın", "artagnan"}, // not dartagnan
 
             // remove common endings before removing one letter before apostrophe
             {"b'dekilere", "b"},
             {"n'ın", "n"},
             {"s'inkilere", "s"},
             {"x'dedir", "x"},
             {"β'ların", "β"},
             {"ϖ'yi", "ϖ"},
             {"ж’den", "ж"},
             {"й'dir", "й"},
 
             // non-word prefixes + clear Turkish suffix -> strip suffix
             {"ch'den", "ch"},
             {"ma'ın", "ma"},
             {"te'de", "te"},
             {"ta'lik", "ta"},
 
             // interaction of non-word stems and other endings
             {"ch'orti's", "chorti"},
             {"ch'ing'in", "ching"},
             {"ma'ali'yi", "maali"},
             {"ma'arretü'n", "maarretü"},
             {"ta'us'un", "taus"},
             {"ta'rifâti'l", "tarifâti"},
             {"te'vilâti'l", "tevilâti"},
             {"te’lifi’l", "telifi"},
 
             // words with non-Turkish Latin or non-Latin characters and no apostrophes should
             // be unchanged
             {"año", "año"}, // Spanish
             {"вищій", "вищій"}, // Ukrainian
             {"위키백과", "위키백과"}, // Korean
             {"əliağa", "əliağa"}, // Azerbaijani
             {"ውክፔዲያ", "ውክፔዲያ"}, // Amharic
             {"ᐅᐃᑭᐱᑎᐊ", "ᐅᐃᑭᐱᑎᐊ"}, // Inuktitut
             {"ვიკიპედია", "ვიკიპედია"}, // Georgian
             {"βικιπαίδεια", "βικιπαίδεια"}, // Greek
             {"аблютомания", "аблютомания"}, // Russian
 
             // Exhaustive test of "common Turkish suffixes", part I. They only get stripped as
             // suffixes after elision prefixes, one-letter stems, or "non-word" stems, so not
             // all examples are easily found. This first batch all come from Turkish Wikipedia.
             {"g’a", "g"},
             {"ş'da", "ş"},
             {"all'daki", "all"},
             {"ı'dan", "ı"},
             {"dell'de", "dell"},
             {"m'deki", "m"},
             {"p’den", "p"},
             {"nell'dir", "nell"},
             {"c'dur", "c"},
             {"w'dı", "w"},
             {"k'dır", "k"},
             {"ı’e", "ı"},
             {"ç'i", "ç"},
             {"b’il", "b"},
             {"v'in", "v"},
             {"v'la", "v"},
             {"a’lar", "a"},
             {"f'le", "f"},
             {"x'ler", "x"},
             {"h'li", "h"},
             {"m'lik", "m"},
             {"c’lu", "c"},
             {"n'luk", "n"},
             {"k’lı", "k"},
             {"w’lık", "w"},
             {"o'na", "o"},
             {"o'ndaki", "o"},
             {"o’ndan", "o"},
             {"π'nin", "π"},
             {"o’nu", "o"},
             {"u'nun", "u"},
             {"μ'nün", "μ"},
             {"å'nın", "å"},
             {"f'si", "f"},
             {"o'su", "o"},
             {"h’sı", "h"},
             {"w'ta", "w"},
             {"s'tan", "s"},
             {"v'te", "v"},
             {"x'teki", "x"},
             {"h'ten", "h"},
             {"v’ti", "v"},
             {"v'tir", "v"},
             {"v’tur", "v"},
             {"w'tır", "w"},
             {"k'u", "k"},
             {"all'un", "all"},
             {"δ'ya", "δ"},
             {"b'ydi", "b"},
             {"ê’ye", "ê"},
             {"φ’yi", "φ"},
             {"b'yken", "b"},
             {"u'yla", "u"},
             {"j'yle", "j"},
             {"u'yu", "u"},
             {"ü'yü", "ü"},
             {"γ'yı", "γ"},
             {"l'ü", "l"},
             {"n'ı", "n"},
             {"ψ'ın", "ψ"},
 
             // Exhaustive test of "common Turkish suffixes", part II. These are "synthetic"
             // examples, x, plus an apostrophe-like character, plus a suffix. The suffixes all
             // occur in Turkish Wikipedia with apostrophe-like characters, but not necessarily
             // after one-letter stems.
             {"x'di", "x"},
             {"xʼdu", "x"},
             {"x’dü", "x"},
             {"x＇dür", "x"},
             {"x’ken", "x"},
             {"x'ki", "x"},
             {"x‘lü", "x"},
             {"x'lük", "x"},
             {"xʼnda", "x"},
             {"x'nde", "x"},
             {"xʼndeki", "x"},
             {"x'nden", "x"},
             {"x＇ne", "x"},
             {"x'ni", "x"},
             {"x’nü", "x"},
             {"x'nı", "x"},
             {"x'sa", "x"},
             {"x’se", "x"},
             {"xʼsü", "x"},
             {"x'taki", "x"},
             {"x'tu", "x"},
             {"x'tü", "x"},
             {"x'tür", "x"},
             {"x'tı", "x"},
             {"x’ul", "x"},
             {"x’ydu", "x"},
             {"x'ydü", "x"},
             {"x'ydı", "x"},
             {"x‘ül", "x"},
             {"x'ün", "x"},
 
         });
     }
 
     @Test
     public void apostropheTester() throws Exception {
         assertEquals(expected, apostrophe.apos(input));
     }
 
 }