feat: Make it easy to convert data to IDLValue and candid text format

[bitdivine]

Overview
It is not well documented how to convert types that implement CandidType into an IDLValue.

Requirements
It is easy to convert instances of Rust types into IDLValues using the best available method.

Considered Solutions

• Provide a Candid library method for the conversion. (implemented in this PR)
• Document the current approach. However the current approach seems inefficient and can probably be improved.

Recommended Solution
Provide a library method for converting data to Candid.

The current method - serializing and parsing - can probably be improved greatly, so it seems preferable to offer a conversion method in the candid library and then improve that method rather than document a method that is sub-optimal and then expect developers to update their code when a better method is documented.

Considerations

• I think that this will have no negative influence on performance.
• This makes conversion of data to text-format candid a one-liner: IDLValue::try_from_candid_type(data).unwrap().to_string()

[github-actions]

Benchmark for adea395

Click to view benchmark

Test	Base	PR	%
Blob/&str	605.5±76.77µs	615.2±94.71µs	+1.60%
Blob/ByteBuf	347.8±89.01µs	342.1±92.79µs	-1.64%
Blob/Bytes	397.8±80.24µs	393.3±74.29µs	-1.13%
Blob/String	633.6±242.90µs	638.5±249.09µs	+0.77%
Collections/vec (text, nat)	51.5±1.85ms	51.5±0.90ms	0.00%
Collections/vec int	23.2±0.25ms	23.2±0.78ms	0.00%
Collections/vec int64	16.2±0.42ms	15.7±0.09ms	-3.09%
Collections/vec nat8	12.1±0.81ms	12.0±0.09ms	-0.83%
option list/1024	978.0±12.43µs	971.9±5.77µs	-0.62%
profiles/1024	1896.0±16.02µs	1941.3±31.64µs	+2.39%
variant list/1024	795.2±11.03µs	793.6±4.43µs	-0.20%

[chenyan-dfinity]

annotate_types is a best-effort implementation, notably subtyping is not supported. A better way is to use from_bytes_with_types which implements everything from the official deserializer.

[github-actions]

Benchmark for c8a2efc

Click to view benchmark

Test	Base	PR	%
Blob/&str	610.2±103.64µs	591.1±90.18µs	-3.13%
Blob/ByteBuf	347.6±89.32µs	353.2±85.53µs	+1.61%
Blob/Bytes	381.2±64.49µs	354.1±61.29µs	-7.11%
Blob/String	324.7±136.71µs	362.0±129.40µs	+11.49%
Collections/vec (text, nat)	53.1±2.59ms	53.0±0.66ms	-0.19%
Collections/vec int	23.0±0.10ms	22.9±0.09ms	-0.43%
Collections/vec int64	16.2±0.06ms	15.8±0.22ms	-2.47%
Collections/vec nat8	11.6±0.99ms	11.5±0.10ms	-0.86%
option list/1024	996.8±21.77µs	985.6±11.31µs	-1.12%
profiles/1024	1865.3±29.68µs	1865.1±35.30µs	-0.01%
variant list/1024	789.9±9.75µs	795.2±10.88µs	+0.67%

[github-actions]

Benchmark for 0c3da76

Click to view benchmark

Test	Base	PR	%
Blob/&str	621.2±91.03µs	607.3±89.08µs	-2.24%
Blob/ByteBuf	357.8±86.55µs	354.9±89.94µs	-0.81%
Blob/Bytes	390.0±62.84µs	363.6±60.91µs	-6.77%
Blob/String	331.5±138.84µs	366.5±130.36µs	+10.56%
Collections/vec (text, nat)	54.0±4.82ms	52.9±1.28ms	-2.04%
Collections/vec int	22.9±0.13ms	22.8±0.14ms	-0.44%
Collections/vec int64	16.7±0.09ms	16.1±0.09ms	-3.59%
Collections/vec nat8	11.6±0.44ms	11.5±0.13ms	-0.86%
option list/1024	984.1±18.32µs	987.4±19.58µs	+0.34%
profiles/1024	1927.1±49.76µs	1934.6±58.63µs	+0.39%
variant list/1024	803.1±13.60µs	809.4±15.99µs	+0.78%

[bitdivine]

@chenyan-dfinity Thank you!