Parsing a custom structured query language?

Question

I'm currently working on developing a simple Query language for basic CRUD functions, it is console based and the user will enter a query in the following form:

FIND WHERE NAME nametofind

and the language will search a vector of items and return the item if it's found. I have developed the following for parsing this query.

int SDIQL::parseQuery(std::string query)
{
    std::vector<std::string> result_items;

    std::istringstream ss(query);
    std::string token;

    while(std::getline(ss, token, ' ')) {
        //std::cout << token << '\n';
        result_items.push_back(token);
    }

    if(result_items[0] == "ADD" || result_items[0] == "INSERT")
    {
        if(result_items[1] != "")
        {
            if(result_items[2] == "WITH")
            {
                if(result_items[3] != "")
                {
                    ADDquery(result_items[1], result_items[3]);
                }
                else
                {
                    std::cout<<"No value entered, please try again."<<std::endl;
                }
            }
            else
            {
                std::cout<<"Invalid paramater entered, expected WITH, please try again."<<std::endl;
            }
        }
        else
        {
            std::cout<<"No Key Specified, please try again."<<std::endl;
        }
    }
    else if(result_items[0] == "FIND" || result_items[0] == "GET")
        //FIND WHERE NAME name
    {
        if(result_items[1] == "WHERE")
        {
            if(result_items[2] == "NAME")
            {
                if(result_items[3] != "")
                {
                    FINDquery(result_items[3]);             
                }
                else
                {
                    std::cout<<"No value entered, please try again."<<std::endl;
                }
            }
            else
            {
                std::cout<<"Invalid paramater entered, expected NAME, please try again."<<std::endl;
            }
        }
        else
        {
            std::cout<<"Invalid paramater entered, expected WHERE, please try again."<<std::endl;
        }
    }
    else if(result_items[0] == "DELETE" || result_items[0] == "REMOVE")
    {
        if(result_items[1] == "WHERE")
        {
            if(result_items[2] == "NAME")
            {
                if(result_items[3] != "")
                {
                    DELETEquery(result_items[3]);               
                }
                else
                {
                    std::cout<<"No value entered, please try again."<<std::endl;
                }
            }
            else
            {
                std::cout<<"Invalid paramater entered, expected NAME, please try again."<<std::endl;
            }
        }
        else
        {
            std::cout<<"Invalid paramater entered, expected WHERE, please try again."<<std::endl;
        }
    }




    return 1;
}

However, there is some serious problems with this it seems... For instance, it currently splits the items by their spaces and then uses them to check what commands are use, but what IF the query to search has spaces in it? e.g.

if(result_items[3] != "")
{
    FINDquery(result_items[3]);             
}

this works fine the query is

FIND WHERE NAME mynametosearch

but what about when its:

FIND WHERE NAME my name to search

It now has additional params exceeding result_items[3] and the search will fail.

What better approach can i take with this?

A few things: simply while (ss >> token) gets the space separated tokens, you should use vector::size() before indexing into the vector or ::at() for each access to avoid undefined behaviour, there's no need to check for empty string elements - your token parsing made sure there wouldn't be any in the vector, you would have noticed that if you'd printed the vector after reading it in - it's a(n obvious ;-P and) useful debugging step. — Tony Delroy
– Tony Delroy, Commented Mar 27, 2014 at 15:25

mkaes · Accepted Answer · 2014-03-27 14:48:11Z

1

Since you develop the language yourself, you should be able to write down the grammar of your language. If not you should really consider doing so.
Because as soon as you have the grammar you can use any parser library you want. I personally find boost::spirit very nice. But the good old flex/bison combo should also do it.

answered Mar 27, 2014 at 14:48

mkaes

14.2k10 gold badges56 silver badges75 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Parsing a custom structured query language?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related