Question

我正在寻找一种方法来清理传递给弹性搜索的搜索词，即逃避所有控制字符。类似于this answer中Ruby所描述的内容。 Scala有这样的事吗？

Answer 1

我已将this answer中找到的ruby解决方案翻译成Scala：

ListBox item

以下是测试：

package util

import java.util.regex.Pattern

trait ElasticSearchSanitizer {
  /** Sanitizes special characters and set operators in elastic search search-terms. */
  def sanitize(term: String): String = (
    escapeSpecialCharacters _ andThen
    escapeSetOperators andThen
    collapseWhiteSpaces andThen
    escapeOddQuote
  )(term)

  private def escapeSpecialCharacters(term: String): String = {
    val escapedCharacters = Pattern.quote("""\/+-&|!(){}[]^~*?:""")
    term.replaceAll(s"([$escapedCharacters])", "\\\\$1")
  }

  private def escapeSetOperators(term: String): String = {
    val operators = Set("AND", "OR", "NOT")
    operators.foldLeft(term) { case (accTerm, op) =>
      val escapedOp = escapeEachCharacter(op)
      accTerm.replaceAll(s"""\\b($op)\\b""", escapedOp)
    }
  }

  private def escapeEachCharacter(op: String): String =
    op.toCharArray.map(ch => s"""\\\\$ch""").mkString

  private def collapseWhiteSpaces(term: String): String = term.replaceAll("""\s+""", " ")

  private def escapeOddQuote(term: String): String = {
    if (term.count(_ == '"') % 2 == 1) term.replaceAll("""(.*)"(.*)""", """$1\\"$2""") else term
  }
}

Answer 2

为此提供的库可能会更好地使用

import org.apache.lucene.queryparser.classic.QueryParserBase
....    
val escapedQ = QueryParserBase.escape(rawQuery)

在Scala中是否有针对Elasticsearch的搜索术语清理程序的实现？

2 个答案: