拆分关键字查询（SELECT/FROM/WHERE/AND/OR）

Question

大家好，下面是我一直在处理的以下代码：

var sql = "SELECT pillers, Balloons, Tacks FROM the_database_file WHERE Balloons != 'small' AND Balloons != 'large' AND Blah = 'bobby';";

$(document).ready(function() {
    var findFROM        = sql.indexOf(" FROM");
    var findWHERE       = sql.indexOf(" WHERE");
    var findAND         = sql.indexOf(" AND");
    var findOR          = sql.indexOf(" OR");
    var findSemicolon   = sql.indexOf(";"); 
    var findCountAND    = sql.match(/\AND\b/g);
    var findCountOR     = sql.match(/\OR\b/g);

    var txtSELECT       = sql.substring(0, findFROM);
    var txtFROM         = sql.substring(findFROM, findWHERE);
    var txtWHERE        = "";
    var txtAND          = "";
    var txtOR           = "";

    if (findAND != -1) {
        var _tmpPos     = 0;
        var _tmpAND     = "";

        findCountAND    = (findCountAND? findCountAND.length : 0);
        findCountOR     = (findCountOR? findCountOR.length : 0);

        for (var i = 1; i < findCountAND; i++) {
            console.log(i);
            _tmpPos = nth_occurrence(sql, ' AND', i);
            _tmpPos = findAND;
            findAND = sql.indexOf(" AND");
            _tmpAND = sql.substring(_tmpPos, findAND);
        }

        txtWHERE    = sql.substring(findWHERE, findAND);
    }

    $('#SELECT').text(txtSELECT);
    $('#FROM').text(txtFROM);
    $('#WHERE').text(txtWHERE);
    $('#test').text(findAND);
});

function nth_occurrence(string, char, nth) {
    var first_index = string.indexOf(char);
    var length_up_to_first_index = first_index + 1;

    if (nth == 1) {
        return first_index;
    } else {
        var string_after_first_occurrence = string.slice(length_up_to_first_index);
        var next_occurrence = nth_occurrence(string_after_first_occurrence, char, nth - 1);

        if (next_occurrence === -1) {
            return -1;
        } else {
            return length_up_to_first_index + next_occurrence;  
        }
    }
}

我正在尝试这样格式化：

SELECT pillers, Balloons, Tacks
FROM the_database_file
WHERE Balloons != 'small'
AND Balloons != 'large'
AND Blah = 'bobby';

但是，我陷入了试图在查询字符串中查找多个 AND 或 OR 的领域。上面的示例只有 2 个 AND 开头，但我似乎无法获得正确的代码。

nth_occurrence函数被发现HERE。

如果有人可以帮助我，那就太好了。可能只需要一个 REGEX 吗？谁知道？

JSFIDDLE 也已设置。

更新

感谢一位善良的论坛成员HERE，我能够更新我的JSFIDDLE，尝试重现他所做的事情：

[ { word: 'SELECT', index: 0, text: 'pillers, Balloons, Tacks' },
  { word: 'FROM', index: 32, text: 'the_database_file' },
  { word: 'WHERE', index: 55, text: 'Balloons != \'small\'' },
  { word: 'AND', index: 81, text: 'Balloons != \'large\'' },
  { word: 'AND', index: 105, text: 'Blah = \'bobby\'' },
  { word: ';', index: 123 } ]

但是，我的尝试并不一样。有什么帮助吗？

Answer 1

你的实现真的很棘手，我可以想象很多情况，当它解析失败时。

我建议不要重新发明轮子，而是使用现有的库来用 JS 解析 SQL：

1. SQL parser
2. simpleSqlParser
3. node-sqljs
4. ....

和其他人。有很多不同复杂性的现成解决方案。

Answer 2

找到解决方案：

Update JSFIDDLE

function parseSql(sql) {
  var found = [];

  ["SELECT", "WHERE", "FROM", "AND", "OR", ";"].forEach(function(word) {
    var idx = sql.indexOf(word);

    while(idx!==-1) {      
      found.push({word:word, index:idx});
      idx = sql.indexOf(word, idx + 1);
      keptIdx = idx;
    }
  });

  found.sort(function(x,y) { return x.index - y.index });
  found.forEach(function(x, i, xs) {
    if (i < xs.length - 1) {
      x.text = sql.substring(x.index, xs[i + 1].index).replace(xs[i].word, "").trim();
    }
  });

  return found;
}