Truncate fractional time precision

[moorereason]

Environment

toml++ version and/or commit hash:
v3 f3bd22b

Compiler:
gcc 11.1.0

C++ standard mode:
c++20

Target arch:
linux x86_64

Library configuration overrides:
n/a

Relevant compilation flags:
meson build_tt --buildtype=release -Dbuild_tt_encoder=true -Dbuild_tt_decoder=true -Dgenerate_cmake_config=false

Describe the bug

Given the TOML document 0=00:00:00.0000000000, the parser returns an error:

$ echo "0=00:00:00.0000000000" | tt_decoder
Error while parsing time: fractional component exceeds maximum precision of 9
        (error occurred at line 1, column 21 of 'stdin')

Steps to reproduce

Compile toml-test decoder and use the following on the command-line:

echo "0=00:00:00.0000000000" | tt_decoder

Additional Information

This may be an implementation detail that should be ignored, but I'm reporting it anyway. I expected the precision to be truncated instead of seeing an error.

The TOML spec:

Millisecond precision is required. Further precision of fractional seconds is implementation-specific. If the value contains greater precision than the implementation can support, the additional precision must be truncated, not rounded.

If this is the desired behavior, let me know. Thanks

[marzer]

Ah, thanks for letting me know. Indeed the parser should be behaving per the spec here, not issuing an error.

[marzer]

Fixed in v3 :)