Proposal: Add formatter for std::error_code.

[phprus]

The std::error_code has are operator<< (https://en.cppreference.com/w/cpp/error/error_code/operator_ltlt), which is equivalent os << ec.category().name() << ':' << ec.value() .

Formatting with ostream is 4-5 (C++11) or 11-24 (C++17) times slower than the custom formatter class.

Simpe and possibly incorrect implementation: https://github.com/phprus/fmt-bench/blob/format-error_code/test_formatter.cpp (I agree that this code are licensed under the {fmt} license, and agree to future changes to the licensing.)
Benchmark (https://github.com/phprus/fmt-bench/tree/format-error_code):

AppleClang 11.0.0.11000033 C++11

Running ./fmt_test
Run on (8 X 2700 MHz CPU s)
CPU Caches:
  L1 Data 32 KiB (x4)
  L1 Instruction 32 KiB (x4)
  L2 Unified 256 KiB (x4)
  L3 Unified 6144 KiB (x1)
Load Average: 1.79, 1.73, 1.69
-----------------------------------------------------------------
Benchmark                       Time             CPU   Iterations
-----------------------------------------------------------------
FMTIostream/22                369 ns          368 ns      1953098
FMTIostreamCompile/22         385 ns          384 ns      1863938
FMTFormatter/22              75.9 ns         75.8 ns      9392947
FMTFormatterCompile/22       79.6 ns         79.5 ns      8838830

AppleClang 11.0.0.11000033 C++17

Running ./fmt_test
Run on (8 X 2700 MHz CPU s)
CPU Caches:
  L1 Data 32 KiB (x4)
  L1 Instruction 32 KiB (x4)
  L2 Unified 256 KiB (x4)
  L3 Unified 6144 KiB (x1)
Load Average: 1.68, 1.71, 1.69
-----------------------------------------------------------------
Benchmark                       Time             CPU   Iterations
-----------------------------------------------------------------
FMTIostream/22                389 ns          388 ns      1854806
FMTIostreamCompile/22         380 ns          378 ns      1794761
FMTFormatter/22              33.7 ns         33.7 ns     20944226
FMTFormatterCompile/22       16.4 ns         16.4 ns     42452029

[vitaut]

Seems like a good idea, could you submit a PR?

[phprus]

Yes, but I'm not sure if my implementation is correct.

I don't know how to write the parse method correctly to follow the logic of the library.

[vitaut]

I'm not sure if my implementation is correct.

Looks OK to me. You could simplify parse by just returning ctx.begin().

[phprus]

PR: #2270