package com.acumenvelocity.ath.common;
   
   import com.ibm.icu.text.BreakIterator;
   import net.sf.okapi.common.LocaleId;
   
   import java.util.ArrayList;
   import java.util.Collections;
   import java.util.List;
   import java.util.TreeSet;
  import java.util.concurrent.ConcurrentHashMap;
  
  /**
   * Returns positions where inline codes can be safely inserted in text.
   * Uses ICU4J word breaking for proper language support, augmented with
   * positions after punctuation and whitespace.
   * 
   * Includes:
   * - Position 0 (start of text)
   * - All ICU4J word boundaries (language-aware)
   * - After each punctuation and whitespace character
   * - Position text.length() (end of text)
   *
   * <pre>
   * "Hello, world!" → [0, 5, 6, 7, 12, 13]
   * "日本語のテスト" → [0, 3, 4, 7] (CJK word breaks)
   * </pre>
   */
  public final class OkapiWordBreaker {
  
    private static final ConcurrentHashMap<String, BreakIterator> WORD_CACHE = new ConcurrentHashMap<>();
  
    private OkapiWordBreaker() {
    }
  
    public static List<Integer> getWordBreakPositions(String text, LocaleId locId) {
      if (text == null || text.isEmpty()) {
        return Collections.emptyList();
      }
      if (locId == null) {
        throw new NullPointerException("LocaleId must not be null");
      }
  
      com.ibm.icu.util.ULocale uLocale = locId.toIcuLocale();
      String cacheKey = uLocale.toString();
  
      BreakIterator wordBreaker = WORD_CACHE.computeIfAbsent(cacheKey,
          k -> BreakIterator.getWordInstance(uLocale));
  
      TreeSet<Integer> positions = new TreeSet<>();
      
      // Always add start and end positions
      positions.add(0);
      positions.add(text.length());
      
      // Add all ICU4J word boundaries (handles all languages properly)
      synchronized (wordBreaker) {
        wordBreaker.setText(text);
        for (int boundary = wordBreaker.first();
             boundary != BreakIterator.DONE;
             boundary = wordBreaker.next()) {
          positions.add(boundary);
        }
      }
      
      // Additionally add positions after punctuation and whitespace
      // This ensures inline codes can be placed around each punctuation mark
      for (int i = 0; i < text.length(); i++) {
        char ch = text.charAt(i);
        if (Character.isWhitespace(ch) || isPunctuation(ch)) {
          positions.add(i + 1);
        }
      }
  
      return new ArrayList<>(positions);
    }
  
    /**
     * Checks if a character is punctuation using Unicode categories.
     */
    private static boolean isPunctuation(char ch) {
      int type = Character.getType(ch);
      return type == Character.DASH_PUNCTUATION
          || type == Character.START_PUNCTUATION
          || type == Character.END_PUNCTUATION
          || type == Character.CONNECTOR_PUNCTUATION
          || type == Character.OTHER_PUNCTUATION
          || type == Character.INITIAL_QUOTE_PUNCTUATION
          || type == Character.FINAL_QUOTE_PUNCTUATION;
    }
  }