Adding Hunspell loading

src/main/java/org/luwrain/controls/EditSpellChecking.java

@@ -28,13 +28,20 @@ public class EditSpellChecking implements EditArea.ChangeListener
  LOG_COMPONENT = "spelling";
 
  private final Luwrain luwrain;
- private final SpellChecker checker = new SpellCheckerFactory().newChecker("ru");
- public EditSpellChecking(Luwrain luwrain )
+ private final SpellChecker checker;
+ public EditSpellChecking(Luwrain luwrain , String lang)
  {
  NullCheck.notNull(luwrain, "luwrain ");
- this .luwrain = luwrain; 
+ NullCheck.notEmpty(lang, "lang");
+ this .luwrain = luwrain;
+ this.checker = new SpellCheckerFactory().newChecker(luwrain, lang);
 }
 
+ public EditSpellChecking(Luwrain luwrain )
+ {
+ this(luwrain, "ru");
+ }
+
  @Override public void onEditChange(EditArea editArea, MarkedLines lines, HotPoint hotPoint)
  {
  final SortedMap<Integer, String> text = new TreeMap<>();

src/main/java/org/luwrain/nlp/Hunspell.java

@@ -1,6 +1,6 @@
-// * Copyright (C) 2021 Pavel Bakhvalov
+/* Copyright (C) 2021 Pavel Bakhvalov*/
 
-package org.luwrain.io.nlp;
+package org.luwrain.nlp;
 
 import dumonts.hunspell.bindings.HunspellLibrary;
 import org.bridj.Pointer;
@@ -14,27 +14,28 @@
 
 public final class Hunspell
 {
-private final Pointer<HunspellLibrary.Hunhandle> handle;
- private final Charset charset;
+ private final Pointer<HunspellLibrary.Hunhandle> handle;
+  private final Charset charset;
 
- public Hunspell(Path dictionary, Path affix) 
-{
- try {
- Pointer<Byte> aff = Pointer.pointerToCString(affix.toString());
- Pointer<Byte> dic = Pointer.pointerToCString(dictionary.toString());
- handle = HunspellLibrary.Hunspell_create(aff, dic);
- charset = Charset.forName(HunspellLibrary.Hunspell_get_dic_encoding(handle).getCString());
- if (this.handle == null) {
- throw new RuntimeException("Unable to create Hunspell instance");
- }
- } catch (UnsatisfiedLinkError e) {
- throw new RuntimeException("Could not create hunspell instance. Please note that LanguageTool supports only 64-bit platforms " +
- "(Linux, Windows, Mac) and that it requires a 64-bit JVM (Java).", e);
+ public Hunspell(String dictionary, String affix) 
+ {
+ try {
+ Pointer<Byte> aff = Pointer.pointerToCString(affix);
+ Pointer<Byte> dic = Pointer.pointerToCString(dictionary);
+ handle = HunspellLibrary.Hunspell_create(aff, dic);
+ charset = Charset.forName(HunspellLibrary.Hunspell_get_dic_encoding(handle).getCString());
+ if (this.handle == null) {
+ throw new RuntimeException("Unable to create Hunspell instance");
+ }
+ } catch (UnsatisfiedLinkError e) {
+ throw new RuntimeException("Could not create hunspell instance. Please note that LanguageTool supports only 64-bit platforms " +
+ "(Linux, Windows, Mac) and that it requires a 64-bit JVM (Java).", e);
+ }
  }
- }
 
- public boolean spell(String word) {
- if (handle == null) {
+ public boolean spell(String word)
+ {
+ if (handle == null) {
  throw new RuntimeException("Attempt to use hunspell instance after closing");
  }
  @SuppressWarnings("unchecked")

src/main/java/org/luwrain/nlp/SpellChecker.java

@@ -22,4 +22,5 @@
 public interface SpellChecker
 {
  List<SpellProblem> check(String text);
+ List<String> suggestCorrections(String word);
 }

src/main/java/org/luwrain/nlp/SpellCheckerFactory.java

@@ -16,19 +16,20 @@
 
 package org.luwrain.nlp;
 
+import org.luwrain.core.*;
 import org.luwrain.nlp.ru.spell.RuSpellChecker;
 
 public final class SpellCheckerFactory
 {
  static private RuSpellChecker ru = null;
 
- public SpellChecker newChecker(String lang)
+ public SpellChecker newChecker(Luwrain luwrain, String lang)
  {
  switch(lang.toUpperCase())
  {
  case "RU": {
  if (ru == null)
- ru = new RuSpellChecker();
+ ru = new RuSpellChecker(luwrain);
  return ru;
  }
  default:

src/main/java/org/luwrain/nlp/SpellText.java

@@ -81,6 +81,16 @@ public List<List<LineMarks.Mark>> buildMarks()
  return res;
  }
 
+ public List<Fragment> getFragments()
+ {
+ return fragments;
+ }
+
+ public List<SpellProblem> getProblems()
+ {
+ return problems;
+ }
+
  static public final class Fragment
  {
  final int posFrom, posTo;

src/main/java/org/luwrain/nlp/ru/spell/RuSpellChecker.java

@@ -30,10 +30,23 @@
 public final class RuSpellChecker implements SpellChecker
 {
  final JLanguageTool langTool;
+ final Hunspell hunspell;
 
- public RuSpellChecker()
+ public RuSpellChecker(Luwrain luwrain)
  {
  this.langTool = new JLanguageTool(new Russian());
+ final File
+ hunspellDataDir = new File(luwrain.getFileProperty(Luwrain.PROP_DIR_DATA), "hunspell"),
+ dictFile = new File(hunspellDataDir, "ru.dict"),
+ affFile = new File(hunspellDataDir, "ru.aff");
+ this.hunspell = new Hunspell(dictFile.getAbsolutePath(), affFile.getAbsolutePath());
+ }
+
+ //Just for unit tests
+RuSpellChecker()
+ {
+ this.langTool = new JLanguageTool(new Russian());
+ this.hunspell = null;
  }
 
  @Override public List<SpellProblem> check(String text)
@@ -50,4 +63,11 @@ public RuSpellChecker()
  throw new RuntimeException(e);
  }
  }
+
+ @Override public List<String> suggestCorrections(String word)
+ {
+ return null;
+ }
+
+
 }

src/test/java/org/luwrain/io/nlp/ru/spell/RuSpellCheckerTest.java

@@ -20,23 +20,47 @@
 //https://languagetool.org/development/api/org/languagetool/rules/spelling/SpellingCheckRule.html
 
 
-import java.io.*;
-import java.net.*;
+import java.util.*;
 
 import org.junit.*;
 
 import org.luwrain.core.*;
+import org.luwrain.nlp.*;
 
 public class RuSpellCheckerTest extends Assert
 {
- private RuSpellChecker c = null;
+ private RuSpellChecker checker = null;
 
- @Test public void main()
+  @Test public void shortSent()
  {
+ final String text = "Это было превосходная день!";
+ final SpellText t = new SpellText(new String[]{text}, checker);
+ assertNotNull(t.getFragments());
+ assertEquals(1, t.getFragments().size());
+ assertNotNull(t.getProblems());
+ assertEquals(1, t.getProblems().size());
+ final List<List<LineMarks.Mark>> res = t.buildMarks();
+ assertNotNull(res);
+ assertEquals(1, res.size());
+ List<LineMarks.Mark> marks = res.get(0);
+ assertNotNull(marks);
+ assertEquals(1, marks.size());
+ final LineMarks.Mark mark = marks.get(0);
+ assertNotNull(mark);
+ final SpellProblem problem = (SpellProblem)mark.getMarkObject();
+ assertNotNull(problem);
+ assertEquals("Прилагательное не согласуется с существительным по роду.", problem.getComment());
+ assertEquals("", problem.getShortComment());
+ final int
+ start = problem.getStart(),
+ end = problem.getEnd();
+ assertEquals("превосходная день", text.substring(start, end));
  }
 
+
+
  @Before public void create()
  {
- c = new RuSpellChecker();
+ checker = new RuSpellChecker();
  }
 }