You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
88 lines
3.1 KiB
88 lines
3.1 KiB
import java.util.ArrayList;
|
|
import java.util.List;
|
|
|
|
import data.Sentence;
|
|
import data.Word;
|
|
|
|
public class Common {
|
|
|
|
public static List<Sentence> corpus;
|
|
public static List<Sentence> minCorpus;
|
|
public static List<Sentence> midCorpus;
|
|
public static List<Sentence> midCorpusSkip;
|
|
public static List<Sentence> josTest;
|
|
|
|
static {
|
|
Sentence testSentence;
|
|
|
|
// full sentence
|
|
ArrayList<String> taxonomy = new ArrayList<>();
|
|
taxonomy.add("#Ft.Z.N.N");
|
|
List<Word> words = new ArrayList<>();
|
|
words.add(new Word("ker", "ker", "Vd", taxonomy));
|
|
words.add(new Word("ima", "imeti", "Ggnste-n", taxonomy));
|
|
words.add(new Word("junak", "junak", "Somei", taxonomy));
|
|
words.add(new Word("v", "v", "Dm", taxonomy));
|
|
words.add(new Word("posesti", "posest", "Sozem", taxonomy));
|
|
words.add(new Word("nekaj", "nekaj", "Rsn", taxonomy));
|
|
words.add(new Word("o", "o", "Dm", taxonomy));
|
|
words.add(new Word("čemer", "kar", "Zz-sem", taxonomy));
|
|
words.add(new Word("se", "se", "Zp------k", taxonomy));
|
|
words.add(new Word("mu", "on", "Zotmed--k", taxonomy));
|
|
words.add(new Word("ne", "ne", "L", taxonomy));
|
|
words.add(new Word("sanja", "sanjati", "Ggnste", taxonomy));
|
|
words.add(new Word("a", "a", "Vp", taxonomy));
|
|
words.add(new Word("se", "se", "Zp------k", taxonomy));
|
|
words.add(new Word("onemu", "oni", "Zk-sed", taxonomy));
|
|
words.add(new Word("zdi", "zdeti", "Ggnste", taxonomy));
|
|
words.add(new Word("ključno", "ključen", "Ppnsei", taxonomy));
|
|
words.add(new Word("pri", "pri", "Dm", taxonomy));
|
|
words.add(new Word("operaciji", "operacija", "Sozem", taxonomy));
|
|
words.add(new Word("666", "666", "Kag", taxonomy));
|
|
|
|
testSentence = new Sentence(words, "#Ft.Z.N.N");
|
|
corpus = new ArrayList<>();
|
|
corpus.add(testSentence);
|
|
|
|
// three word sentence
|
|
testSentence = new Sentence(corpus.get(0).getSublist(0, 3), "#Ft.Z.N.N");
|
|
minCorpus = new ArrayList<>();
|
|
minCorpus.add(testSentence);
|
|
|
|
// five word sentence
|
|
words = new ArrayList<>();
|
|
words.add(new Word("ker", "ker", "Vd", taxonomy));
|
|
words.add(new Word("ima", "imeti", "Ggnste-n", taxonomy));
|
|
words.add(new Word("junak", "junak", "Somei", taxonomy));
|
|
words.add(new Word("ima", "imeti", "Ggnste-n", taxonomy));
|
|
words.add(new Word("posesti", "posest", "Sozem", taxonomy));
|
|
testSentence = new Sentence(words, "#Ft.Z.N.N");
|
|
|
|
midCorpus = new ArrayList<>();
|
|
midCorpus.add(testSentence);
|
|
|
|
// five word sentence - for skipgrams
|
|
words = new ArrayList<>();
|
|
words.add(new Word("ker", "ker", "Vd", taxonomy));
|
|
words.add(new Word("ima", "imeti", "Ggnste-n", taxonomy));
|
|
words.add(new Word("junak", "junak", "Somei", taxonomy));
|
|
words.add(new Word("v", "v", "Dm", taxonomy));
|
|
words.add(new Word("posesti", "posest", "Sozem", taxonomy));
|
|
testSentence = new Sentence(words, "#Ft.Z.N.N");
|
|
|
|
midCorpusSkip = new ArrayList<>();
|
|
midCorpusSkip.add(testSentence);
|
|
|
|
// JOS test
|
|
words = new ArrayList<>();
|
|
words.add(new Word("junak", "junak", "Somei", taxonomy));
|
|
words.add(new Word("ima", "imeti", "Ggnste-n", taxonomy));
|
|
words.add(new Word("posesti", "posest", "Sozem", taxonomy));
|
|
testSentence = new Sentence(words, "#Ft.Z.N.N");
|
|
|
|
josTest = new ArrayList<>();
|
|
josTest.add(testSentence);
|
|
}
|
|
|
|
}
|