C/C++ Decimal Formatting: From printf to std::format

In 2015, I answered a question on Stack Overflow in Portuguese about limiting decimal places in C and C++. It scored 5 upvotes.

The 2015 Answer: Two Approaches

C-style with printf:

double value = 3.14159;
printf("%.2f\n", value); // Output: 3.14

C++ with iomanip:

#include <iostream>
#include <iomanip>

double value = 3.14159;
std::cout << std::fixed << std::setprecision(2) << value << std::endl; // 3.14

Both worked, but each had its annoyances. printf required memorizing format specifiers. The iomanip approach was verbose and had the quirky behavior of std::fixed and std::setprecision persisting across subsequent output operations — easy to accidentally affect other output.

The 2026 Approach: std::format

C++20 introduced std::format, which brought Python’s f-string-style formatting to C++:

#include <format>

double value = 3.14159;
std::string result = std::format("{:.2f}", value); // "3.14"
std::println("{:.2f}", value); // C++23: prints with newline

The format string syntax is consistent and learnable. If you know Python’s f"{value:.2f}", you already know std::format("{:.2f}", value).

Format Specifier Reference

Format	Result for 3.14159	Meaning
{:.2f}	3.14	Fixed, 2 decimal places
{:.4f}	3.1416	Fixed, 4 decimal places
{:.2e}	3.14e+00	Scientific notation
{:.2g}	3.1	General (shortest)
{:8.2f}	3.14	Width 8, 2 decimals

Key Takeaway

printf("%.2f") still works and is the right choice in C or when interfacing with C APIs. In modern C++, std::format offers cleaner syntax, type safety at compile time, and no risk of format string vulnerabilities. C++23’s std::println makes it even simpler.