package gw.solr.mapper.query

uses org.apache.commons.lang.CharUtils
uses java.lang.Character


/**
 * Simple search value encoder that converts "special" characters to
 * whitespace and then compresses each whitespace block into a single
 * 'space.
 * 
 * Use for encoding term queries, such as roles or phone numbers.
 * For edismax subqueries, use PhraseEncoder.
 */
@ReadOnly
class DefaultEncoder extends AbstractEncoder {

  /**
   * Encode unsupported characters as ' ' (whitespace) for term queries.
   */
  public final static var ENCODED_CHAR : char = ' '

  override function isSearchable(c : char) : boolean {
    return 
        Character.isLetterOrDigit(c) ||
        // Ascii symbols except double quotes
        (CharUtils.isAsciiPrintable(c) && !Character.isWhitespace(c) && c != DOUBLE_QUOTE) ||
        Character.isLowSurrogate(c) ||
        Character.isHighSurrogate(c)
  }
  
  override property get EncodedChar() : char {
    return ENCODED_CHAR
  }

  override function needsEscape(c : char) : boolean {
    return LUCENE_SPECIAL_CHARS.contains(c)
  }

}