Hi all, Looks like the initial code [1] for the Generators API in Apache Commons Functor was created based on the Python generators [2].
After playing for a while with Python generators, I realized many examples have a 'generator', a 'range' and a 'yield' statement. Then I looked for existing Java API's with generators, ranges and/or *yields* to have more ideas for [functor]'s Generator API. I found one [3] that requires some bytecode manipulation, and another one [4] that uses a threading mechanism to yield objects. I tried both of them [5] and the latter, java-yielder, seems to provide a useful Yield API, as well as Generator and Ranges. There are few modifications that I would like to see in its Range, like the possibility to make the boundary values inclusive or exclusive (like in Google Guava). As well as include a StoppableGenerator interface to the Generator API, so we could maintain the existing implementations in [functor] of Generators. If there are no objections, I will fill an issue in JIRA for enhancements on the Generator API in [functor], where we can further discuss on this issue as well as attach patches :-) I believe we could have similar separation in [functor], Generators, Ranges, and maybe, Yields. However, it would be good if someone else could have a look on java-yielder and other existing Yields implementations for Java. And if everybody thinks it is appropriate and the API is good enough, maybe we could ping the author of java-yielder and see if he would be willing to donate the code to ASF? It is licensed under the BSD-License [6], not sure if it is compatible with the Apache License. Looking forward to hearing your thoughts on this! [1] http://markmail.org/message/rtnn2xpudw3ygtvr#query:+page:1+mid:ynfmtd4shswtuvz7+state:results [2]http://www.dabeaz.com/generators/ [3] http://code.google.com/p/infomancers-collections/ [4] http://code.google.com/p/java-yield/ [5] https://github.com/kinow/yields-tests/tree/master/src/main/java/sample/yield [6] http://www.opensource.org/licenses/bsd-license.php Cheers, Bruno P. Kinoshita http://kinoshita.eti.br http://tupilabs.com >________________________________ > From: Adrian Crum <adrian.c...@sandglass-software.com> >To: Commons Developers List <dev@commons.apache.org> >Sent: Tuesday, 24 January 2012 7:21 PM >Subject: Re: [functor] CharacterRange, FloatRange, DoubleRange and open/closed >intervals > >After looking at the Lang 3 Range API, I realized an Iterator factory >should be kept external to the Range class: > >Range<Date> dateRange = Range.between(thisMonth, nextMonth); >Iterator<Date> dayIterator = RangeIteratorFactory.getInstance(dateRange, >Calendar.DATE); > >Range<Integer> intRange = Range.between(firstInt, lastInt); >Iterator<Integer> intSeqIterator = >RangeIteratorFactory.getInstance(intRange, 3); > >-Adrian > >On 1/24/2012 9:10 PM, Bruno P. Kinoshita wrote: >> >> Sounds great for me too :) >> >> >> Cheers, >> >> >> Bruno P. Kinoshita >> http://kinoshita.eti.br >> http://tupilabs.com >> >> >> ----- Mensagem original ----- >>> De: Adrian Crum<adrian.c...@sandglass-software.com> >>> Para: Commons Developers List<dev@commons.apache.org> >>> Cc: >>> Enviadas: Terça-feira, 24 de Janeiro de 2012 18:30 >>> Assunto: Re: [functor] CharacterRange, FloatRange, DoubleRange and >>> open/closed intervals >>> >>> On 1/24/2012 8:15 PM, Matt Benson wrote: >>>> Thanks for your interest, Adrian! Comments inline: >>>> >>>> On Tue, Jan 24, 2012 at 1:58 PM, Adrian Crum >>>> <adrian.c...@sandglass-software.com> wrote: >>>>> On 1/24/2012 6:58 PM, Matt Benson wrote: >>>>>> IMO the difference between [lang]'s and [functor]'s range >>> concepts is that >>>>>> those in [lang] are intended to be able to say whether a given >>> value is in >>>>>> the Range, but, with the exception of CharRange (this class is not >>> even >>>>>> public), they make no provision for iterating over members of the >>> range. I >>>>>> would therefore design a [functor]-targeted Range builder API to >>> allow the >>>>>> user to specify how next values are obtained. Range may not even be >>> the best >>>>>> word to describe the behavior we are trying to model here, so >>> let's be open >>>>>> to reinterpretation as we proceed. >>>>> From my perspective, Range + Step = Sequence. >>>> I'll buy that if noone else objects. :) >>>> >>>>> A Range iterator would be >>>>> useful for implementing mathematical sets. >>>> e.g. integer range a..b step 3 yields the sequence of integers n >>>> greater than or equal to a and less than or equal to b where (n - a) % >>>> 3 == 0 ? Is this useful? >>>> >>>> Matt >>> It would be useful for a mathematician. Another possibility is a Range >>> of Dates, with a step of one hour, or one day or... >>> >>> The Range class could accept an Iterator factory so various steps can be >>> implemented in client code. >>> >>> -Adrian >>> >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org >>> For additional commands, e-mail: dev-h...@commons.apache.org >>> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org >> For additional commands, e-mail: dev-h...@commons.apache.org >> > >--------------------------------------------------------------------- >To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org >For additional commands, e-mail: dev-h...@commons.apache.org > > > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
