new algorithm(xjb) for double to shortest decimal.

[xjb714]

hello，

link : https://github.com/xjb714/f2dec_bench

If it doesn't work, please give me feedback. It seems a little difficult to prove that the algorithm is correct, but I can't find a counterexample. The algorithm may be correct, and I have tested a large number of data with correct output. I have tested all the float values and the output results are correct. If you are interested, you can study my algorithm code, and the function name is xjb_f64_to_dec. If you know that others are interested in this kind of algorithm, you can also share it with others.

Thanks for reading.