Documentation: BigFloat #5639
Documentation: BigFloat #5639
Conversation
src/big/big_float.cr
Outdated
| @@ -217,11 +243,21 @@ struct BigFloat < Float | |||
| mpf | |||
| end | |||
|
|
|||
| # Returns an inspected representation of self. | |||
src/big/big_float.cr
Outdated
| def inspect(io) | ||
| to_s(io) | ||
| io << "_big_f" | ||
| end | ||
|
|
||
| # Returns a string representation of self. |
| # Raises `ArgumentError` if the string doesn't denote a valid float. | ||
| # | ||
| # ``` | ||
| # BigFloat.new("-0.123456789101101987654321") # => -0.123456789101101987654321 |
There was a problem hiding this comment.
nope, it's added only in #inspect.
There was a problem hiding this comment.
Yeah, that's what the expression value should show, see the example of the arg-less constructor above.
There was a problem hiding this comment.
@straight-shoota Don't you like BigFloat.new.inspect # => 0_big_f ?
There was a problem hiding this comment.
I'm not sure if I like it in general, because it looks like a literal but there is none.
However, if you use that format for the value of the arg-less constructor example, this one should be equal.
There was a problem hiding this comment.
Another reason for removing suffixes... there are none for big numbers. Unless we inspect them to BigInt.new("#{string}") but that's unreadable and ugly.
|
Is there anything else to be done here? |
src/big/big_float.cr
Outdated
| @@ -217,11 +243,21 @@ struct BigFloat < Float | |||
| mpf | |||
| end | |||
|
|
|||
| # Returns an inspected representation of `self`. | |||
There was a problem hiding this comment.
Returns a human-readable representation of self.
src/big/big_float.cr
Outdated
| def inspect(io) | ||
| to_s(io) | ||
| io << "_big_f" | ||
| end | ||
|
|
||
| # Returns a string representation of `self`. |
There was a problem hiding this comment.
Specify base 10.
We really should support other bases here, but thats another issue.
There was a problem hiding this comment.
Would you like a # TODO: support other bases here ?
There was a problem hiding this comment.
@RX14 Actually on reflection does it really make much sense in talking about float representation in anything other than base 10, and by extension should I maybe revert the change to specify base 10 ? Certainly there doesn't appear to be any use or mention of alternate bases in Float32 / Float64 for instance. Perhaps you were thinking of BigInt?
+ # Returns a string representation (base 10) of `self`.
+ #
+ # ```
+ # BigFloat.new("0.123456789101101987654321").to_s # => 0.123456789101101987654321
+ # ```
def to_s(io : IO)
|
Specified base 10, rebased and squashed |
|
Is there anything else to be done here? |
|
@mjago lgtm @RX14 @straight-shoota is this OK? |
|
Please don't close things due to inactivity. Especially because you've deleted the original repo so I cannot reopen this PR. |
Documentation for BigFloat