Format woes

[seeplus]

int main() {
	std::cout <<  std::format("{:< 8f}", .111111);
}

displays 0.111111 but I need .111111 (no initial 0). Is this possible just using format or do I need to manipulate the produced string?

[George PlusPlus]

With a quick 'net search using multiple search terms I'd have to opine there is no built-in method to not have that leading zero, in std::format or printf.

Interestingly coincidental many of the results showed how to add leading zeros.

You could create a custom formatter, essentially treating your decimal number as a custom type.

https://www.cppstories.com/2022/custom-stdformat-cpp20/

That looks to be a lot more work than I'd care to do. Whacking off any leading zeros by manipulating the string is probably the quickest and simplest route to go.

Creating a custom formatter is quite beyond my self-taught hobbyist capabilities currently. I'd have to spend a lot of time with a hammer, blow-torch and crowbar 55 gallons of lubricant, 40 feet of rubber tubing and a Yak to do some tests on creating one.

My current interests are ones that lie in different directions. Eventually I might wander back and look at creating custom formatters, when dealing with custom class output.

[seeplus]

Whacking off any leading zeros by manipulating the string is probably the quickest and simplest route to go.

That's what I've done. But I was interested if there was a way of doing it just by using a a format string. It seems not...

You could create a custom formatter ... That looks to be a lot more work than I'd care to do

Yeah - but IMO that's massive over-engineering for the simple issue in this case.

[George PlusPlus]

I'm more than likely just not understanding things but maybe you could try std::format_to or std::format_to_n?

https://en.cppreference.com/w/cpp/utility/format/format_to
https://en.cppreference.com/w/cpp/utility/format/format_to_n

TBH I look at those examples and to me they are as complicated as writing a custom formatter from scratch. Nigh incomprehensible gobbledygook and argyle-gargle flummery.

[seeplus]

format_to sends it's output to an output buffer iterator rather than returning a string. format-to_n is format_to with the maximum number of characters to be written to the buffer also specified. They both use the same format specifiers as format.

[Manga]

How about a little mess of code like this...

#include <iostream>
#include <string>
using namespace std;

int main()
{
  
  double a = 0.110000;
  
  
  if(to_string(a)[0] == '0') {
    for(int i = 1; i < to_string(a).length(); i++) {
      cout << to_string(a)[i];   
    }
  }
  else {
     cout << to_string(a);   
  }
 
  return 0;
}

[helios]

If you're going to write less code then write less code.

auto s = std::to_string(a);
std::cout << (s.size() > 0 && s[0] == '0' ? s.substr(1) : s);

[Manga]

ok less code...

auto s = std::to_string(a);
std::cout << (s.size() && s[0] == '0' ? s.substr(1) : s);

I win...

[seeplus]

std::cout << s.substr(s.size() && s[0] == '0');

[mbozzi]

std::cout << s.substr(*s.c_str() == '0');