search

fastfloat / fast_float

Fast and exact implementation of the C++ from_chars functions for number types: 4x to 10x faster than strtod, part of GCC 12, Chromium, Redis and WebKit/Safari
Apache License 2.0
1.54k stars 132 forks source link

max_mantissa_double problem #258

Closed wjr-z closed 2 months ago

wjr-z commented 2 months ago

1 << 53 that's equal with 0x20000000000000, not 0x10000000000000. Maybe this's an error ?

lemire commented 2 months ago

If you believe that there is a bug, please provide a test case where the library gives an incorrect result. Note that this library has been in production for many years without any issue.

lemire commented 2 months ago

Oh I see. Reopening.

wjr-z commented 2 months ago

哦，我明白。重新。

I mean an error of performance. Because the range of his fast path is smaller than the correct one

lemire commented 2 months ago

Yes. I think that you are correct.