iteratee enumPair incorrectly propagating the stream

[John Lato]

Thanks for these comments, I also agree that being consistent with
Prelude/Data.List would be the best option here.  I've pushed these to the
repo.

John

On Fri, Mar 4, 2011 at 3:42 AM, Maciej Wos <maciej.wos at gmail.com> wrote:

> +1 for zip, zip2, zip3, sequence ...
>
> One has to import Iteratee.ListLike qualified anyway (or hide a myriad
> of Prelude's functions). Plus, we already have filter, not enumFilter
> or take, not enumTake, etc. So, enumPair/enum2/enumZip wouldn't really
> fit in.
>
> -- Maciej
>
> On Fri, Mar 4, 2011 at 12:19 PM, Conrad Parker <conrad at metadecks.org>
> wrote:
> > On 3 March 2011 17:40, John Lato <jwlato at gmail.com> wrote:
> >> Thanks very much for this patch.  It's definitely a better
> implementation
> >> than the old version and it's been pushed.
> >>
> >> I wonder if enum3, enum4, and enum5 would be better names than the
> >> spelled-out version.  It would be consistent with the "zip" functions.
> And
> >> maybe enumPair should have its name changed also, maybe "enumZip", or
> just
> >> "eZip".  That would have to wait for a major release unfortunately, but
> the
> >> new name could be introduced now.
> >>
> >> Thoughts?  I'll put a new release on Hackage after I get some feedback
> about
> >> the names.
> >
> > consistency would be good, eg.:
> >
> > enum2, enum3, enum4 ...
> >
> > or
> >
> > enumZip, enumZip3, enumZip4 ...
> >
> > also it would be worth noting in the documentation that enumSequence
> > is a similar operation over lists.
> >
> > bikeshedding, I don't think eZip is very descriptive. OTOH if we want
> > short names to use qualified, perhaps just use zip, zip2, sequence
> > etc. (ie. I.zip, I.zip2, I.sequence ...)
> >
> > Conrad.
> >
> >>
> >> John
> >>
> >> On Thu, Mar 3, 2011 at 6:10 AM, Maciej Wos <maciej.wos at gmail.com>
> wrote:
> >>>
> >>> Hello,
> >>>
> >>> I've noticed that enumPair can behave a little strange. In some cases
> >>> it returns a part of the stream it has already consumed; this is tied
> >>> to the length of the chunk it is given:
> >>>
> >>> > run =<< I.enumPureNChunk [1..10 :: Int] 1 ((I.enumPair I.head (I.head
> >>
> >>> > I.head >> I.head)) >> I.stream2list)
> >>> [4,5,6,7,8,9,10]
> >>> > run =<< I.enumPureNChunk [1..10 :: Int] 2 ((I.enumPair I.head (I.head
> >>
> >>> > I.head >> I.head)) >> I.stream2list)
> >>> [4,5,6,7,8,9,10]
> >>> > run =<< I.enumPureNChunk [1..10 :: Int] 3 ((I.enumPair I.head (I.head
> >>
> >>> > I.head >> I.head)) >> I.stream2list)
> >>> [2,3,4,5,6,7,8,9,10]
> >>> > run =<< I.enumPureNChunk [1..10 :: Int] 5 ((I.enumPair I.head (I.head
> >>
> >>> > I.head >> I.head)) >> I.stream2list)
> >>> [2,3,4,5,6,7,8,9,10]
> >>>
> >>> I found the implementation in the library hard to follow, so I written
> >>> a new version from scratch. I also added enumThree, enumFour and
> >>> enumFive for a good measure -- I've used something like enumFive in my
> >>> code before and I think they would be nice to have.
> >>>
> >>> I've attached a patch with the changes.
> >>>
> >>> -- Maciej
> >>>
> >>> _______________________________________________
> >>> Iteratee mailing list
> >>> Iteratee at projects.haskell.org
> >>> http://projects.haskell.org/cgi-bin/mailman/listinfo/iteratee
> >>>
> >>
> >>
> >> _______________________________________________
> >> Iteratee mailing list
> >> Iteratee at projects.haskell.org
> >> http://projects.haskell.org/cgi-bin/mailman/listinfo/iteratee
> >>
> >>
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://projects.haskell.org/pipermail/iteratee/attachments/20110307/2d58641b/attachment.htm>