[Haskell-cafe] Re: iteratee: Do I need to roll my own?

Valery V. Vorotyntsev valery.vv at gmail.com
Wed Mar 31 14:15:39 EDT 2010

> I'm looking at iteratee as a way to replace my erroneous and really
> inefficient lazy-IO-based backend for an expect like Monad DSL I've
> been working for about 6 months or so now on and off.
> The problem is I want something like:
> expect "some String"
> send "some response"
> to block or perhaps timeout, depending on the environment, looking for
> "some String" on an input Handle, and it appears that iteratee works
> in a very fixed block size.

Actually, it doesn't. It works with what enumerator gives him.
In case of `enum_fd'[1] this is a fixed block, but generally this is
a ``value'' of some ``collection''[2].  And it is up to programmer to
decide of what should become a value.

  [1] http://okmij.org/ftp/Haskell/Iteratee/IterateeM.hs
  [2] http://okmij.org/ftp/papers/LL3-collections-enumerators.txt

> While a fixed block size is ok, if I can put back unused bytes into
> the enumerator somehow (I may need to put a LOT back in some cases,
> but in the common case I will not need to put any back as most
> expect-like scripts typically catch the last few bytes of data sent
> before the peer is blocked waiting for a response...)

I don't quite get this ``last few bytes'' thing. Could you explain?

I was about writing that there is no problem with putting data back to
Stream, and referring to head/peek functions...  But then I thought,
that the ``not consuming bytes from stream'' approach may not work
well in cases, when the number of bytes needed (by your function to
accept/reject some rule) exceeds the size of underlying memory buffer
(4K in current version of `iteratee' library[3]).

  [3] http://hackage.haskell.org/packages/archive/iteratee/0.3.4/doc/html/src/Data-Iteratee-IO-Fd.html

Do you think that abstracting to the level of _tokens_ - instead of
bytes - could help here? (Think of flex and bison.)  You know, these
enumerators/iteratees things can be layered into
_enumeratees_[1][4]... It's just an idea.

  [4] http://ianen.org/articles/understanding-iteratees/

> Otherwise, I'm going to want to roll my own iteratee style library
> where I have to say "NotDone howMuchMoreIThinkINeed" so I don't over
> consume the input stream.

What's the problem with over-consuming a stream? In your case?

BTW, this `NotDone' is just a ``control message'' to the chunk
producer (an enumerator):

    IE_cont k (Just (GimmeThatManyBytes n))

> Does that even make any sense?  I'm kind of brainstorming in this
> email unfortunately :-)

What's the problem with brainstorming? :)



More information about the Haskell-Cafe mailing list