如何从字符串中提取单词

Question

We have some logging service from an API that's having some trouble catching the database a query is accessing.

Say, a query comes like this:

Select Top 10 * From DataBase..Table

What I'd like to achieve is to get "DataBase" or the text prior to the ".." on the string.

Also, sometimes a query comes like this:

Select Top 10 * From DataBase.Schema.Table

Is it possible to get the "DataBase" string on both cases?

Here's what I've been trying to do, but I not that good in regular expressions.

([A-Z]+\.+?([A-Z])*(\.)+[A-Z])

But this matches the following (on square brackets):

Select Top 10 * From [DataBase..T]able

Thanks for the help!

Edit: This is done in C#, prior to sending the query/request to the database, we are trying to log every request this API method processes.

Answer 1

As it is mentioned in comment, parsing SQL this way is a bad idea due complexities that you can encounter. But if this is something you want to do, following should work. It will work for select or delete statements, insert/update won't work as they have different structure.

public string GetDbName(string sql)
{
    var sqlLower = sql.ToLower();
    var parts = sqlLower.Split(new string[] { "from " }, StringSplitOptions.RemoveEmptyEntries);

    if (parts.Length < 2)
    {
        //something is wrong in sql;
        //may be its not a select statement
        return null;
    }

    var tableName = parts[1]
        .Split(new char[] { ' ' }, StringSplitOptions.RemoveEmptyEntries)
        .FirstOrDefault();

    var posOfTableName = tableName.LastIndexOf('.');
    var dbName = tableName.Substring(0, posOfTableName);

    //We can return here, but it will return lower case db name parts
    //We can also get the original value as well by looking into original parameter
    var indexOfStart = sqlLower.IndexOf(dbName);
    return sql.Substring(indexOfStart, dbName.Length);

}