C++ checking textbox value in application - c++

Question

I'm working on an app with a map and do not want a person to be able to enter an invalid point on the map (think Latitude and longitude).
Say you have a panel that looks something like this. Note not codewise how it looks, just how it would look like if you looked at the application:
[My great application]
Textbox1 [XX:XX:XX]
Textbox2 [XX:XX:XX]
Textbox3 [XX:XX:XX]
Pushbutton [APPLY]
My main problem
I have no way of knowing if what someone enters is right in the text boxes. My TextBox is expecting 3 sets integers [XX:XX:XX]. Nothing below 00, and nothing above 99 (strictly 2 digits). For this example I will only show on int to keep it short:
int myInt = NULL; //Don't want garbage data here
bool isValid = false;
//Gets input and sets textbox to myInt
std::string myString = std::to_string(myInt); //Cast int to a string
If(myString.length() - 1 == 0) //NULL intToString.length() == 1?
{
std::cout << "Incorrect, please try again" << std::endl;
}
else
{
isValid = true;
}
if(isValid)
{
//applyChanges(); <-- MY INTS WOULD GO INTO THIS
}
then I quickly realized that 0-9 are totally valid. (00, 01, ..., 09). So this wouldn't work.
My question
How am I able to test if what a user inputs is correct? It's not quite making sense to me. I'm not sure what else I can do besides test the length.
All in all I want to find a way to get a condition so that when all is well THEN isValid = true;
Notes
I can directly access all the data in the textboxes
I have the ability to check for gaining or loss of focus

Answer 1

Here's a solution I came up with. Let me know if I misunderstood your problem.
The algorithm is simple:
Splits the input into 3 parts;
Make sure each is 2 digits long;
Convert to an integer and make sure its within range;
#include <iostream>
#include <string>
#include <vector>
#include <sstream>
// Simple string split algorithm
std::vector<std::string> string_split(
std::string const& str,
char delimiter)
{
std::istringstream ss{str};
std::vector<std::string> result;
std::string line;
while(std::getline(ss, line, delimiter)) {
result.emplace_back(std::move(line));
}
return result;
}
// Validates the input
bool isValid(std::string const& input) {
// Split string on each ':'
// 00:00:00
// ^ ^
auto substr = string_split(input, ':');
// If there are not 3 parts, its malformed
if(substr.size() != 3)
return false;
// For reach result...
for(auto const& str : substr) {
// If it is not 2 digits long, error.
if(str.size() != 2)
return false;
try {
// Try converting it to an integer
// This throws on failure
auto value = std::stoi(str);
// make sure its within range
return value >= 0 && value <= 99;
} catch(...) {
// Failed to convert
return false;
}
}
return true;
}
int main()
{
// Some tests
std::vector<std::string> inputs{
"00:00:00",
"99:99:99",
"1:00:00",
"AA:00:00",
"0B:0000",
"0B:00c00",
};
// Run tests and print result
for(auto& input : inputs) {
if(isValid(input)) {
std::cout << input << " - Valid!"<< std::endl;
} else {
std::cout << input << " - Invalid!"<< std::endl;
}
}
}
Results:
$g++ -o main *.cpp
$main
00:00:00 - Valid!
99:99:99 - Valid!
1:00:00 - Invalid!
AA:00:00 - Invalid!
0B:0000 - Invalid!
0B:00c00 - Invalid!
Hope this helps,
Cheers.

Answer 2

I agree with markhc but you might also like using a regular expression way of doing things. The advantage is that the same code can be used to validate other type of input also and of course is much shorter.
Here is the code (parts stolen from markhc!):
#include <regex>
#include <string>
#include <iostream>
using namespace std;
int main()
{
string regx = R"([0-9]{2}:[0-9]{2}:[0-9]{2})";
smatch matches;
// some test data
std::vector<std::string> inputs{
"00:00:00",
"99:99:99",
"1:00:00",
"AA:00:00",
"0B:0000",
"0B:00c00",
};
// Run tests and print result
for (auto& input : inputs) {
if (regex_search(input, matches, regex(regx)))
{
std::cout << input << " - Valid!" << std::endl;
}
else {
std::cout << input << " - Invalid!" << std::endl;
}
}
}
This also produces as output:
00:00:00 - Valid!
99:99:99 - Valid!
1:00:00 - Invalid!
AA:00:00 - Invalid!
0B:0000 - Invalid!
0B:00c00 - Invalid!
The regular expression [0-9]{2}:[0-9]{2}:[0-9]{2} is explained below.
[0-9] means digit 0-9
{2} means 2 times
: means the separator you need
To show how flexible the regular expressions are:
..If you wanted a different format like 000:0000 for example you just have to change the regex to
[0-9]{3}:[0-9]{4} and the code remains the same.