Use Advanced Search to search the entire archive.
[jsr363-experts] Re: Quantity.doubleValue(...) ?
- From: Werner Keil <
>
- To: "
" <
>
- Subject: [jsr363-experts] Re: Quantity.doubleValue(...) ?
- Date: Mon, 15 Dec 2014 12:37:00 +0100
I fixed the only remaining sentence in the spec where doubleValue(Unit)
etc. were still mentioned in the context of the API Quantity.
Also moved some footnote into the RI level AbstractQuantity. Please remove
if you feel it no longer matters.
Werner
On Mon, Dec 15, 2014 at 12:14 PM, Werner Keil
<
>
wrote:
>
>
I recall, especially doubleValue() was found of no real value since you
>
get it directly via the Number of getValue().
>
>
The API must not use any of the new Java SE 8 stuff like Optional (in
>
fact, for a case like Range in uom-se it turned out to be flawed and
>
useless, so the range currently has to return null if there's no upper or
>
lower bound;-) so should there be a case where getValue() of a Quantity
>
would be null, what should a parallel doubleValue() then return?;-O
>
Or would we throw an exception?
>
>
Similar with doubleValue(Unit) this is just a convenience method for
>
to(Unit).getValue().doubleValue().
>
And should anything other than the obvious unit argument be null, then we
>
have the same issue here.
>
There are maybe just one or two concrete classes like
>
>
https://github.com/unitsofmeasurement/unit-ri/blob/master/src/main/java/tec/units/ri/quantity/DoubleQuantity.java
>
where doubleValue() directly based on the double member variable could
>
help.
>
>
You'll notice, many of the low level methods went into
>
https://github.com/unitsofmeasurement/unit-ri/blob/master/src/main/java/tec/units/ri/AbstractQuantity.java
>
(I am not sure, why one of them is public, either it's used from another
>
RI package or it is a mistake) so the concrete and final (value like)
>
classes that benefit most like DoubleQuantity may override them as public
>
while others don't (where it rarely makes sense)
>
>
It may clutter the API, see JSR-275, but if there is one or another method
>
we really feel like adding through the next stages, that can evolve.
>
>
Werner
>
>
On Mon, Dec 15, 2014 at 5:26 AM, Martin Desruisseaux <
>
>
>
wrote:
>
>
>
> The specification contains a discussion about the following Quantity
>
> methods, which existed in JSR-275:
>
>
>
> - double doubleValue();
>
> - double doubleValue(Unit<Q> unit);
>
>
>
> Those methods do not exist in current JSR-363 Quantity interface. I do
>
> not remember what was the rational for removing them. Is there plan to
>
> re-introduce them, or shall we remove the discussion about those methods
>
> from the spec.?
>
>
>
> Martin
>
>
>
>