On Saturday, 28 July 2018 at 16:02:22 UTC, Gary Willoughby wrote:
I just saw this on hacker news:
We present Ryū, a new routine to convert binary floating point
numbers to their decimal representations using only fixed-size
integer operations, and prove its correctness. Ryū is simpler
and approximately three times faster than the previously
fastest implementation.
Links:
https://pldi18.sigplan.org/event/pldi-2018-papers-ry-fast-float-to-string-conversion
https://www.youtube.com/watch?v=kw-U6smcLzk
https://github.com/ulfjack/ryu
Thanks, it looks very interesting (and hey, it is a bit too late,
I've just finished to implement errol in D today :)).
After a fast look at the texts, I can't find information about
the optimality. They compare ryu mostly to grisu and grisu
produces always correct, but not always optimal results, so it
fallbacks to a slower algorithm for some values.
It would be also nice to have some paper; I see only videos.
Maybe a paper will come soon.
