Re: Newbie needs help

• Previous message: Cephas: "NaN question"
• In reply to: Rich Strang: "Newbie needs help"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Date: Tue, 18 Nov 2003 01:57:36 +0000 (UTC)

In article <hKcub.3211$vb6.28441643@news-text.cableinet.net>,
Rich Strang <riky_rang@hotmail.com> wrote:
>
>[...] Without the loop the program works fine with the
>loop the cout doesn't seem to work,
[snip]
> if(cInput[0] != 'e' && cInput[1] != 'x' && cInput[2] != 'i' && cInput[3]
>!= 't' )

Here's your problem. You want to exit the loop when the user types
"exit", right? That means you want to exit when

cInput[0] == 'e' && cInput[1] == 'x' && cInput[2] == 'i' &&
cInput[3] == 't'

Your loop design means that you actually need the *opposite* of this
condition, which is *not* what you wrote above, but rather

cInput[0] != 'e' || cInput[1] != 'x' || cInput[2] != 'i' ||
cInput[3] != 't'

Think about it a bit, and if you don't "get" it, look up DeMorgan's Rule
in a logic primer or a Google search.

Now, here's a piece of general advice. You must be using an old C++
textbook (pre-1998 or thereabouts), because standard header files don't
have .h in them, and the standard came into existence around 1998.
Post-1998 standard C++ has stuff in it that greatly simplifies what you're
trying to do. Parsing text character by character just to split up
whitespace-separated tokens is *sooo* 1990s. ;-)

Here's my stab at what I think you're trying to accomplish, in more or
less 21st-century C++ style:

#include <iostream>
#include <cctype>
#include <vector>
#include <string>
#include <sstream>

using namespace std;

// no need for your GetLine function, standard C++ has one already
void ParseLine (const string& cInput, vector<string>& cTokens);
void PrintTokens (const vector<string>& cTokens);

int main ()
{
    // strings are like arrays of char on steroids
    string cInput;
    // vectors are like arrays on steroids
    vector<string> cTokens;
    int iExit = 0;
    while (iExit == 0)
    {
        cout << "$ ";
        getline (cin, cInput); // cInput automatically expands as needed
                                // to accommodate the input
        ParseLine (cInput, cTokens);
        if (cTokens[0] == "exit") // can compare strings w/o any fuss
        {
            iExit = 1;
        }
        else
        {
            PrintTokens (cTokens);
        }
    }
    return 0;
}

void ParseLine (const string& cInput, vector<string>& cTokens)
{
    cTokens.resize(0); // empty out the vector from the previous line
    istringstream linestream (cInput);
    string token;
    // read from an istringstream just like you read from cin;
    // this reads whitespace-separated tokens until it hits the
    // end of the stream
    while (linestream >> token)
    {
        cTokens.push_back(token); // cTokens automatically expands
                                   // as needed
    }
}

void PrintTokens (const vector<string>& cTokens)
{
// note vectors and strings can tell you how big they are, so you
// don't have to keep track of it yourself
    for (int iIndex1 = 0; iIndex1 < cTokens.size(); iIndex1++)
    {
        for (int iIndex2 = 0; iIndex2 < cTokens[iIndex1].length();
          iIndex2++)
        {
            // isalnum means "isalpha || isdigit"
            if (isalnum (cTokens[iIndex1][iIndex2]))
            {
                cout << cTokens[iIndex1][iIndex2];
            }
        }
        cout << '\n';
    }
}

Personally, I'd write a separate function to strip the non-alphanumeric
stuff out of a token:

string StripNonAlnum (const string& token)
{
    string newToken;
    for (int k = 0; k < token.size(); ++k)
    {
        if (isalnum(token[k]))
        {
            newToken += token[k];
        }
    }
    return newToken;
}

void PrintTokens (const vector<string>& cTokens)
{
    for (int iIndex1 = 0; iIndex1 < cTokens.size(); iIndex1++)
    {
        cout << StripNonAlnum (cTokens[iIndex1]) << '\n';
    }
}

-- 
Jon Bell <jtbellap8@presby.edu>                     Presbyterian College
Dept. of Physics and Computer Science        Clinton, South Carolina USA

• Previous message: Cephas: "NaN question"
• In reply to: Rich Strang: "Newbie needs help"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]