jyapayne/Nim
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Document the difference between toFloat/toInt and type conversion (#17894)

Browse source
This commit is contained in:
Gabriel Huber 2021-04-30 22:39:25 +02:00 • committed by GitHub
commit 34a09574ce
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

5
lib/system.nim
View file

@ -1472,7 +1472,7 @@ type # these work for most platforms:
PInt32* = ptr int32 ## An alias for `ptr int32`.
proc toFloat*(i: int): float {.noSideEffect, inline.} =
## Converts an integer `i` into a `float`.
## Converts an integer `i` into a `float`. Same as `float(i)`.
##
## If the conversion fails, `ValueError` is raised.
## However, on most platforms the conversion cannot fail.
@ -1494,7 +1494,8 @@ proc toInt*(f: float): int {.noSideEffect.} =
##
## Conversion rounds `f` half away from 0, see
## `Round half away from zero
## <https://en.wikipedia.org/wiki/Rounding#Round_half_away_from_zero>`_.
## <https://en.wikipedia.org/wiki/Rounding#Round_half_away_from_zero>`_,
## as opposed to a type conversion which rounds towards zero.
##
## Note that some floating point numbers (e.g. infinity or even 1e19)
## cannot be accurately converted.