GroongaStringUtils, for extraction of matched terms on the client side.

GroongaStringUtils.js

/**
 * Usage:
 *   var matcher = GroongaStringUtils.getTermsMatcher('検索 OR けんさく');
 *   // => /(a pattern to match all search terms in the given query)/g
 *   document.body.innerHTML = document.body.innerHTML
 *     .replace(matcher, function(term) { return '<em>' + term + '</em>'; });
 *
 * License: The MIT License
 * Copyright (c) 2013 ClearCode Inc.
 */

var GroongaStringUtils = {
  getTermsMatcher: function(source) {
    var terms = this.extractTerms(source);
    terms = terms.map(this.expandSimilarCharacters, this);
    var matcher = '(' + terms.join('|') + ')';
    return new RegExp(matcher, 'gi');
  },

  // http://groonga.org/ja/docs/reference/grn_expr/query_syntax.html
  extractTerms: function(source) {
    var terms = [];
    var escaped = false;
    var doubleQuoted = false;
    var singleQuoted = false;
    var lastTerm = '';
    source.replace(/\s+/g, ' ').trim().split('').forEach(function(character) {
      if (escaped) {
        lastTerm += character;
        escaped = false;
        return;
      }

      switch (character) {
        case '\\':
          escaped = true;
          return;
          
        case '(':
          if (doubleQuoted || singleQuoted) {
            lastTerm += character;
          } else {
            groupLevel++;
          }
          return;

        case ')':
          if (doubleQuoted || singleQuoted) {
            lastTerm += character;
          } else {
            groupLevel--;
          }
          return;

        case '"':
          if (doubleQuoted) {
            terms.push(lastTerm);
            lastTerm = '';
            doubleQuoted = false;
          } else {
            doubleQuoted = true;
          }
          return;

        case "'":
          if (singleQuoted) {
            terms.push(lastTerm);
            lastTerm = '';
            singleQuoted = false;
          } else {
            singleQuoted = true;
          }
          return;

        case " ":
          if (doubleQuoted || singleQuoted) {
            lastTerm += character;
          } else {
            terms.push(lastTerm);
            lastTerm = '';
          }
          return;

        default:
          lastTerm += character;
          return;
      }
    });
    if (lastTerm)
      terms.push(lastTerm);

    terms = terms.filter(function(term) {
      return term && term != 'OR' && term != '+' && term != '-';
    }).map(function(term) {
      return term.replace(/^[^:]+:(?:[@^$!<>]|<=|>=)?/, '');
    });

    return terms.filter(function(term) {
      return term;
    });
  },

  hankakuZenkakuCharacters: {
    'ｱ': ['ア'],
    'ｲ': ['イ'],
    'ｳ': ['ウ'],
    'ｴ': ['エ'],
    'ｵ': ['オ'],
    'ｶ': ['カ'],
    'ｷ': ['キ'],
    'ｸ': ['ク'],
    'ｹ': ['ケ'],
    'ｺ': ['コ'],
    'ｻ': ['サ'],
    'ｼ': ['シ'],
    'ｽ': ['ス'],
    'ｾ': ['ケ'],
    'ｿ': ['ソ'],
    'ﾀ': ['タ'],
    'ﾁ': ['チ'],
    'ﾂ': ['ツ'],
    'ﾃ': ['テ'],
    'ﾄ': ['ト'],
    'ﾅ': ['ナ'],
    'ﾆ': ['ニ'],
    'ﾇ': ['ヌ'],
    'ﾈ': ['ネ'],
    'ﾉ': ['ノ'],
    'ﾊ': ['ハ'],
    'ﾋ': ['ヒ'],
    'ﾌ': ['フ'],
    'ﾍ': ['ヘ'],
    'ﾎ': ['ホ'],
    'ﾏ': ['マ'],
    'ﾐ': ['ミ'],
    'ﾑ': ['ム'],
    'ﾒ': ['メ'],
    'ﾓ': ['モ'],
    'ﾔ': ['ヤ'],
    'ﾕ': ['ユ'],
    'ﾖ': ['ヨ'],
    'ﾗ': ['ラ'],
    'ﾘ': ['リ'],
    'ﾙ': ['ル'],
    'ﾚ': ['レ'],
    'ﾛ': ['ロ'],
    'ﾜ': ['ワ'],
    'ｦ': ['ヲ'],
    'ﾝ': ['ン'],
    'ｧ': ['ァ'],
    'ｨ': ['ィ'],
    'ｩ': ['ゥ'],
    'ｪ': ['ェ'],
    'ｫ': ['ォ'],
    'ｯ': ['ッ'],
    'ﾞ': ['゛'],
    'ﾟ': ['゜'],
    'ｶﾞ': ['ガ', 'カ゛'],
    'ｷﾞ': ['ギ', 'キ゛'],
    'ｸﾞ': ['グ', 'ク゛'],
    'ｹﾞ': ['ゲ', 'ケ゛'],
    'ｺﾞ': ['ゴ', 'コ゛'],
    'ｻﾞ': ['ザ', 'サ゛'],
    'ｼﾞ': ['ジ', 'シ゛'],
    'ｽﾞ': ['ズ', 'ス゛'],
    'ｾﾞ': ['ゼ', 'セ゛'],
    'ｿﾞ': ['ゾ', 'ソ゛'],
    'ﾀﾞ': ['ダ', 'タ゛'],
    'ﾁﾞ': ['ヂ', 'チ゛'],
    'ﾂﾞ': ['ヅ', 'ツ゛'],
    'ﾃﾞ': ['デ', 'テ゛'],
    'ﾄﾞ': ['ド', 'ト゛'],
    'ﾊﾞ': ['バ', 'ハ゛'],
    'ﾋﾞ': ['ビ', 'ヒ゛'],
    'ﾌﾞ': ['ブ', 'フ゛'],
    'ﾍﾞ': ['ベ', 'ヘ゛'],
    'ﾎﾞ': ['ボ', 'ホ゛'],
    'ﾊﾟ': ['パ', 'ハ゜'],
    'ﾋﾟ': ['ピ', 'ヒ゜'],
    'ﾌﾟ': ['プ', 'フ゜'],
    'ﾍﾟ': ['ペ', 'ヘ゜'],
    'ﾎﾟ': ['ポ', 'ホ゜']
  },

  get hankakuZenkakuPatterns() {
    if (!this._hankakuZenkakuPatterns) {
      this._hankakuZenkakuPatterns = {};
      Object.keys(this.hankakuZenkakuCharacters).forEach(function(hankaku) {
        var zenkakus = this.hankakuZenkakuCharacters[hankaku];
        var pattern;
        if (hankaku.length > 1 ||
            zenkakus.some(function(zenkaku) { return zenkaku.length > 1; })) {
          pattern = '(?:' + hankaku + '|' + zenkakus.join('|') + ')';
        } else {
          pattern = '[' + hankaku + zenkakus.join('') + ']';
        }
        this._hankakuZenkakuPatterns[hankaku] = pattern;
      }, this);
    }
    return this._hankakuZenkakuPatterns;
  },
  _hankakuZenkakuPatterns: null,

  get zenkakuHankakuCharacters() {
    if (!this._zenkakuHankakuCharacters) {
      this._zenkakuHankakuCharacters = {};
      Object.keys(this.hankakuZenkakuCharacters).forEach(function(hankaku) {
        var zenkaku = this.hankakuZenkakuCharacters[hankaku][0];
        this._zenkakuHankakuCharacters[zenkaku] = hankaku;
      }, this);
    }
    return this._zenkakuHankakuCharacters;
  },
  _zenkakuHankakuCharacters: null,

  get hankakuMatcher() {
    if (!this._hankakuMatcher) {
      var patterns = [];
      Object.keys(this.hankakuZenkakuCharacters).forEach(function(hankaku) {
        patterns.push(hankaku);
      }, this);
      this._hankakuMatcher = new RegExp('(' + patterns.reverse().join('|') + ')', 'g');
    }
    return this._hankakuMatcher;
  },
  _hankakuMatcher: null,

  get zenkakuMatcher() {
    if (!this._zenkakuMatcher) {
      var patterns = [];
      Object.keys(this.hankakuZenkakuCharacters).forEach(function(hankaku) {
        var zenkakus = this.hankakuZenkakuCharacters[hankaku];
        patterns = patterns.concat(zenkakus);
      }, this);
      this._zenkakuMatcher = new RegExp('(' + patterns.reverse().join('|') + ')', 'g');
    }
    return this._zenkakuMatcher;
  },
  _zenkakuMatcher: null,

  zenkaku2hankaku: function(string) {
    var me = this;
    string = string.replace(/[０-９ａ-ｚＡ-Ｚ]/gi, this.zenkaku2hankakuAscii);
    string = string.replace(this.zenkakuMatcher, function(matched) {
      return me.zenkakuHankakuCharacters[matched];
    });
    return string;
  },
  zenkaku2hankakuAscii: function(character) {
    return String.fromCharCode(character.charCodeAt(0) - 0xFEE0);
  },

  hankaku2zenkaku: function(string) {
    var me = this;
    string = string.replace(/[0-9a-z]/gi, this.hankaku2zenkakuAscii);
    string = string.replace(this.hankakuMatcher, function(matched) {
      return me.hankakuZenkakuCharacters[matched][0];
    });
    return string;
  },
  hankaku2zenkakuAscii: function(character) {
    return String.fromCharCode(character.charCodeAt(0) + 0xFEE0);
  },

  expandSimilarCharacters: function(term) {
    var me = this;
    // step 1: normalize to hankaku
    term = this.zenkaku2hankaku(term);
    // step 2: expand to hankaku and zenkaku
    term = term.replace(/[A-Za-z0-9]/g, function(matched) {
      return '[' + matched + me.hankaku2zenkakuAscii(matched) + ']';
    });
    term = term.replace(this.hankakuMatcher, function(matched) {
      return me.hankakuZenkakuPatterns[matched];
    });
    return term;
  }
};