Closed GoogleCodeExporter closed 8 years ago
It sounds a good idea, but I would like to better define the features that each
responsibility should cover before. For example, as I replied to the issue 2,
the
difference between select and filter is not clear.
Original comment by mario.fu...@gmail.com
on 19 Jan 2009 at 8:40
it's right. filter is an alias of select.. or viceversa
Original comment by luca.marrocco
on 20 Jan 2009 at 3:37
Original comment by mario.fu...@gmail.com
on 26 Jun 2009 at 7:10
Hi,
Splitting Lamdaj class can be a good and bad idea. It is a facade, so it is
reasonable to be very big class. Splitting them, IMHO, could decrease the Lamdaj
usability.
Original comment by filosgan...@gmail.com
on 27 Jun 2009 at 1:11
I agree with filosganga80, it can be good and it can be a bad idea. Mockito
uses the
same facade strategy; one large "entry point" class with static methods for all
(or
at least most of) the features of the library. In Eclipse I can set up
ch.lambdaj.Lambda as a "favorite" for static import lookups, meaning I get code
completion by just typing in the method names. If the Lambda class is split up,
I
must set up more classes as "favorites".
Original comment by flobakk
on 16 Sep 2009 at 7:20
my suggestion is to divide lambdaj logic internally
* ch.lambdaj.lang.Join
* ch.lambdaj.lang.Select
* ch.lambdaj.lang.Filter
and have a facade ch.lambdaj.LambdaJ that remain the entry point of the
functionality.
Original comment by luca.marrocco
on 16 Sep 2009 at 7:25
I agree that splitting the Lambda class will make the library less usable.
Moreover the class ch.lambdaj.Lambda is already not much more than a facade that
calls the features internally implemented by lambdaj. In other words each
method of
the Lambda class is typically one or two line of codes, so I don't see a need to
create another layer of delegation.
Original comment by mario.fu...@gmail.com
on 10 Oct 2009 at 3:57
Original issue reported on code.google.com by
luca.marrocco
on 9 Jan 2009 at 10:44