v1.2.0

#22

• Now handles any ~float{32|64} for the edge cases (was only edge case check for +Inf etc using actual float32/float64)
• Fix strange issue only appearing in armhf with maxuint32

Full Changelog: v1.1.0...v1.2.0

v1.1.0

What's Changed

#17

• Added a simpler safecast.Conv and safecast.MustConv that only deal with Integers (leaving Convert for backward compatibility and to/from floats)
• Fix issue with max int32/int64 conversion to float 32

Also

• Allow NaN<->NaN 32<->64 bits by @ldemailly in #15
• Optimize out 1 comparison for unsigned types by @hrissan in #13

Thanks @hrissan for digging up a number of interesting corner cases which we fixed now.

New Contributors

• @hrissan made their first contribution in #13

Full Changelog: v1.0.0...v1.1.0

v1.0.0

This is functionally same as v0.1.1 except with slightly smaller API surface (removed Negative() and SameSign()) as well as cover the odd case of MaxUInt64 which should error when converting to float64 and float32.

No (other) bugs have been found in the previous release so this should be now ready for 1.0.0 🎉

What's Changed

• Update README.md by @ldemailly in #6
• Switch back to main branch which has the token passing for codecov by @ldemailly in #7
• Adding test and fix for ^uint64(0) -> float by @ldemailly in #8
• Add link to gosec and CWE by @ldemailly in #9
• Create SECURITY.md by @ldemailly in #10
• Add wikipedia links to IEEE standard and check the 53 and 24 numbers by @ldemailly in #11

Full Changelog: v0.1.1...v1.0.0

v0.1.1: Initial release from new location

🎉

Full Changelog: 6cabc28...v0.1.1